Volume 72, Number 1, January-February 2020
Siriraj Medical Journal
SMJ
ISSN 2629-995XE-ISSN 2228-8082
ORIGINAL ARTICLE
1
Achievement of Metabolic Goals among Different Health
Insurance Schemes in Thai Patients with Type 2 Diabetes
Mellitus: a Nationwide Study
Lukana Preechasuk, et al
10 Breastmilk Macronutrient Levels and Infant Growth
During the First Three Months: Cohort Study
Martini, et al.
18 Suitability of Enhanced Recovery after Surgery (ERAS)
Protocols for Elderly Colorectal Cancer Patients
Gregory Heng, et al.
24 Multifaceted Intervention to Improve the Quality of Care
for Postpartum Hemorrhage
Tippawan Liabsuetrakul, et al.
33 Role of VEGF and Doppler Sub-endometrial Parameters as
Predictors of Successful Implantation in ICSI Patients
Wafaa Salih Rahmatullah, et al.
41 The Good Skin Care Practices and Emollient Use since
Early Infancy as the Primary Prevention of Infantile Atopic
Dermatitis among Infants at Risk:
Pajaree Thitthiwong, et al.
47 The Impact of Active Nutritional Support for Head and
Neck Cancer Patients Receiving Concurrent
Chemoradiotherapy
Vutisiri Veerasarn, et al
59 Depression and Quality of Life in Spinal Cord Injury
Patients Living in the Community After Hospital Discharge
Cholavech Chavasiri, et al.
67 Comparison of Computed Tomography Angiography (CTA)
Findings in Post-Endovascular Aortic Aneurysm Repair
CTA Between Persistent and Transient Type II Endoleak
Karnkawin Patharateeranart, et al
74 Performance of Two Commercial Dengue NS1 Rapid Tests
for the Diagnosis of Adult Patients with Dengue Infection
Ekkarat Wongsawat, et al.
79 Comparison of an Automated Thermodynamic Treatment
System (LipiFlow) and Warm Compresses for the Treatment
of Moderate Severity of Meibomian Gland Dysfunction
Wipawee Booranapong, et al.
REVIEW ARTICLE
87 Overview of Surgical Therapies for Obstructive Sleep
Apnea: a Concise Review Literature
Lee Kian Khoo, et al
www.smj.si.mahidol.ac.th
E-mail: sijournal@mahidol.ac.th
International Association of Surgeons
Gastroenterologists & Oncologists
Thailand Chapter
Thai Association for Gastrointestinal
Endoscopy
Indexed by
By Karnkawin Patharateeranart, et al.
SIRIRAJ MEDICAL JOURNAL
First Editor: Ouay Ketusinh Emeritus Editors: Somchai Bovornkitti, Adulya Viriyavejakul, Sommai Toongsuwan,
Nanta Maranetra, Niphon Poungvarin, Prasit Watanapa, Vithya Vathanophas, Pipop Jirapinyo, Sanya Sukpanichnant,
Somboon Kunathikom
Executive Editor: Prasit Watanapa Editorial Director: Manee Rattanachaiyanont
Managing Editor: Gulapar Srisawasdi, Chenchit Chayachinda
Editor-in-Chief: awatchai Akaraviputh
Associate Editor: Varut Lohsiriwat, Prapat Wanitpongpan Online Editor: Puttinun Patpituck
SIRIRAJ MEDICAL JOURNAL is published bimonthly, 6 issues a year (Jan-Feb, Mar-Apr, May-Jun, Jul-Aug, Sep-Oct and Nov-Dec)
and distributed by the end of the last month of that issue.
SIRIRAJ MEDICAL JOURNAL is listed as a journal following the Uniform Requirements for Manuscripts Submitted to Biomedical Journals (URM)
by the International Committee of Medical Journal Editors (ICMJE) since 9 July 2010 [http://www.icmje.org/journals.html].
Philip Board (Australian National University, Australia)
Richard J. Deckelbaum (Columbia University, USA)
Yozo Miyake (Aichi Medical University, Japan)
Yik Ying Teo (National University of Singapore, Singapore)
Harland Winter (Massachusetts General Hospital, USA)
Philip A. Brunell (State University of New York At Bualo, USA)
Noritaka Isogai (Kinki University, Japan)
Yuji Murata (Aizenbashi Hospital, Japan)
Keiichi Akita (Tokyo Medical and Dental University Hospital, Japan)
Shuji Shimizu (Kyushu University Hospital, Japan)
David S. Sheps (University of Florida, USA)
Robin CN Williamson (Royal Postgraduate Medical School, UK)
Tai-Soon Yong (Yonsei University, Korea)
Anusak Yiengpruksawan (e Valley Robotic Institute, USA)
Stanlay James Rogers (University of California, San Francisco, USA)
Kyoichi Takaori (Kyoto University Hospital, Japan)
Tomohisa Uchida (Oita University, Japan)
Yoshiki Hirooka (Nagoya University Hospital, Japan)
Hidemi Goto (Nagoya University Graduate School of Medicine, Japan)
Kazuo Hara (Aichi Cancer Center Hospital, Japan)
Shomei Ryozawa (Saitama Medical University, Japan)
Christopher Khor (Singapore General Hospital, Singapore)
Yasushi Sano (Director of Gastrointestinal Center, Japan)
Mitsuhiro Kida (Kitasato University & Hospital, Japan)
Seigo Kitano (Oita University, Japan)
Ichizo Nishino (National Institute of Neuroscience NCNP, Japan)
Masakazu Yamamoto (Tokyo Women’s Medical University, Japan)
Dong-Wan Seo (University of Ulsan College of Medicine, Korea)
George S. Baillie (University of Glasgow, UK)
G. Allen Finley (Delhousie University, Canada)
Sara Schwanke Khilji (Oregon Health & Science University, USA)
Matthew S. Dunne (Institute of Food, Nutrition, and Health, Switzerland)
Marianne Hokland (University of Aarhus, Denmark)
Marcela Hermoso Ramello (University of Chile, Chile)
Ciro Isidoro (University of Novara, Italy)
Moses Rodriguez (Mayo Clinic, USA)
Robert W. Mann (University of Hawaii, USA)
Wikrom Karnsakul (Johns Hopkins Children’s Center, USA)
Frans Laurens Moll (University Medical Center Ultrecht, Netherlands)
James P. Dolan (Oregon Health & Science University, USA)
John Hunter (Oregon Health & Science University, USA)
Nima Rezaei (Tehran University of Medical Sciences, Iran)
Dennis J. Janisse (Subsidiary of DJO Global, USA)
Folker Meyer (Argonne National Laboratory, USA)
David Wayne Ussery (University of Arkansas for Medical Sciences, USA)
Intawat Nookaew (University of Arkansas for Medical Sciences, USA)
Victor Manuel Charoenrook de la Fuente
(Centro de Oalmologia Barraquer, Spain)
Karl omas Moritz
(Swedish University of Agricultural Sciences, Sweden)
Nam H. CHO (University School of Medicine and Hospital, Korea)
www.smj.si.mahidol.ac.th
SMJ
Statistician: Saowalak Hunnangkul (Mahidol University, ailand)
Medical Illustrator: Chananya Hokierti (Nopparat Rajathanee Hospital, ailand)
Online Assistant: Surang Promsorn, Wilailuck Amornmontien, Hatairat Ruangsuwan Editorial Oce Secretary: Amornrat Sangkaew
International Editorial Board
Editorial Board
Watchara Kasinrerk (Chiang Mai University, ailand)
Rungroj Krittayaphong (Siriraj Hospital, Mahidol University, ailand)
Wiroon Laupattrakasem (Khon Kaen University, ailand)
Anuwat Pongkunakorn (Lampang Hospital, ailand)
Nopporn Sittisombut (Chiang Mai University, ailand)
Vasant Sumethkul (Ramathibodi Hospital, Mahidol University, ailand)
Yuen Tanniradorm (Chulalongkorn University, ailand)
Saranatra Waikakul (Siriraj Hospital, Mahidol University, ailand)
Pa-thai Yenchitsomanus (Siriraj Hospital, Mahidol University, ailand)
Surapol Issaragrisil (Siriraj Hospital, Mahidol University,ailand)
Jaturat Kanpittaya (Khon Kaen University, ailand)
Suneerat Kongsayreepong (Siriraj Hospital, Mahidol University, ailand)
Pornchai O-Charoenrat (Siriraj Hospital, Mahidol University, ailand)
Nopphol Pausawasdi (Siriraj Hospital, Mahidol University, ailand)
Supakorn Rojananin (Siriraj Hospital, Mahidol University, ailand)
Jarupim Soongswang (Siriraj Hospital, Mahidol University, ailand)
Suttipong Wacharasindhu (Chulalongkorn University, ailand)
Prapon Wilairat (Mahidol University, ailand)
Pornprom Muangman (Siriraj Hospital, Mahidol University, ailand)
Ampaiwan Chuansumrit
(Ramathibodi Hospital, Mahidol University, ailand)
Sayomporn Sirinavin
(Ramathibodi Hospital, Mahidol University, ailand)
Vitoon Chinswangwatanakul
(Siriraj Hospital, Mahidol University, ailand)
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
1
Original Article
SMJ
Lukana Preechasuk, M.D.*, Pimrapat Tengtrakulcharoen, MPH.Biostatictics**,Khemajira Karaketklang, MPH.
Biostatictics ***, Ram Rangsin, M.D.****, Tada Kunavisarut, M.D.*****
*Siriraj Diabetes Center of Excellence, **Department of Research and Development, ***Department of Medicine, Faculty of Medicine Siriraj Hospital,
Mahidol University, Bangkok, ****Department of Military and Community Medicine, Phramongkutklao College of Medicine, Bangkok, *****Department
of Medicine, Faculty of Medicine Siriraj Hospital, Bangkok, ailand.
Achievement of Metabolic Goals among Different
Health Insurance Schemes in Thai Patients with
Type 2 Diabetes Mellitus: a Nationwide Study
ABSTRACT
Objective: ailand has three healthcare insurance schemes, including: Universal Health Coverage (UHC), Social
Health Insurance (SHI), and Civil Servant Medical Benet (CSMB). UHC has the narrowest drug coverage, SHI
uses its own list, and CSMB oers the greatest drug coverage. e aim of this study was to investigate metabolic
goal achievement in patients with type 2 diabetes mellitus (T2DM) compared among the three healthcare schemes
in ailand.
Methods: Data were obtained from a nationwide survey administered by MedResNet during 2011 to 2012. A cross-
sectional survey was conducted in patients with T2DM aged >35 years who were treated for at least 12 months. e
data were retrospectively collected from medical records.
Results: Of 49,303 T2DM patients that were recruited, 69.8% were female. CSMB patients were the oldest and had
the longest diabetes duration. Achievement of BP, HbA1c, and LDL goals was 32.8%, 33.3%, and 42.5%, respectively.
UHC patients had the highest percentage of BP control achievement (<130/80 mmHg; 33.1%), while CSMB patients
had the highest percentage of HbA
1
c
(<7%; 40.4%) and LDL (<100 mg/dL; 49%) achievement. CSMB patients had
the highest prevalence of 2-goal (16.8% for T2DM without HT) and 3-goal achievement (8.2% for T2DM with HT).
Multivariate analysis revealed the CSMB scheme to be signicantly associated with 3-goal achievement compared
to the UHC scheme (odds ratio: 1.48, 95% condence interval: 1.30-1.67; p<0.01).
Conclusion: e prevalence of metabolic goal achievement in patients with T2DM in ailand is low. CSMB patients
have the highest prevalence of 2-goal and 3-goal achievement.
Keywords: Blood pressure; glycemic level; health insurance; lipid level; type 2 diabetes mellitus (Siriraj Med J 2020;
72: 1-9)
Corresponding author: Tada Kunavisarut
E-mail: kunavisarut@gmail.com
Received 4 June 2019 Revised 23 September 2019 Accepted 15 October 2019
ORCID ID: http://orcid.org/0000-0001-8711-5852
http://dx.doi.org/10.33192/Smj.2020.01
INTRODUCTION
Type 2 diabetes mellitus (T2DM) is a signicant
non-communicable disease. It can cause micro- and
macro-vascular complications that lead to disability and
mortality.
1
In addition to glycemic control, blood pressure
and lipid control are also very important for reducing
these vascular complications. Previous study demonstrated
that intensied multifactorial intervention aimed at
multiple risk factors reduced the risk of microvascular
complications, cardiovascular events, and cardiovascular
death.
2,3
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
2
Preechasuk et al.
Mortality and morbidity rates in patients with
T2DM have fallen considerably over recent decades.
4-6
ese improvements in diabetes care may be attributed
to improvements in several factors, including level of
patient (e.g., literacy, patient education, motivation), level
of healthcare professional (e.g., knowledge, awareness),
and level of healthcare system (e.g., broader list of drugs,
increased budgets). For the healthcare professional,
the ai Clinical Practice Guideline for Diabetes was
developed and it is revised approximately every 3 years.
is guideline provides basic knowledge about diabetes,
treatment goals, medication adjustment, and complication
management. It is distributed to primary and secondary
care physicians located in all regions of ailand. Regarding
the healthcare system, access to medical care and essential
medicine are very important for optimizing outcomes.
e cost of medicine consumed approximately 40% of
the total direct medical cost in a developed country
7
,
and approximately 25-65% of total spending on health
in developing countries.
8
ailand implemented universal health coverage
in 2002. ere are three main health insurance schemes,
including: 1) the Civil Servant Medical Benet (CSMB)
scheme for government employees and their dependents,
2) the Social Health Insurance (SHI) scheme for private
sector employees, and 3) the Universal Health Coverage
(UHC) scheme for the rest of ai population. In 2017,
CSMB, SHI, and UHC covered approximately 4.4, 10.6,
and 48.0 million ai people, respectively.
9
e expansion
of health insurance coverage increased access to medicines
in primary care, especially for chronic diseases, such as
diabetes and hypertension.
8
However, the drug lists,
budgets, and payment methods all dier among the
three health coverage schemes in ailand. UHC has
the narrowest drug coverage, as dictated by the ailand
National List of Essential Medicines (NLEM), with a
closed-end budget system; SHI uses its own drug list
with a closed-end budget system; and, CSMB oers the
greatest drug coverage with an open-end budget system.
9
UHC and CSMB beneciaries can receive medical service
without any copayment while SHI beneciaries have to
pay a monthly deductible around 25 US dollars. us,
the aim of this study was to investigate metabolic goal
achievement in patients with T2DM compared among
the three healthcare schemes in ailand.
MATERIALS AND METHODS
Study design
e data evaluated in this study were obtained from
a nationwide survey, entitled “An Assessment on Quality
of Care among Patients Diagnosed with Type 2 Diabetes
and Hypertension Visiting Ministry of Public Health and
Bangkok Metropolitan Administration Hospitals in ailand
(ailand DM/HT)”. is survey was administered by the
Medical Research Foundation ailand (MedResNet).
is survey was a cross-sectional study conducted in
patients with T2DM and/or hypertension who were
aged >35 years and treated for more than 12 months in
hospitals across ailand during 2011 to 2012. A two-
stage stratied cluster proportional to the size sampling
technique was used to select a representative sample of
patients with T2DM and/or hypertension in ailand.
e rst stage of sample collection was from the 77
provinces in ailand that comprised 77 strata. e
second stage of sample collection was from hospitals and
clinics, which were stratied into ve strata according
to their size, in each province. University hospitals were
excluded from this survey. At each hospital or clinic,
health care personnel, usually a registered nurse would
invite patients to participate into the study by signing
a consent form for providing permission to review and
abstract his or her medical records. e most recent data
within 12 months, including general information, the
status of diabetes complications, and laboratory results,
were retrospectively collected from medical records by
trained health care professionals. is national survey was
approved by the Institutional Ethical Review Committee
for Research in Human Subjects, ailand Ministry of
Public Health, and by the Royal ai Army Medical
Department Ethical Review Board. e authors of this
study received approval to use the MedResNet database.
e protocol for this study was approved by the Siriraj
Institutional Review Board [SIRB] (Si 376/2014).
Study population
is study included patients with T2DM who were
enrolled in the UHC, CSMB, and SHI public health
insurance schemes in ailand. Baseline characteristics
and metabolic goal achievement were compared among
health insurance schemes.
Goals of treatment
At the time of data collection, the targets of metabolic
control were BP <130/80 mmHg, HbA
1c
<7%, and LDL
<100 mg/dL, as recommended by the American Diabetes
Association (ADA) in 2011.
10
e BP and HbA
1c
targets
were the same targets as the current recommendations
by the 2018 ESC/ESH Guidelines for the Management
of Arterial Hypertension
11
and a consensus report by
the ADA and the European Association for the Study
of Diabetes.
12
However, the current recommendation
for treatment of hypercholesterolemia recommends
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
3
Original Article
SMJ
using moderate- or high-intensity statin in people with
diabetes.
13
ere were 2 goals set for patients in the T2DM
without HT group (HbA
1c
and LDL), and there were 3
goals set for patients in the T2DM with HT group (BP,
HbA
1c
, and LDL).
Statistical analysis
Statistical analyses were performed using SPSS Statistics
version 18.0 (SPSS, Inc., Chicago, IL, USA). Continuous
data are reported as mean ± standard deviation (SD) or
median and interquartile range (IQR), and categorical
data are reported as number and percentage. One-way
analysis of variance (ANOVA) or Kruskal-Wallis test
was used to analyze continuous data, and chi-square
test was used to evaluate categorical data. Univariate
logistic regression analysis was used to identify factors
signicantly associated with 3-goal achievement. Multiple
logistic regression analysis was employed to adjust factors
associated with 3-goal achievement. A p-value less than
0.05 was considered statistically signicant.
RESULTS
Baseline characteristics
Of the 49,303 T2DM patients recruited, 76.3%
had UHC coverage, 19.5% had CSMB coverage, and
4.2% had SHI coverage. Around 69% (n=33,887) of
patients had hypertension (DM with HT group). e
mean ± standard deviation age of patients was 59±10.7
years, and 70% of patients were female. ere were
signicant dierences in baseline characteristics among
the 3 healthcare schemes, as shown in Table 1. CSMB
patients were the oldest and had the longest duration of
diabetes, whereas SHI patients were the youngest and had
the shortest duration of diabetes. e majority of UHC
patients were followed-up at community hospitals. In
contrast, CSMB and SHI patients were followed-up in
relatively equal proportions at all 3 levels of hospitals.
UHC patients had the lowest mean BP (131.5±16/75±10
mmHg), and CSMB patients had the lowest mean HbA
1c
(7.7±1.8%) and LDL (106.0±37.0 mg/dL) levels.
Metabolic goal achievement
e overall achievement of BP, HbA
1c
, and LDL
goals was 32.8%, 33.3%, and 42.5%, respectively. Among
the evaluated schemes, UHC patients had the highest
percentage of overall BP control achievement (33.1%).
CSMB patients had the highest percentage of overall
glycemic (40.4%) and LDL control (49.0%) as shown
in Fig 1. e achievement of HbA
1c
and LDL goals was
not dierent between UHC and SHI patients. CSMB
patients had the highest prevalence of 2-goal and 3-goal
achievement as shown in Fig 2A & B. Specic to 1-goal
achievement, the percentage of LDL achievement was
greater than that of HbA
1c
and BP achievement. e
overall percentage of 1-goal achievement was not dierent
among the schemes. Regarding 2-goal achievement,
the percentage of HbA
1c
and LDL achievement was the
greatest compared to other combinations. Metabolic
goal achievement among DM with hypertension and
without hypertension is shown in Table 2.
Univariate logistic regression analysis showed
scheme, gender, age, duration of diabetes, and BMI to
be associated with 3-goal achievement. Multiple logistic
regression analysis revealed CSMB scheme (compared
with UHC scheme), older age, shorter duration of diabetes,
and lower BMI to be associated with 3-goal achievement
(Table 3).
DISCUSSION
All ai citizens have been covered by public universal
health coverage since 2002. All three ailand insurance
schemes provide essential laboratory investigations and
medications for blood pressure, glycemic, and lipid
control. However, in the present study, the prevalence of
each metabolic goal achievement in patients with T2DM
was approximately 30-40% and 3-goal achievement was
only 6.1%, which were close to the rates reported from
developing countries
14,15
but lower than the rates reported
from developed countries.
16,17
e 3-goal achievement
percentage of study from India
14
and Ethiopia
15
was 9%
and 8.5%, respectively. However, the BP goal of those
studies was < 140/90 mmHg which was higher than BP
goal of the present study. In the US during 2007-2010,
the percentage of BP, HbA
1c
, and LDL achievement was
51.3%, 52.2%, and 56.8%, respectively, while the rate of
3-goal achievement plus non-smoking status was 14.3%.
e prevalence of metabolic goal achievement in the US
increased markedly during 1999-2010.
16
In ailand, the
percentage of glycemic goal achievement increased from
26.3% in 2003 (data from ailand Diabetes Registry
(TDR) Project)
18
to 33.3% in the present study (2011-
2012). However, the data in the TDR project came from
university hospitals, which may treat cases of T2DM that
are more complicated than those included in our study.
In this study, the prevalence of LDL goal achievement
was the highest, which is similar to the ndings of a US
study
16
, followed by HbA
1c
and BP goal achievement.
However, there was no conclusion which risk factor had the
greatest inuence to morbidities and mortality in patients
with T2DM. Lizheng Shi, et al suggested LDL level to be
the strongest risk factor for diabetes complications and
mortality in veterans with T2DM.
19
Soa Gudbjörnsdottir,
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
4
Preechasuk et al.
TABLE 1. Baseline characteristics compared among dierent ailand health schemes.
Universal Health Civil Servant Social Health
Characteristics Total Coverage MedicalBenet Insurance P-value
(N=49,303) Scheme Scheme Scheme
(n=37,633) (n=9,641) (n=2,029)
Hospital size, %
Regional hospital (>500 beds) 15.6% 12.8% 23.0% 31.6%
General hospital (200-500 beds) 22.9% 19.5% 33.7% 35.4% <0.01
Community hospital (<120 beds) 61.5% 67.7% 43.4% 33.0%
Female gender, % 69.8% 72.9% 61.4% 51.5% <0.01
Age (years) 59±10.7 58.3±10.4 64.3±9.8* 49.4±8.8
†‡
<0.01
<40 years, % 3.2% 3.3% 0.7% 13.8% <0.01
40-49 years, % 14.9% 16.3% 5.4% 34.1%
50-59 years, % 34.4% 36.2% 26.3% 40.2%
60-69 years, % 29.7% 29.3% 35.4% 10.6%
≥70years,% 17.7% 14.9% 32.1% 1.4%
BMI (kg/m
2
), % 25.7±4.4 25.7±4.4 25.7±4.3 26.6±4.5
†‡
<0.01
Duration of diabetes (years), 6 (3, 9) 6 (3, 9) 6 (4, 10)* 4 (3, 7)
†‡
<0.01
median (IQR)
<5 years, % 37.4% 38.3% 31.1% 51.0% <0.01
5-9 years, % 41.4% 41.6% 41.6% 36.4%
10-14 years, % 15.2% 14.6% 18.6% 9.9%
≥15years,% 6.0% 5.5% 8.7% 2.8%
Blood pressure (mmHg)
Hypertension group (n=33,887)
Systolic 132.0±16 131.5±16.0 132.0±16.0* 133.0±16.0
†‡
<0.01
Diastolic 75.0±11 75.0±10.0 74.0±11.0* 79.0±11.0
†‡
<0.01
HbA
1c
, % 8.1±2.0 8.2±2.0 7.7±1.8* 8.0±1.9
‡
<0.01
LDL (mg/dL) 110±37 112±37 106±37* 111±37
‡
<0.01
Triglyceride (mg/dL), 157 (113, 221) 155 (111, 128) 138 (100,192)* 143 (103, 199)
†‡
<0.01
median (IQR)
GFR (ml/min/1.73 m
2
) 68±26 68±26 65±24* 84±24
†‡
<0.01
Hypoglycemia, % 4.4% 4.6% 3.9%* 2.9%
†‡
<0.01
Data presented as percentage (%) or mean ± standard deviation
* indicates p<0.05 between UHC and CSMB,
†
indicates p<0.05 between UHC and SHI,
‡
indicates p<0.05 between CSMB and SHI
Abbreviations: BMI = body mass index; IQR = interquartile range; HbA
1c
= glycated hemoglobin; LDL = low-density lipoprotein cholesterol;
GFR = glomerular ltration rate; UHC = Universal Health Coverage; CSMB = Civil Servant Medical Benet; SHI = Social Health Insurance
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
5
Original Article
SMJ
Fig 1. Metabolic goal achievement compared
among dierent health schemes.
*indicates p<0.05 between UHC and CSMB ,
**indicates p<0.05 between UHC and SHI ,
***indicates p<0.05 between CSMB and SHI
Abbreviations: CSMB = Civil Servant Medical
Benet; SHI = Social Health Insurance; UHC
= Universal Health Coverage
Fig 2A. Number of goal achievement of DM
without hypertension group compared among
dierent health schemes.
*indicates p<0.05 between UHC and CSMB ,
***indicates p<0.05 between CSMB and SHI
Abbreviations: CSMB = Civil Servant Medical
Benet; SHI = Social Health Insurance; UHC
= Universal Health Coverage
Fig 2B. Number of goal achievement of DM
with hypertension group compared among
dierent health schemes.
*indicates p<0.05 between UHC and CSMB ,
***indicates p<0.05 between CSMB and SHI
Abbreviations: CSMB = Civil Servant Medical
Benet; SHI = Social Health Insurance; UHC
= Universal Health Coverage
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
6
Preechasuk et al.
TABLE 2. Metabolic goal achievement among DM with hypertension and without hypertension.
Universal Health Civil Servant Social Health
Metabolicgoals Total Coverage MedicalBenet Insurance P-value
(n=49,303) Scheme Scheme Scheme
(n=37,633) (n=9,641) (n=2,029)
DM without hypertension (n=15,416) (n=12,383) (n=2,319) (n=714)
1 goal achieved, %
HbA
1c
<7% 15.1% 14.6% 17.7%* 16.5% 0.03
LDL <100 mg/dL 27.6% 27.7% 27.5% 26.4% 0.84
2 goals achieved, %
HbA
1c
and LDL 12.3% 11.3% 16.8%* 12.2%
‡
<0.01
DM with hypertension (n=33,887) (n=25,250) (n=7,322) (n=1,315)
1 goal achieved, %
HbA
1c
<7% 12.3% 12.1% 13.1% 12.6% 0.12
LDL <100 mg/dL 17.4% 17.0% 18.5% 17.3% 0.06
BP <130/80 mmHg 11.9% 12.9% 8.9%* 11.7%
‡
<0.01
2 goals achieved, %
HbA
1c
and LDL 11.5% 10.3% 15.0%* 12.7%
†‡
<0.01
LDL and BP 9.7% 9.5% 10.3% 9.1% 0.19
HbA
1c
and BP 5.8% 5.9% 6.0% 4.1% 0.052
3 goals achieved, %
HbA
1c
LDL and BP 6.1% 5.5% 8.2%* 5.1%
‡
<0.01
*indicates p<0.05 between UHC and CSMB,
†
indicates p<0.05 between UHC and SHI,
‡
indicates p<0.05 between CSMB and SHI
Abbreviations: BP = blood pressure; HT = hypertension; HbA
1c
= glycated hemoglobin; LDL = low-density lipoprotein cholesterol; DM =
diabetes mellitus; GFR = glomerular ltration rate; UHC = Universal Health Coverage; CSMB = Civil Servant Medical Benet; SHI = Social
Health Insurance
et al reported HbA
1c
level to be the strongest predictor of
stroke and acute myocardial infarction.
20
However, both
of those studies conrmed that the higher the number
of goals achieved to reduce cardiovascular risk factors,
the lower the rate of morbidities and mortality.
19,20
is
supports the importance of metabolic goal achievement
in this patient population.
Aer adjusting for gender, age, duration of diabetes,
and BMI, the CSMB scheme was still found to be associated
with 3-goal achievement. is dierence in metabolic
control among schemes might be explained by one or
more of the following factors: 1) dierences in the lists
of available medicine, 2) opportunity or not to consult
a specialist, and/or 3) socioeconomic and/or geographic
barrier.
First, UHC patients have to use the medicines listed
in the ailand NLEM (supplementary data), and most of
these medicines are lower-cost alternatives. e drug list
for SHI patients diers among hospitals, but it should be
at least comparable to the NLEM. In contrast, the use of
non-essential medicines is permitted for CSMB patients
if physicians conrm them to be clinically indicated.
21
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
7
Original Article
SMJ
TABLE 3. Multiple logistic regression analysis of 3-goal achievement.
Factors Odds ratio 95% CI P-value
UHC scheme 1.00
CSMB scheme 1.48 1.30-1.69 <0.01
SHI scheme 0.99 0.71-1.41 0.99
Female 1.00
Male 1.097 0.97-1.24 0.14
Age <40 years 1.00
Age 40-49 years 2.55 1.11-5.88 0.03
Age 50-59 years 2.87 1.27-6.50 0.01
Age 60-69 years 3.34 3.34-1.48 <0.01
Age≥70years 4.29 4.29-1.89 <0.01
DM duration <5 years 1.00
DM duration 5-9.9 years 0.95 0.83-1.08 0.44
DM duration 10-14.9 years 0.82 0.68-0.97 0.02
DMduration≥15years 0.64 0.49-0.83 <0.01
BMI≥30kg/m
2
1.00
BMI 25-29.9 kg/m
2
1.21 0.99-1.48 0.55
BMI 23-24.9 kg/m
2
1.43 1.15-1.77 <0.01
BMI <23 kg/m
2
1.84 1.51-2.25 <0.01
A p-value<0.05 indicates statistical signicance
Abbreviations: CI = condence interval; UC = Universal Health Coverage; CSMB = Civil Servant Medical Benet; SHI = Social Health
Insurance; DM = diabetes mellitus; BMI = body mass index
For BP achievement, although only generic drugs are
available in the NLEM, all antihypertensive drug classes
are included. Contrarily, only moderate-intensity statin
(NLEM 2008-2012) and low-cost anti-hyperglycemic
agents are available in the NLEM for lipid and glycemic
control. is might causes dierences in lipid and glycemic
control among schemes, but similar for BP control between
UHC and CSMB. Second, UHC and SHI use a primary
healthcare gatekeeping system. is means that patients
can only access hospitals where they are registered, while
CSMB allows patients to access specialists directly.
21
It
should be mentioned that care given by specialists does
not always ensure better metabolic control than care given
by general practitioners.
22,23
ird, most CSMB patients
live in urban settings and have higher socioeconomic
status, whereas most UHC patients live in rural areas
and have lower socioeconomic status.
21
erefore, CSMB
patients might have higher health awareness and health
literacy. Although health centers and district hospitals are
distributed across ailand, geographic barriers remain
a problem in some rural areas. However, data about the
specialty of doctor, residency and socioeconomic status
among health scheme were not collected in this study.
To reduce the disparities among health insurance
schemes, the inclusion of some new anti-hyperglycemic
agents and high potency statin with lowest acquisition
cost to the NELM might improve glycemic and lipid
control in UHC and SHI patients. ere should also be
strict indications and verication for the use of high-
cost medicines to control the budget in all 3 schemes.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
8
Preechasuk et al.
Otherwise, the expansion of the drug list will cause rapid
cost escalation, as observed in the CSMB scheme. In
2014, expenditure per CSMB member was 4 times higher
than that per UHC member.
21
Furthermore, appropriate
referral to specialists might optimize metabolic goal
achievement.
e strength of this study is a large number study
population that represents all 5 regions of ailand and
dierent sized hospitals except university hospitals.
Moreover, data were abstracted from medical records by
trained healthcare professionals. us, the data used in
this study should be considered more accurate than self-
reported data. Although this study represents the context
in 2011-2012, the trend of metabolic goal achievement
among the 3 ailand healthcare schemes may currently
be the same. is hypothesis is based on the fact that the
drug classes for BP, cholesterol and, glycemic control for
type 2 diabetes in the NELM 2018 are similar to those
listed in the NELM 2008-2012, except for the addition of a
high-intensity statin for hypercholesterolemia treatment.
ere are some limitations in this study. First, there
might be some selection bias because of a nature of cross
sectional study. erefore, it can only demonstrate the
association not cause and eect. Second, there were
no data about other factors that might inuence the
metabolic goal achievements such as educational level,
health literacy, income or residency. erefore, we did
not adjust those factors in the multiple logistic regression
analysis. Finally, there were unequal sample sizes in each
scheme since majority of ai population use UHC
scheme.
CONCLUSION
e prevalence of blood pressure, glycemic, and lipid
control achievement in patients with T2DM in ailand
is low. CSMB patients have the highest prevalence of
HbA
1c
, LDL, 2-goal, and 3-goal achievement. Exploring
the factors that contribute to dierences among healthcare
schemes may improve diabetes care in ailand.
ACKNOWLEDGMENTS
is research utilizes data provided by the study:
“An Assessment on Quality of Care among Patients
Diagnosed with Type 2 Diabetes and Hypertension Visiting
Ministry of Public Health and Bangkok Metropolitan
Administration Hospitals in ailand (ailand DM/
HT)”, a collaborative clinical study supported by the
ailand National Health Security Oce (NHSO) and
the ailand Medical Research Network (MedResNet).
e data was archived at the web site http://www.damus.
in.th maintained by MedResNet.
is manuscript was not prepared in collaboration
with Investigators of the ailand DM/HT study and
does not necessarily reect the opinions or views of the
ailand DM/HT study, the ailand NHSO or the
ailand MedResNet.
Conict of interest declaration: All authors declare
no personal or professional conicts of interest, and no
nancial support from the companies that produce and/
or distribute the drugs, devices, or materials described
in this report.
Funding disclosure: is was an unfunded study.
REFERENCES
1. International Diabetes Federation. IDF Diabetes Atlas, 8
th
edn. Brussels, Belgium: International Diabetes Federation;
2017.
2. Gæde P, Vedel P, Larsen N, Jensen GV, Parving H-H, Pedersen
O. Multifactorial intervention and cardiovascular disease in
patients with type 2 diabetes. N Engl J Med 2003;348:383-93.
3. Gæde P, Lund-Andersen H, Parving H-H, Pedersen O. Eect
of a multifactorial intervention on mortality in type 2 diabetes.
N Engl J Med 2008;358:580-91.
4. Rawshani A, Rawshani A, Franzen S, Eliasson B, Svensson AM,
Miaraj M, et al. Mortality and Cardiovascular Disease in
Type 1 and Type 2 Diabetes. N Engl J Med. 2017;376:1407-18.
5. Cheng YJ, Imperatore G, Geiss LS, Saydah SH, Albright AL,
Ali MK, et al. Trends and Disparities in Cardiovascular Mortality
Among U.S. Adults With and Without Self-Reported Diabetes,
1988-2015. Diabetes Care 2018;41:2306-15.
6. Gregg EW, Li Y, Wang J, Burrows NR, Ali MK, Rolka D, et al.
Changes in diabetes-related complications in the United States,
1990-2010. N Engl J Med 2014;370:1514-23.
7. American Diabetes A. Economic Costs of Diabetes in the U.S.
in 2017. Diabetes Care 2018;41:917-28.
8. Quick JD. Ensuring access to essential medicines in the
developing countries: a framework for action. Clinical Pharmacology
& erapeutics. 2003;73(4):279-83.
9. Tangcharoensathien V, Witthayapipopsakul W, Panichkriangkrai
W, Patcharanarumol W, Mills A. Health systems development in
ailand: a solid platform for successful implementation of
universal health coverage. e Lancet. 2018.
10. Prevention I. Standards of medical care in diabetes—2011.
Diabetes care. 2011;34:S11.
11. Williams B, Mancia G, Spiering W, Agabiti Rosei E, Azizi M,
Burnier M, et al. 2018 ESC/ESH Guidelines for the management
of arterial hypertension. Eur Heart J. 2018;39(33):3021-104.
12. Davies MJ, D’Alessio DA, Fradkin J, Kernan WN, Mathieu C,
Mingrone G, et al. Management of hyperglycaemia in type
2 diabetes, 2018. A consensus report by the American Diabetes
Association (ADA) and the European Association for the
Study of Diabetes (EASD). Diabetologia. 2018;61(12):2461-98.
13. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK,
Blumenthal RS, et al. 2018 AHA/ACC/AACVPR/AAPA/
ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline
on the Management of Blood Cholesterol: A Report of the
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
9
Original Article
SMJ
American College of Cardiology/American Heart Association
Task Force on Clinical Practice Guidelines. Journal of the
American College of Cardiology. 2018:25709.
14. Kumar KH, Modi K. A1c, blood pressure and cholesterol goal
achievement in patients of Type 2 diabetes. Medical Journal
of Dr DY Patil University. 2016;9(2):195.
15. Belay E, Abera A, Mehari A, Gebremeskel G, Endrias A, Endris
K. Achievements of Diabetes Goals and eir Determinants
in Type 2 Diabetic Patients Attending Outpatient Diabetic
Clinic in Northern Ethiopia. International journal of chronic
diseases. 2017;2017.
16. Ali MK, Bullard KM, Saaddine JB, Cowie CC, Imperatore G,
Gregg EW. Achievement of goals in U.S. diabetes care, 1999-
2010. N Engl J Med. 2013;368(17):1613-24.
17. Yu SH, Kang JG, Hwang YC, Ahn KJ, Yoo HJ, Ahn HY, et al.
Increasing achievement of the target goals for glycemic, blood
pressure and lipid control for adults with diagnosed diabetes
in Korea. J Diabetes Investig. 2013;4(5):460-5.
18. Kosachunhanun N, Benjasuratwong Y, Mongkolsomlit S,
Rawdaree P, Plengvidhya N, Leelawatana R, et al. ailand
diabetes registry project: glycemic control in ai type 2 diabetes
and its relation to hypoglycemic agent usage. Journal of the
Medical Association of ailand = Chotmaihet thangphaet.
2006;89 Suppl 1:S66-71.
19. Shi Q, Liu S, Krousel-Wood M, Shao H, Fonseca V, Shi L.
Long-term outcomes associated with triple-goal achievement
in patients with type 2 diabetes mellitus (T2DM). Diabetes
research and clinical practice. 2018;140:45-54.
20. Rawshani A, Rawshani A, Franzen S, Sattar N, Eliasson B, Svensson
AM, et al. Risk Factors, Mortality, and Cardiovascular Outcomes
in Patients with Type 2 Diabetes. N Engl J Med. 2018;379(7):633-44.
21. Patcharanarumol W, Panichkriangkrai W, Sommanuttaweechai
A, Hanson K, Wanwong Y, Tangcharoensathien V. Strategic
purchasing and health system eciency: A comparison of two
nancing schemes in ailand. PLoS One. 2018;13(4):e0195179.
22. De Berardis G, Pellegrini F, Franciosi M, Belglio M, Di Nardo
B, Greeneld S, et al. Quality of care and outcomes in type 2
diabetic patients: a comparison between general practice and
diabetes clinics. Diabetes Care. 2004;27(2):398-406.
23. Arai K, Hirao K, Matsuba I, Takai M, Matoba K, Takeda H,
et al. e status of glycemic control by general practitioners
and specialists for diabetes in Japan: a cross-sectional survey of
15,652 patients with diabetes mellitus. Diabetes research and
clinical practice. 2009;83(3):397-401.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
10
Martini, M.D.*, Irwanto, M.D., Ph.D.*,**, Roedi Irawan, M.D., Ph.D.*, Nur Aisiyah Widjaja, M.D.*
*Department of Child Health, Faculty of Medicine, Universitas Airlangga. Jl Mayjen Prof Moestopo 47, Surabaya 60132 East Java-Indonesia,
** Universitas Airlangga Hospital, Kampus C Mulyorejo, Surabaya 60115 East Java-Indonesia.
BreastmilkMacronutrientLevelsandInfantGrowth
During the First Three Months: a Cohort Study
ABSTRACT
Objective: e rst three months aer birth is a critical time interval for growth and development. Breastmilk is a
natural nutrition source for infants. However, studies on the practice of exclusive breastfeeding and infant growth
tend to result in contradictions. e objective of this study was to investigate the correlation between breastmilk
macronutrient levels and infant growth during the rst three months.
Methods: We conducted an observational cohort study at Universitas Airlangga Hospital from June-October 2018.
Subjects were enrolled using total sampling. Infant anthropometry, as dened by body weight, body length, and head
circumference, were measured. Breastmilk specimens were collected using a breast pump and then sent directly
for analysis. Lactose, protein, fat, and total calorie levels were obtained using a human milk analyzer. Procedures
were repeated three times, once per month. e Pearson correlation coecient was used for statistical analysis.
Results: Forty participants were enrolled in this study. ere was a positive correlation between breastmilk total
calories and head circumference growth during the rst (p = 0.039), second (p = 0.020), and third month (p =
0.020). Breastmilk protein level was positively correlated with body length (p < 0.05) and head circumference
(p < 0.05) during the rst month. ere was no correlation between body weight and breastmilk macronutrients
or total calories (p > 0.05).
Conclusion: Breastmilk macronutrient levels correlate to infant growth in a unique pattern. Total calories and rst
month protein correlated positively with infant head circumference. However, calorie source, e.g., lactose or fat,
did not correlate with infant body weight and length.
Keywords: Breastmilk; nutrition; growth; infant; macronutrient (Siriraj Med J 2020; 72: 10-17)
Corresponding author: Irwanto
E-mail: irwanto.idris@gmail.com
Received 23 April 2019 Revised 24 July 2019 Accepted 9 September 2019
ORCID ID: http://orcid.org/0000-0002-7573-8793
http://dx.doi.org/10.33192/Smj.2020.02
Martini et al.
INTRODUCTION
Growth and development are the main features
that distinguish between children and adults. e rst
three months aer birth is a critical time interval for
growth and development. e brain volume increases
quickest during this time, at roughly 1% per day aer
birth; the rate slows to 0.4% per day by the end of third
month.
1
Child growth itself is commonly associated with
nutrition intake. Abnormal nutrition intake will result
in aberrant growth. In return, deviation from normal
growth may indicate serious health problems.
2
Hence,
nutrition is an important aspect to be observed during
childhood.
Breastmilk is a natural nutrition source for infants.
Human breastmilk contains specic components that
avert malnutrition and oer optimal growth for infants.
3,4
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
11
Original Article
SMJ
Lactose, a main breastmilk component, is positively
correlated to infant weight, adiposity, and body mass
index in the 3-month-old infant.
5
Although various proteins in breastmilk are lower
than in formulated milk,
6
breastmilk contains specic
free amino acids (such as glutamine) that are positively
correlated with length gain.
7,8
Another study showed
that a higher n-3/n-6 ratio in breastmilk positively
correlates with infant length.
9
However, no matter how
good breastmilk may appear to be for the infant, studies
on the practice of exclusive breastfeeding and infant
growth tend to result in contradictions. According to
an Indonesian Health Department report, although
exclusive breastfeeding increased from 15.3% to 30.2%
between 2010-2013, stunting and malnutrition remain
serious problems. One study showed that exclusive
6-month breastfeeding leads to underweight infants.
10
However, other studies instead reported overweightness
in exclusively breastfed infants.
9,11
ese dierences do
not apparently correlate with the breastfeeding itself, but
rather are more correlated with breastmilk macronutrient
content.
11,12
Hence, in this study we aimed to investigate
the correlation between breastmilk macronutrients and
infant growth during the rst three months.
MATERIALS AND METHODS
Study design
is study was an observational cohort deign that
aimed to explain the correlation between breastmilk
macronutrients, dened by carbohydrates (in the form of
lactose), protein, and fat, and infant anthropometry, dened
by body weight, body length, and head circumference.
is study was conducted at the Universitas Airlangga
Hospital from June to October 2018. Ethical clearance
was issued by the Ethics Committee at the Universitas
Airlangga Hospital, Surabaya, Indonesia, with clearance
number 108/KEH/2018.
Study subjects
Subjects were enrolled using total sampling methods,
where all candidates who met inclusion criteria were
recruited during the study period. Inclusion criteria
were term neonates with a birthweight > 2,500 g. e
infant’s mother must exclusively breastfeed without any
additional nutritional intake for at least three months.
Exclusive breastfeeding was not defined by a daily
measurable dose, but rather as natural as possible where
the infant was satiated. e mothers also had to have
their Mother and Child Health Handbook still available
for data tracking and cross-checking. e Mother and
Child Health Handbook is a book ocially given by the
Indonesian government to each pregnant mother in order
to help keep track of her and her baby’s health status.
Informed consent signed by the mother was required to
participate in this study. Preterm infants, infants with
congenital disease, and mothers with HIV infection
who were unable to breastfeed were excluded from this
study. Participants were considered dropped-out when
the mother did not exclusively breastfeed, the Mother
and Child Health Handbook was missing, the mother
or infant did not survive during our observation, or she
decided to withdraw from the study.
Study questionnaire
A questionnaire was provided to obtain characteristics
about the mothers and infants, as well as a tool for
monitoring inclusion criteria. We sought data about
the mother’s age, parity, education, and employment,
while for the child we conrmed sex and method of
birth. e monitoring questions were assessed each
month and consisted of infant general condition and
breastfeeding status. To obtain body mass index data
for mothers during pregnancy, data from the Mother
and Child Health Handbook were used. Questionnaires
were completed by the investigator aer participants
provided written informed consent.
Anthropometrical measurements
We measured the infant’s body weight, length, and
head circumference as anthropometric indicators. All
measurements used metric scales and were performed at
the Universitas Airlangga Hospital directly by investigator.
e measurement was repeated once per month for
three months. Measurement of body weight used an
analogue infant scale with an accuracy of ± 0.1 kg. e
infant scale was placed on a at table that was not easy to
sway. Before body weight was measured, the needle was
set to zero. Babies were weighed naked without clothing
and accessories. Body weight was recorded as the point
where needle was most stable. Body length measurement
was performed using an infantometer. e infants were
rst positioned lying on a at bed. e infant’s head was
attached to the numerical barrier while the examiner’s
le hand pressed the knee straight and the right hand
pressed the toe to the foot. Length was measured based
on the pointed number on the outer edge of the gauge.
Head circumference was measured with metered tape.
e tape was looped over the forehead, between eyebrows,
above the two ears, and passing the occiput. e metered
tape was then pulled slightly tighter, and the value that
crossed the zero point was read.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
12
Breastmilk specimen collection
Breastmilk specimens were collected individually by
each mother using a breast pump at random times. Full
breastmilk expression, including foremilk and hindmilk,
was used as a sample. To collect the sample, pumped
breastmilk was rst collected in a 50 mL sterile bottle. e
minimum analyzable amount for each specimen was 3 mL,
which was collected using sterile plastic tube. Collected
breastmilk was then sent directly to the laboratory for
analysis. We did not store any samples because they were
all directly analyzed aer collection. is procedure was
repeated once a month for three months.
Breastmilk macronutrient analysis
Laboratory analysis was conducted at the Universitas
Airlangga Hospital. Macronutrient analysis was performed
using the Miris Human Milk Analysis instrument set
(Miris, Sweden). Samples were first reheated to the
optimal temperature and then homogenized with a
20 J/mL ultrasonic wave. Homogenized samples were
analyzed under mid-infrared (mid-IR) transmission
spectroscopy. Results are presented for lactose (g/100
mL), protein (g/100 mL), soluble fat (g/100 mL), and
total calories (kcal/100 mL). e Pearson correlation
coefficient was used to analyze the data using SPSS
version 22.0. Results are presented in tables.
RESULTS
Subject characteristics
We enrolled 40 subjects for this study. A total of
10 participants were excluded, all due to non-exclusive
breastfeeding. e remaining subjects nished all the
observations and were included in the study. e mothers
average age was 29.5 years old. Most of them were on
their second parity (43%). e majority of the mothers
were high school graduates (60%), followed by middle
school graduates (17%). Most of the mothers (60%)
were unemployed and taking a role as a housewife. With
regards to the children, the number of male and female
subjects was approximately the same (binomial non-
parametric test, p > 0.05). Caesarean section occurrence
was relatively high (47%) and served as the only assisted
method of delivery in our study. Detailed information
about subject characteristics is provided in Table 1, while
the anthropometry summary is presented in Table 2.
Breastmilk macronutrient prole
We measured the breastmilk lactose, protein, fat,
and total calories prole. We found that lactose was the
most abundant breastmilk macronutrient (compared to
protein and fat). e highest lactose concentration was
detected in the third month of observations, with 4.17
± 0.81 g/100 mL breastmilk. In our population, protein
was the smallest component found in breastmilk. Its
concentration peaked at the rst month of observation
at 1.89 ± 0.38 g/100 mL. We observed a slow but gradual
decrease in the protein concentration each month. Fat was
the second most abundant macronutrient; its concentration
was the highest in the second month at 3.56 ± 0.96 g/100
mL, but dropped rapidly to 2.80 ± 1.16 g/100 mL (unlike
the other macronutrients that did not exhibit such sharp
temporal changes). Total breastmilk calories changed
each month; it generally followed the same pattern as
protein. e peak was 56.73 ± 9.77 kcal/100 mL in the
rst month, followed by a gradual decrease over the
time. e complete breastmilk macronutrient prole
is provided in Table 3.
Correlation between breastmilk macronutrients and
infant anthropometry
Prior to performing the correlation tests, the
Kolmogorov-Smirnov test was used to test the data
distribution for all variables. All anthropometric
measurements as well as breastmilk carbohydrates,
protein, fat, and total calories were normally distributed
(p > 0.05). Hence, the Pearson correlation coecient was
calculated for all variables. A detailed summary of the
p-values, Pearson r correlation strength value, and the
respective time of measurement is presented in Table 4.
We did not nd any correlation between body weight
and breastmilk macronutrients (carbohydrates, protein,
or fat p > 0.05). We also did not nd any correlation
between breastmilk total calories and infant body weight
(p > 0.05). ese results were consistently found for all
monthly observations.
Protein was related to infant growth, especially
during the rst month; protein was positively correlated
with body length (p < 0.05) and head circumference (p <
0.05) for rst month. However, there were no correlations
between these measures and protein levels during the
second and third months (p > 0.05).
ere was also a positive correlation between breast
milk total calories and infant head circumference for all
three months of observations; (p < 0.05; Table 4).
DISCUSSION
During their rst three months of life, infants grow
faster than at older ages due to the larger body volume-
to-surface ratio. Hence, young infants have an average
higher average metabolic rate per kg body weight and
require greater nutrient intake for body temperature
regulation at this time.
12
is pattern of metabolism
Martini et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
13
Original Article
SMJ
TABLE 1. Participant characteristics.
Characteristic Total (n = 30)
Mothers
Age (years; mean ± SD) 29.5 ± 7.19
Parity
First birth 7 (23%)
Second birth 13 (43%)
Third birth 5 (17%)
Fourth birth 3 (10%)
Fifth birth 2 (7%)
Highest Education
Elementary school 3 (10%)
Middle school 5 (17%)
High school 18 (60%)
Vocational 3 (10%)
Bachelor’s 1 (3%)
Mother’s employment
Housewife 18 (60%)
Employee 6 (20%)
Self-employed 6 (20%)
Infants
Sex
Male 17 (57%)
Female 13 (43%)
Mode of delivery
Normal 16 (53%)
Caesarean section 14 (47%)
Birth weight (g; mean ± SD) 3123 ± 47.5
Birth length (cm; mean ± SD) 48.7 ± 0.50
TABLE 2. Infant anthropometric characteristics.
Components
Mean ± SD
1
st
month 2
nd
month 3
rd
month
Weight (kg) 3.46 ± 0.54 (normal
1
) 4.51 ± 0.83 (normal
1
) 5.60 ± 0.59 (normal
1
)
Length (cm) 51.95 ± 2.40 (normal
1
) 55.40 ± 2.11 (normal
1
) 59.60 ± 2.14 (normal
1
)
Head circumference (cm) 34.73 ± 1.16 (normal
1
) 36.90 ± 1.11 (normal
1
) 38.84 ± 1.10 (normal
1
)
1
Normal range is based on World Health Organization Z-score chart of body weight-for-age, length-for-age, and head circumference-for-
age for children aged 0-2 years.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
14
TABLE 3. Breastmilk macronutrients and total calories prole.
TABLE 4. BCorrelations between breastmilk macronutrients and infant growth.
Components
Mean ± SD
1
st
month 2
nd
month 3
rd
month
Lactose (g/100 mL) 3.77 ± 0.72 3.72 ± 0.79 4.17 ± 0.81
Protein (g/100 mL) 1.89 ± 0.38 1.83 ± 0.43 1.72 ± 0.37
Fat (g/ 100 mL) 3.41 ± 0.97 3.56 ± 0.96 2.80 ± 1.16
Total calories (kcal/100 mL) 56.73 ± 9.77 55.50 ± 9.62 53.37 ± 10.72
Lactose Protein Fat Total calories
r p r p r p r p
1
st
month 0.051 0.79 -0.264 0.158 0.141 0.457 0.276 0.139
Weight 2
nd
month 0.247 0.189 -0.037 0.847 0.152 0.423 0.269 0.151
3
rd
month 0.488 0.189 -0.002 0.847 -0.023 0.423 -0.006 0.151
1
st
month 0.291 0.118 0.501 0.005* 0.024 0.898 0.109 0.566
Length 2
nd
month 0.193 0.308 -0.044 0.817 0.101 0.597 0.283 0.129
3
rd
month 0.577 0.308 -0.226 0.817 -0.177 0.597 -0.181 0.129
Head
1
st
month 0.222 0.238 0.468 0.009* 0.047 0.804 0.379 0.039*
Circumference
2
nd
month 0.278 0.136 -0.08 0.673 0.276 0.14 0.423 0.020*
3
rd
month 0.445 0.136 -0.015 0.673 0.189 0.14 0.264 0.020*
Notes: r, Pearson’s correlation coecient; *statistically signicant correlation (p < 0.05).
requires higher caloric input, and this need is met by
macronutrients. Breastmilk is a natural source of such
nutrition, and its content is critical for infant nutrition.
Weight and length are important anthropometric
values in growing infants. We observed normal weight
and length increases of exclusively breastfed infants
during the study time interval. Our results dier from
previous studies. One study showed 3.1% of infants were
underweight in a cohort of exclusively breastfed infants at
their third month of age
10
, while another reported excessive
overweightness in an exclusively breastfed cohort.
9,11
Our
disparate results might be due to dierences in subject
characteristics, research methodology, and/or breastmilk
quality. Hence, assessing breastmilk macronutrient levels
is crucially important.
Head circumference has a relationship with brain
volume, and it is oen used to assess brain growth aer
a baby is born.
13,14
Evidence shows that total energy in
breastmilk supports the energy requirements for brain
growth,
15
which we will discuss later. However, the
growth trajectory of the exclusively-breastfed infant
brain is still poorly studied. Further studies are needed
to assess how a specic breastmilk component could
aect infant brain growth.
Martini et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
15
Original Article
SMJ
Compared to several previous studies in developed
and developing countries, breastmilk in this study had
lower lactose levels. Research in South Korea showed that
breastmilk lactose levels in the third month averaged 7.1
± 0.5 g/100 ml.
16
Similarly, research in the Philippines
showed a higher mean lactose level, 7.3 ± 0.58 g/100 ml.
17
Lactose levels in breast milk are known to be higher in
mothers who have more breast milk.
7
A study showed
that formula feeding can increase the quantity of breast
milk a breastfeeding mother produces.
18
is observation
further reinforces the importance of maternal formula
feeding during pregnancy.
In this study, the rst month breastmilk fat content
was lower compared to other reports. Of the analyzed
macronutrients, our measured third-month fat level was
the lowest when compared with other studies.
17,19
Fat
content is the most variable macronutrient in breastmilk.
20
An Indonesian study of breastmilk fat levels reported
an average of 5.94 ± 15.9 g/100 ml,
21
which is similar
with studies conducted in South Korea (3.2 ± 1.0 g/100
ml)
16
and China (3.11 ± 1.13) g/100 ml).
22
ese fat level
dierences are linked to variations in the maternal diet,
age, type of analyzed samples (hindmilk or foremilk),
breastfeeding time, and breastmilk storage conditions
post-pumping.
16,22-25
Unfortunately, daily diet data and
cohort data of maternal body mass index were not available
in our study, and thus we could not conrm whether fat
content variability in our samples was due to that factor.
In our study, breastmilk fat level was not directly
correlated toward any child growth anthropometric
measures. However, total calories were positively correlated
to head circumference growth. Hence, we emphasize
that both lactose and fat are important calorie sources
used for brain development. ere is limited evidence
on how total caloric intake aects brain volume, which
can be represented as head circumference. However,
in the 1970s, a study showed that caloric intake is very
important to achieve normal head circumference. e
study showed that exposure to subnormal temperature
(< 35°C) and failure to achieve a caloric intake of 120
cal/kg/day leads to abnormal head circumference growth
in asymptomatic neonates. Recent findings showed
that head circumference does not solely rely on total
caloric intake. Rather, an adequate and balanced diet
for optimal child brain development is recommended
rather than pursuing a specic total caloric intake. e
study showed that cumulative fat and caloric intake
are positively associated with fractional anisotropy in
the posterior limb of the internal capsule, a measure
that relates to an increase in brain volume. e same
study also showed that cumulative protein intake is
positively associated with higher cognitive and motor
scores.
26
erefore, from these points of view and based
on our ndings, we conclude that breastmilk contains
the necessary calories for adequate infant brain growth
as well as the required macronutrients level needed to
achieve daily caloric needs.
We observed a decrease in breastmilk protein levels
as the infants aged. Protein levels contained in mature
breastmilk are between 0.9-1.1 g/100 ml.
27
However,
breastmilk protein levels are higher at the beginning of
the lactation: they start at 1.4-1.6 g/100 ml, decrease to
0.8-1.0 g/100 ml when infants reaches 3-4 months old,
and further decline to 0.7-0.8 g/100 ml when the baby
reaches 6 months.
28
Interestingly, the breastmilk in this
study had higher protein levels even during the third
month of observation (1.72 ± 0.37) g/100 ml compared
to that previous report. Again, maternal diet, age, type of
analyzed samples (hindmilk or foremilk), breastfeeding
time, and breastmilk storage conditions post-pumping
might play role in our nding, since other macronutrients
(especially fat) depend on those factors.
16,22-25
Our ndings on the correlation between breastmilk
protein level and body weight suggest that protein is an
important energy source in earlier months, but it is less
crucial compared to fat and carbohydrates at older ages.
ere are several studies that support this view. One
showed that weight gain is lower in infants who received
breastmilk with less protein compared to higher protein.
However,anthropometric measurements are higher in
the group that received formula milk with higher protein,
a measure that is associated with a higher-than-normal
body mass index.
29
Our nding is also in line within the
context of a study that stated protein intake in formula-fed
infants exceeds the required amount aer 1-2 months of
age. Further, this excess is probably responsible for the
higher adiposity in older infants. e same study also
showed that the breastfed infants are leaner compared to
other groups.
30
However, it must be noted that most of
these previous studies compared protein levels of human
breastmilk to formula milk. erefore, careful deduction
is still needed when concluding whether a dierence in
human breastmilk protein level alone might have the
same eect.
Our study has some limitations that hinder us from
drawing a more generalized conclusion. First, there is still
no Indonesian standard regarding breastmilk macronutrient
levels, a deciency that makes our study prone to bias
when compared to other studies due to dierences in
population characteristics. Second, although we used a
total-sampling method to enroll participants, we were
only able to register thirty participants (approximately
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
16
one new participant per day). Further studies with longer
sampling periods may be able to overcome this limitation.
ird, our measurement method yielded no information
on total breastmilk quality over 24 h, a decit that made
it dicult to calculate daily total calories available from
the breastmilk. Fourth, perhaps the main drawback in our
study was the unavailability of maternal diet data, which
may aect breastmilk production and its quality during
pregnancy and aer childbirth. Finally, breastfeeding
volume and frequency was dened “as the infant is
satised with the breastfeeding” rather than a controlled
bottled volume. erefore, further studies with dierent
designs are required to investigate about breastmilk and
infant growth and overcome these limitations.
CONCLUSION
We conclude that breastmilk macronutrient levels
correlate to infant growth in unique patterns. Breastmilk
total calories were positively correlated with infant head
circumference. Individual macronutrient levels yielded
no signicant correlations to infant body weight and
length. Rather, the combined eect of the macronutrients
manifested as total calories apparently aected growth.
We also demonstrated that even lower concentrations of
caloric sources (lactose and fat) compared to previous
studies did not negatively aect infant growth. erefore,
we cannot fully support the idea that breastmilk causes
underweight or overweight in infants less than 3 months
old. Based on our ndings, we support the idea of exclusive
breastfeeding, since the evidence showed a positive
relationship toward increased brain volume. Further
studies that are able to overcome our limitations are
needed to conrm our ndings.
ACKNOWLEDGMENTS
e authors wish to thank the Director of Universitas
Airlangga Hospital, Surabaya, Indonesia, for the ethical
approval and laboratory support in this study. All authors
declare no conict of interest related to nancial support or
relationships during the proposal writing, data collection,
analysis, and manuscript writing of this study
REFERENCES
1. Holland D, Chang L, Ernst TM, Curran M, Buchthal SD,
Alicata D, et al. Structural growth trajectories and rates of
change in the rst 3 months of infant brain development.
JAMA Neurol 2014;71:1266-74.
2. Koletzko B, Chourdakis M, Grote V, Hellmuth C, Prell C,
Rzehak P, et al. Regulation of early human growth: impact on
long-term health. Ann Nutr Metab 2014;65:101-9.
3. American Academy of Pediatrics (AAP), American College of
Obstetricians and Gynecologists (ACOG). Breastfeeding
Handbook for Physicians. 2
nd
ed. Washington, DC: AAP; 2014
4. Schwarzenberg SJ, Georgie MK, Committee on nutrition.
Advocacy for improving nutrition in the first 1000 days
to support childhood development and adult health. Pediatrics
2018;141:e20173716.
5. Prentice P, Ong KK, Schoemaker MH, van Tol EAF, Vervoort
J, Hughes IA, et al. Breast milk nutrient content and infancy
growth. Acta Paediatr 2016;105:641-7.
6. Kim J, Donovan SM. Human milk: composition and nutritional
value. In: Caballero PM, Finglas PM, Toldrá, editors. Encyclopedia
of Food and Health. Amsterdam: Elsevier; 2016. p. 357-62.
7. Ballard O, Morrow AL. Human milk composition. Pediatr
Clin North Am 2013;60:49-74.
8. Gruca MA. Breast milk macronutrient composition and infant
growth in rural west Africa. [dissertation on the internet].
Boulder, CO, USA: University of Colorado; 2017. [cited 18 July
2019]. Availble from: https://scholar.colorado.edu/anth_
gradetds/66/
9. Rudolph MC, Young BE, Lemas DJ, Palmer CE, Hernandez
TL, Barbour LA, et al. Early infant adipose deposition is
positively associated with the n-6 to n-3 fatty acid ratio in
human milk independent of maternal BMI. Int J Obes 2017;41:
510-7.
10. Haider R, Saha KK. Breastfeeding and infant growth outcomes in
the context of intensive peer counselling support in two
communities in Bangladesh. Int Breastfeed J 2016;11:1-10.
11. Saure C, Armeno M, Barcala C, Giudici V, Mazza CS. Excessive
weight gain in exclusively breast-fed infants. J Pediatr Endocrinol
Metab2017;30:719-24.
12. Geddes D, Hassiotou F, Wise M, Hartmann P. Human milk
composition and function in the infant. In: Polin RA, Abman
SH, Rowitch DH, Benitz WE, Fox WW, editors. Fetal and
Neonatal Physiology. 5
th
ed. Amsterdam: Elsevier; 2017.p.273-
80.e3. https://doi.org/10.1016/B978-0-323-35214-7.00026-3.
13. Bartholomeusz HH, Courchesne E, Karns CM. Relationship
between head circumference and brain volume in healthy
normal toddlers, children, and adults. Neuropediatrics 2002;33:
239-41.
14. Baker J, Workman M, Bedrick E, Frey MA, Hurtado M, Pearson
O. Brains versus brawn: an empirical test of barkers brain
sparing model. Am J Hum Biol 2010;22:206-15.
15. Aumeistere L, Ciprovica I, Zavadska D, Celmalniece K. Lactose
content of breast milk among lactating women in Latvia.
Foodbalt 2017.p.169-73. https://doi.org/10.22616/foodbalt.2017.023.
16. Chang N, Jung JA, Kim H, Jo A, Kang S, Lee SW, et al.
Macronutrient composition of human milk from Korean
mothers of full term infants born at 37-42 gestational weeks.
Nutr Res Pract 2015;9:433-8.
17. Quinn EA, Largado F, Power M, Kuzawa CW. Predictors of
breast milk macronutrient composition in Filipino mothers.
Am J Hum Biol 2012;24:533-40.
18. Huynh DTT, Tran NT, Nguyen LT, Berde Y, Low YL. Impact
of maternal nutritional supplementation in conjunction with
a breastfeeding support program on breastfeeding performance,
birth, and growth outcomes in a Vietnamese population. J
Matern Fetal Neonatal Med2018;31:1586-94.
19. Gidrewicz DA, Fenton TR. A systematic review and meta-
analysis of the nutrient content of preterm and term breast
milk. BMC Pediatr 2014;14:216.
Martini et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
17
Original Article
SMJ
20. Sever O, Mandel D, Mimouni FB, Marom R, Cohen S, Lubetzky
R. Macronutrients in human milk: colostrum lactose but not
fat or protein predicts mature human milk content. ICAN
Infant, Child, Adolesc Nutr 2015;7:162-5.
21. Kurniati AM, Sunardi D, Sungkar A, Bardosono S, Kartinah
NT. Associations of maternal body composition and nutritional
intake with fat content of Indonesian mothers’ breast milk.
Paediatr Indones 2016;56:298-304.
22. Yang T, Zhang L, Bao W, Rong, S. Nutritional composition of
breast milk in Chinese women: a systematic review. Asia Pac
J Clin Nutr 2018;27:491-502.
23. Gidrewicz DA, Fenton TR. A systematic review and meta-
analysis of the nutrient content of preterm and term breast
milk. BMC Pediatr 2014;14:216.
24. García-Lara NR, Escuder-Vieco D, García-Algar O, De la Cruz
J, Lora D, Pallás-Alonso C. Eect of freezing time on macronutrients
and energy content of breastmilk. Breastfeed Med 2012;7:295-301.
25. Lubetzky R, Sever O, Mimouni FB, Mandel D. Human milk
macronutrients content: eect of advanced maternal age.
Breastfeed Med 2015;10:433-6.
26. Coviello C, Keunen K, Kersbergen KJ, Groenendaal F, Leemans
A, Peels B, et al. Eects of early nutrition and growth on brain
volumes, white matter microstructure, and neurodevelopmental
outcome in preterm newborns. Pediatr Res 2018;83:102-10.
27. Stam J, Sauer PJ, Boehm G. Can we dene an infant’s need from
the composition of human milk? Am J Clin Nutr 2013;98:
521S-8S.
28. Martin CR, Ling PR, Blackburn GL. Review of infant feeding:
key features of breast milk and infant formula. Nutrients
2016;8: 279.
29. Koletzko B, von Kries R, Closa R, Escribano J, Scaglioni S,
Giovannini M, et al. Lower protein in infant formula is associated
with lower weight up to age 2 y: a randomized clinical trial.
Am J Clin Nutr 2009;89:1836-45.
30. Ziegler EE. Growth of breast-fed and formula-fed infants.
Nestle Nutr Workshop Ser Pediatr Program 2006;58:51-63.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
18
Gregory Heng, M.D.*, Varut Lohsiriwat, M.D.**, Kok-Yang Tan, M.D.*
*Department of Surgery, Khoo Teck Puat Hospital, Singapore, **Department of Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University,
Bangkok, ailand.
Suitability of Enhanced Recovery after Surgery
(ERAS) Protocols for Elderly Colorectal Cancer
Patients
ABSTRACT
Objective: Enhanced recovery aer surgery (ERAS) provides a multimodal approach to postsurgical recovery,
seeking to reduce a patient’s stress response and promoting recovery. is study aimed to determine the suitability
of ERAS protocols for elderly patients above 75 years of age.
Methods: is is a retrospective analysis of all patients who had undergone major colorectal resections under ERAS
protocols in Khoo Teck Puat Hospital, Singapore and Faculty of Medicine Siriraj Hospital, ailand between 2013
and 2014. Data collected included patient characteristics and outcomes, including length of hospitalization, and
time to rst atus and mobilization.
Results: Of the 196 patients studied, 38 were above 75 years of age. Elderly patients were more likely to have more
comorbidities, a higher ASA score and a higher POSSUM predicted mortality. ey also had an increased risk of
developing Clavien 2 complications (OR 2.41, 95% CI 1.10-5.29). Compared to their younger counterparts, elderly
patients did not have a delay in rst atus or mobilization. However, they tended to stay longer (7.89 vs. 5.16 days,
p<0.001). On multivariate analysis, ASA score of 3 and above was an independent risk factor for a length of stay
over 1 week while age was not.
Conclusion: is study has shown that elderly patients achieve comparable functional recovery under an enhanced
recovery approach. Enhanced recovery aer surgery can be adopted regardless of a patient’s age.
Keywords: Enhanced recovery; colorectal resections; elderly (Siriraj Med J 2020; 72: 18-23)
Corresponding author: Gregory Heng
E-mail: maniufp@gmail.com
Received 26 February 2019 Revised 7 October 2019 Accepted 10 October 2019
ORCID ID: http://orcid.org/0000-0003-3796-0083
http://dx.doi.org/10.33192/Smj.2020.03
Heng et al.
INTRODUCTION
Enhanced recovery aer surgery (ERAS) had its roots
from the 1990s in the early work of Henrik Kehlet
1
, who
proposed a multimodal approach to postsurgical recovery
that could reduce both postoperative complications
and duration of recovery. Enhanced recovery or ‘fast-
track’ surgery challenges the conventional approach
to postoperative care, and seeks to reduce the patient’s
stress response to surgery and promote recovery. Its key
elements include preoperative counseling, avoidance
of bowel preparation and premedication, reduction
in preoperative fasting, maintenance of intraoperative
normothermia, laparoscopic access where possible,
avoidance of opioid analgesics, nasogastric and intra-
abdominal drains, early removal of urinary catheters and
early mobilization and feeding. At the beginning, these
measures challenged what was then conventional wisdom
and were met with resistance, but numerous publications
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
19
Original Article
SMJ
over the years have validated its ecacy. A meta-analysis
of six randomized controlled trials with 452 patients by
Varadhan et al
2
in 2010 showed a reduction in hospital
stay by 2.55 days (95% CI 1.85-3.24) and a reduction of
complications with a relative risk of 0.53 (95% CI 0.44-
0.64). While enhanced recovery was initially developed
for colorectal surgery, it has now been evaluated or
adopted in a number of other surgeries, including radical
cystectomy, hepatobiliary and pancreatic surgery, total
knee replacement and cardiovascular surgery.
3-7
e management of an elderly patient undergoing
surgery is an evolving challenge. In 2010, 524 million
people were above the age of 65 years - 8 percent of the
world’s population. is number is expected to increase to
1.5 billion people, or 16 percent of the world’s population
by 2050.
8
A natural consequence of the demographic
change has seen more major surgery being performed for
elderly patients.
9,10
Surgery for the elderly is fraught with
complexities as elderly patients have a higher incidence
of co-morbidities, reduced functional reserves and are
more likely to be frail. Duron et al
11
studied 3322 patients
undergoing major digestive surgery in a multicenter
prospective study and found that age above 65 years was
shown to be an independent risk factor for mortality
with an odds ratio of 2.21 (95% CI 1.36-3.59). Surgery
in an elderly patient calls for a holistic approach with
measures taken to address their increased postoperative
risks.
is study aims to evaluate the comparative benet
of enhanced recovery aer surgery (ERAS) protocols
between elderly and young patients undergoing elective
major colorectal resections in an Asian setting.
MATERIALS AND METHODS
All patients who had undergone elective major
colorectal resections under ERAS protocols in the Khoo
Teck Puat Hospital, Singapore and the Faculty of Medicine
Siriraj Hospital, ailand between January 2013 and
December 2014 were reviewed. e elements of ERAS that
were adopted in their care are detailed in Table 1. Data
from our prospectively collected computer database was
extracted and further clinical information was extracted
from a review of the clinical notes. Individual comorbidities
were recorded and quantied using the American Society
of Anesthesiologist (ASA) system. Preoperative surgical
risk was quantied with the Colorectal Physiological and
Operative Severity for the enUmeration of Mortality
and Morbidity (CR-POSSUM)
12
score. Body mass index
(BMI), stage of malignant disease, site of surgery relative
to the peritoneal reection and surgical approach (open
vs. laparoscopic) were noted.
TABLE 1. Enhanced Recovery aer Surgery (ERAS) Protocol.
Variable Age
Preoperative Patient education and counselling
Anaesthesia assessment
Lowresiduedietuntildaybeforesurgery,clearuidsuntil2hoursbeforesurgery
Fluids for carbohydrate loading
Avoidance of bowel preparation (except for low rectal lesion)
Intraoperative Epidural analgesia for open surgery (upper midline)
Mechanical venous thromboemolism and antibiotic prophylaxis
Zero-balanceuidmanagement
Bilateral TAP block for laparoscopic/subumbilical incisions +/- PCA
Removal of NG tube before extubation
Postoperative Allowfreeuidsattheearliestopportunity
Giveuidsat1ml/kg/hr-PromptcessationofIVuids
Discontinue PCA by POD2, Epidural/TAP cathethers by POD3
Oral analgesia prescribed - Paracetamol +/- NSAIDS
Sit out of bed and ambulation by physiotherapist from POD1
Removeurinarycathetheriftherearenospecicrequirements
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
20
Outcome measures of morbidity were quantied with
the Clavien scoring system.
13
Other outcome measures
included the length of stay, day of rst atus or mobilization,
the need for re-operations and postoperative mortality.
Analysis for factors correlating to the development of
postoperative complications and mortality were performed
using factors that were identied to be useful predictors
by previous studies. e outcomes of patients with age
over 75 years were compared with those with a younger
age.
Bivariate analysis was performed using Chi-square
test in Statistical Package for the Social Sciences for
Windows (SPSS Inc, Chicago, USA), version 20.0 on
a personal computer. Results were expressed as odds
ratios with 95% condence intervals. Stepwise logistic
regression analysis was used in multivariate analysis
to identify parameters that independently had aected
outcomes.
Institutional review board statement: e study was
approved by the institutional review board [DSRB Reference
Number: 2017/00130 and SIRB Number: Si 482/2560].
RESULTS
Among the 196 patients who underwent colorectal
resections, 38 were above the age of 75 years. The
demographic and characteristics of the patients who
were operated under ERAS protocols are shown in Table 2.
Elderly patients above 75 years of age were more likely to
have two or more co-morbidities, with 44.7% of elderly
patients having an ASA score of 3 and above compared
to 17.7% (p<0.001) for younger patients. Elderly patients
also had a higher mean CR-POSSUM predicted mortality
of 5.86% compared to 1.79% (p<0.001) for younger
patients. ere was no signicant dierence in gender,
BMI, stage of disease and site of surgery.
Bivariate analysis showed that elderly patients had
increased risk of postoperative morbidity. e odds ratio
for developing Clavien 2 complications and above was
2.41 (95% CI 1.10 – 5.29) for elderly patients (Table 3).
However, age was not shown to be an independent risk
factor for developing Clavien 2 complications and above
on multivariate analysis (Table 4). Instead, having two or
more co-morbidities and BMI above 25 were found to be
independent risk factors. e elderly were not shown to
TABLE 2. Patient Demographics and Operation Characteristics.
Variable Age P-value
Above 75 Below 75
Male gender 39.5% 55.5% 0.072
(15/38) (88/158)
2 or more co-morbidities 63.2% 28.5% <0.001
(24/38) (45/158)
ASA 3 and above 44.7% 17.7% <0.001
(17/38) (28/158)
Mean POSSUM predicted Mortality (SD) 5.86% 1.79% <0.001
(4.43) (1.52)
BMI > 25 21.1% 32.3% 0.176
(8/38) (51/158)
Stage 3 disease and above 36.8% 47.5% 0.238
(14/38) (75/158)
Belowperitonealreection 39.5% 50.9% 0.244
(15/38) (79/158)
Laparoscopic 18.4% 36.1% 0.037
(7/38) (57/158)
Heng et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
21
Original Article
SMJ
TABLE 3. Patient outcomes on bivariate analysis (Odds Ratio).
TABLE 4. Multivariate analysis correlating to Clavien 2 or more complications and length of stay ≥1 week.
Variable Age Odds Ratio P-value
Above 75 Below 75 (95% CI)
Reoperation 2.6% 1.9% 1.40 0.774
(1/38) (3/158) (0.14 - 13.81)
Clavien≥2complications 34.2% 17.7% 2.41 0.025
(13/38) (28/158) (1.10-5.29)
Clavien≥3complications 5.3% 3.8% 1.41 0.682
(2/38) (6/158) (0.94 - 1.10)
Hospitalization death 2.6% 0.6% 4.24 0.271
(1/38) (1/158) (0.26 - 69.42)
Factor Clavien≥2Complications Lengthofstay≥1week
OR (95% CI, P-value) OR (95% CI, P-value)
Age >75 1.46 1.22
(0.59 – 3.63, 0.414) (0.45 – 3.33, 0.701)
Male gender 1.02 1.00
(0.49 – 2.16, 0.950) (0.44 – 2.27, 0.992)
2 or more co-morbidities 2.61 2.06
(1.13 – 6.00, 0.024) (0.82 – 5.18, 0.127)
ASA 3 and above 1.66 2.58
(0.69 – 3.97, 0.256) (1.02 – 6.54, 0.460)
BMI >25 2.62 2.50
(1.05 – 6.52, 0.038) (0.91 – 6.89, 0.076)
Stage 3 disease and above 0.85 1.45
(0.40 – 1.81, 0.670) (0.64 – 3.31, 0.373)
Belowperitonealreection 0.82 0.83
(0.37 – 5.89, 0.577) (0.36 – 1.89, 0.652)
Laparoscopic 1.29 0.98
(0.58 – 2.87, 0.529) (0.40 – 2.42, 0.964)
be at signicant risk for developing more major (Clavien
3 and above) complications. ere was no statistically
signicant dierence in the rate of reoperations or in-
hospital death.
Elderly patients were found to have an increased
length of stay at an average of 7.89 days compared to
5.16 days (p<0.001) in younger patients (Table 5), but
demonstrated good functional recovery, with no signicant
dierence in the day of rst atus or rst mobilization.
Once again, age was not shown to be an independent
risk factor for a length of stay beyond one week on
multivariate analysis (Table 4). An ASA score of 3 and
above was shown to be an independent risk factor for
increased length of stay.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
22
TABLE 5. Patient outcomes (comparison of means).
Variable Age
Above 75 Below 75 P-value
Length of stay (SD) 7.89 days 5.16 days <0.001
(6.24) (3.42)
Dayofrstatus(SD) 1.97days 1.88days 0.638
(1.30) (1.22)
Dayofrstmobilization(SD) 1.50days 1.32days 0.052
(0.95) (0.72)
DISCUSSION
is review of two cohort of patients from Singapore
and ailand showed that elderly patients 75 years of age
and above undergoing major colorectal resections under
ERAS guidelines were more likely to develop Clavien
2 and above complications and to have an increased
length of stay. Although elderly patients experienced
more complication and longer hospital stay, they could
achieve comparable functional recovery as of a younger
age group within an ERAS protocol. Interestingly, old
age was not shown to be an independent risk factor for
either outcome on multivariate analysis. In fact, two
or more co-morbidities and BMI above 25 were found
to be independent risk factors for major postoperative
complication, and ASA score of 3 and above was an
independent risk factor for increased length of stay.
While age above 75 years was associated with the
development of Clavien 2 and above complications, it was
not found to lead to more major complications (Clavien
3 and above) requiring admission to an intensive care
unit or further procedural or surgical re-interventions.
Elderly patients were also not found to have a signicant
dierence in mortality rates. is suggests that enhanced
recovery interventions can be safely applied to elderly
patients. Several previous studies have shown that ERAS
reduced complications and duration of hospitalization
when applied to the elderly, and there was no signicant
difference in outcomes between the elderly and the
young.
14-16
While the present ndings are comparable to
those in the literature, it may be subject to Type 2 sampling
error given the sample size and the low probabilities
of major adverse events, with only one patient in each
cohort (age above or below 75 years) suering from
in-hospital death.
Another consideration for an increased length of stay
in elderly patients was the reduced use of a laparoscopic
approach, which reduces postoperative pain, improves
functional recovery and shortens hospital stay.
17-18
In the present study, elderly patients were more likely
to be selected for open surgery partly due to relative
contraindications to a laparoscopic approach relating
to their co-morbidities.
Despite an increased length of stay, there was no
signicant dierence in the functional recovery of an
elderly patient under an enhanced recovery program.
ey had similar average day of rst atus and rst
mobilization as their younger cohort. is suggests that
the elderly derive a comparable benet from enhanced
recovery measures. Furthermore, what dierentiate
elderly patients from their younger counterparts is their
own expectations when they agree to undergo surgery.
Long-term survival is oen not as important as a return
to pre-morbid function and the retention of functional
independence.
19
e application of ERAS protocols to
reduce functional impairment aer surgery could have
an even greater signicance in the elderly population.
CONCLUSION
Compared to a younger patient, an elderly patient
may develop increased complications or require a longer
duration of hospitalization if they have increased co-
morbidities or a higher ASA score. Nevertheless, this
study has shown that elderly patients achieve comparable
functional recovery under an enhanced recovery approach.
Enhanced recovery aer surgery should be adopted
regardless of a patient’s age.
ACKNOWLEDGMENTS
e authors do not have any potential nancial and
non-nancial conicts of interest. e authors contributed
to concept design, data preparation and analysis, and
the writing of the manuscript. e authors declare that
Heng et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
23
Original Article
SMJ
the Department of General Surgery, Khoo Teck Puat
Hospital, Singapore and Faculty of Medicine Siriraj
Hospital, Mahidol University, ailand intellectually
supported us to write this manuscript but there was no
funding source for this review.
REFERENCES
1. Kehlet H. Multimodal approach to control postoperative
pathophysiology and rehabilitation. Br J Anaesth 1996;78:
606-17.
2. Kehlet H, Wilmore DW. Fast-track surgery. Br J Surg 2005;92:
3-4.
3. Arumainayagam N, McGrath J, Jeerson KP, Gillatt DA.
Introduction of an enhanced recovery protocol for radical
cystectomy. BJU Int 2008;101:698-701.
4. Hall TC, Dennison AR, Bilku DK, Metcalfe MS,Garcea G.
Enhanced recovery programmes in hepatobiliary and pancreatic
surgery: a systematic review. Ann R Coll Surg Engl 2012;94:
318-26.
5. Husted H, Troelsen A, Otte K, Kristensen BB, Holm G, Kehlet
H. Fast-track surgery for bilateral total knee replacement. J
Bone Joint Surg Br 2011;93B:351-6.
6. Muehling B, Schelzig H, Steen P, Meierhenrich R,Sunder-
Plassmann L,Orend KH. A prospective randomized trial
comparing traditional and fast-track patient care in elective
open infrarenal aneurysm repair. World J Surg 2009;33:577-85.
7. Salhiyyah K, Elsobky S, Raja S, Attia R,Brazier J,Cooper GJ. A
clinical and economic evaluation of fast-track recovery aer
cardiac surgery. Heart Surg Forum 2011;14:E330-4.
8. World Health Organization (2011). Global Health and Aging.
In: National Institute on Aging, National Institute of Health,
NIH Publication no. 11-7737, October 2011.
9. Etzioni DA, Liu JH, Maggard MA, Ko CY. e aging population
and its impact on the surgery workforce. Ann Surg 2003;238:
170-7.
10. Christmas C, Makary MA, Burton JR. Medical considerations
in older surgical patients. J Am Coll Surg 2006;203:746-51.
11. Duron JJ,Duron E,Dugue T, Pujol J,Muscari F,Collet D,et al.
Risk factorsformortalityinmajordigestive surgeryin theelderly:
a multicenter prospective study. Ann Surg 2011;254:375-82.
12. Neary WD, Heather BP, Earnshaw JJ. e Physiological and
Operative Severity Score for the enUmeration of Mortality
and morbidity (POSSUM). Br J Surg 2003;90:157-65.
13. Dindo D, Demartines N, Clavien PA. Classication of surgical
complications: a new proposal with evaluation in a cohort of
6,336 patients and results of a survey. Ann Surg 2004;240:
205-13.
14. Tejedor P, Pastor C, Gonzalez-Ayora S, Ortega-Lopez M,
Guardalajara H, Garcia-Olmo. Short-term outcomes and
benets of ERAS program in elderly patients undergoing
colorectal surgery: a case-matched study compared to conventional
care. Int J Colorectal Dis 2018;33:1251-8.
15. Bagnall NM,Malietzis G,Kennedy RH, Athanasiou T,Faiz
O,Darzi A. A systematic review ofenhancedrecoverycare aer
colorectal surgery inelderlypatients. Colorectal Dis 2014;16:
947-56.
16. Kisialeuski M, Pędziwiatr M, Matłok M, Major P,Migaczewski
M,Kołodziej D,et al. Enhanced recovery aer colorectal
surgery in elderly patients. Wideochir Inne Tech Maloinwazyjne
2015;10:30-36.
17. Perrin M, Fletcher A. Laparoscopic abdominal surgery. Contin
Educ Anaesth Crit Care Pain 2004;4:107-10.
18. Weber DM. Laparoscopic surgery: an excellent approach in
elderly patients. Langenbecks Arch Surg 2004;389:204-8.
19. Chee J, Tan KY. Outcome studies on older patients undergoing
surgery are missing the mark. J Am Geriatr Soc 2010;58:
2238-40.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
24
Tippawan Liabsuetrakul, M.D., Ph.D.*, Gunilla Lindmark, M.D., Ph.D.**, Pattarawalai Talungchit, M.D., Ph.D.***
*Epidemiology Unit, Faculty of Medicine, Prince of Songkla University, Songkhla 90110, ailand, **International Maternal and Child Health,
Department of Women’s and Children’s Health, Uppsala University Hospital, SE-751 85 Uppsala, Sweden, ***Department of Obstetrics and Gynecology,
Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, ailand.
Multifaceted Intervention to Improve the Quality of
Care for Postpartum Hemorrhage
ABSTRACT
Objective: To evaluate the eect of a multifaceted intervention on the quality of care and clinical outcomes for
postpartum hemorrhage (PPH) as measured by adherence and acceptability to indicators of PPH management at
district and referral levels.
Methods: A quasi-experimental study of interventions was conducted to improve healthcare for PPH in 6 district
and 3 referral hospitals in southern ailand. Multifaceted intervention included educational outreach, audit
and feedback, reminders, and the involvement of an opinion leader. Physicians and nurses responsible for case
management and care policies (145 individuals) in the aforementioned hospitals participated in the intervention.
Medical records were reviewed considering the checklist of indicators for district and referral levels.
Results: All indicators for PPH management were accepted by at least 80% of the participants, except for surgical
intervention. We reviewed the medical records of 805 women diagnosed with PPH. Of these, during the pre- and
post-intervention periods, 132 and 142 were from district hospitals, and 228 and 303 from referral hospitals,
respectively. e use of an indwelling bladder catheter and hemodynamic monitoring were signicantly increased
in the post-intervention period.
Conclusion: Quality of care for postpartum hemorrhage improved aer implementing a multifaceted intervention
targeting healthcare providers. e results of this study can be applied to other hospitals with a similar setting
regarding the eligible criteria. Severe maternal morbidity and mortality from PPH should be monitored.
Keywords: Indicator; multifaceted intervention; postpartum hemorrhage; quality of care (Siriraj Med J 2020; 72: 24-32)
Corresponding author: Pattarawalai Talungchit
E-mail: pattarawalai.tal@mahidol.ac.th
Received 20 March 2019 Revised 8 November 2019 Accepted 11 November 2019
ORCID ID: http://orcid.org/0000-0002-3668-8754
http://dx.doi.org/10.33192/Smj.2020.04
Liabsuetrakul et al.
INTRODUCTION
Postpartum hemorrhage (PPH), dened as a blood
loss of 500 mL or more within 24 h following parturition,
is a common obstetric complication contributing to
maternal near-miss morbidity and mortality.
1
A systematic
review of the literature showed that PPH is a leading
cause of maternal death in developed countries.
2
e
global prevalence of hemorrhage in maternal deaths
was reported as 13.4%, but 10%-20% in Thailand.
3
e World Health Organization (WHO) updated the
guidelines for PPH prevention and treatment in 2012
focusing on uterine atony and a retained placenta, which
are the most common causes of PPH.
4
However, these
guidelines are recommended practices in general and
are not focused on hospital contexts in particular. As a
result, the guidelines are not used eciently and clinical
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
25
Original Article
SMJ
practice based on evidence-based guidelines remains a
challenge.
5
Moreover, the translation of knowledge into
action is limited in most developing countries.
6
Although evidence-based guidelines for best practice to
prevent and treat PPH have been promoted,
4,7
these practice
guidelines are not always adhered to appropriately.
8-10
To improve the adherence to standard practices, indicators
are an important tool to aid evaluating and monitoring
quality improvement.
11
Scientic evidence for developing
appropriate indicators to improve the quality of care is
essential.
12
Acceptability of the indicators by healthcare
providers is vital to develop appropriate indicators to
measure the quality of a process.
13–14
We previously
developed process indicators for PPH management for
district- and referral-level hospitals from the evidence-
based guidelines, which a healthcare provider agreed were
feasible to monitor.
15
e report of a study conducted in e
Netherlands described the development of quality indicators
to prevent and manage PPH, but their ultimate indicators
diered from those in our present study.
16
A systematic
review found that didactic traditional continuing medical
education or mailing the dissemination of knowledge alone
did not change the behavior of healthcare professionals
signicantly.
17
erefore, comprehensive methods should
be used to introduce to healthcare providers the practice
of implementing and correctly utilizing indicators.
A clinical audit is a commonly used method to reduce
the frequency of severe complications and improve the
management of PPH.
18,19
However, it has been argued
that conducting an audit alone is not sucient to change
the practice of medical professionals and that the use of
multiple approaches, known as a multifaceted intervention,
including audit and feedback, educational outreach,
the involvement of opinion leaders, and reminders, is
eective.
20,21
Multifaceted intervention has also been
reported as an eective strategy to improve the adherence
to standard practice related to PPH.
20
However, there
are a variety of interventions that inuence the clinical
practice depending on the context of the facilities.
21
erefore, the present study aimed to evaluate the eect
of a multifaceted intervention on the quality of care for
PPH as measured by the adherence and acceptability to
management indicators as well as the clinical outcomes
at the district and referral levels.
MATERIALS AND METHODS
Study design and context
We conducted a quasi-experimental before-and-aer
study in Songkhla province, southern ailand, which
has various hospital levels and contexts, to assess the
change of practice related to PPH based on management
indicators. is study was approved by the Institutional
Ethics Committee of the Faculty of Medicine, Prince of
Songkla University (approval No. 53-108-18-1-3) and
the Ethical Review Committee for Research in Human
Subjects, Ministry of Public Health, ailand (Si 45/2010)
in June 2010. All the physicians and nurses participated
in the study were informed and signed their consent.
A total of 9 hospitals, including 3 referral hospitals
(a university, a regional, and a provincial hospital) and 6
district hospitals in Songkhla province, with a previous
record of at least 30 deliveries per month and the highest
complications in the province, were chosen to acquire
sucient cases of PPH. e obstetricians and the general
doctors had responsibilities as the team leaders at the
referral and provincial hospital, respectively. Although
guidelines for PPH management existed in the participating
hospitals, they were described variously, and unmonitored.
A checklist of indicators, which were considered feasibly
relevant, measurable, and improvable, was used to audit
PPH management. Details of the development process,
reliability, and validity are described in our previously
published manuscript.
15
e indicators for district hospitals included a correct
diagnosis of PPH, and general management, such as the
administration of intravenous uids, monitoring of vital
signs, and use of an indwelling bladder catheter. For
referral hospitals, blood cross-matching was added as an
indicator of general management. Specic management
indicators common to both hospital levels included
performing a uterine massage and providing oxytocin
or methylergometrine in the case of uterine atony and
aer removal of the placenta in the case of a retained
placenta. Prostaglandin E2, surgical and radiological
interventions for uterine atony, and removal of the
placenta under general anesthesia were additionally
indicated at the referral level.
Intervention
e intervention included educational outreach,
audit and feedback, reminders, and the involvement
of an opinion leader from August to September 2010.
Before educational outreach, an audit was conducted
by reviewing the medical records of women with PPH
in the participating hospitals. en the research team
visited the study hospitals to give information on the
magnitude of PPH and the need for prompt appropriate
PPH management. In addition, feedback on the audit
information related to PPH management and the outcomes
of practice based on the indicators was provided to
healthcare providers in this educational outreach visit.
In addition, a chart of quality indicators was placed in
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
26
the labor room, postpartum ward, and emergency room
as a reminder of the PPH guidelines. ese activities
were facilitated by the opinion leaders in the hospitals.
Data collection
e medical records of women with PPH who had
delivered in the participating hospitals from October
2008 to September 2009 (pre-intervention period) were
audited by 2 standardized researchers using an indicator
checklist. en, educational outreach activities were
conducted in each participating hospital during August
and September 2010. All the physicians and nurses who
were mainly responsible for the management of PPH at
each hospital were invited to participate in the study.
Discussions on each indicator for the management of
PPH, and the challenges with possible solutions were
recorded. Aer the intervention, the medical records of
women with PPH who had delivered between October
2010 to September 2011 (post-intervention period) were
audited by the same researchers using the same checklist.
A diagram of the study ow and time frame is shown in
Fig 1.
Variables
Information on the healthcare providers participating
in the intervention, including gender, age, occupational
status, and duration of work experience, were collected.
The acceptability of each indicator was determined
using a constructed self-administered questionnaire.
e acceptability was assessed as “the degree to which
they would accept the indicator in their practice” and
was measured using a 5-item Likert-type rating scale,
ranging from strongly unacceptable to strongly acceptable.
Patient characteristics, including maternal age, parity,
gestational age at delivery, and route of delivery, were
recorded. Practicing each indicator was evaluated as
“yes”, “no”, or “no data”. Maternal outcomes, such as
amount of blood loss in 24 h aer delivery, receiving
a blood transfusion, shock, hysterectomy, length of
hospital stay, and referral status, were assessed. Neonatal
outcomes included Apgar score < 7 at 5 min and NICU
admission.
Data analysis
e sample size for the study was based on a target
for improvement dened as 90% adherence to each
indicator, compared with an average baseline adherence
of 70% at the district level and 80% at the referral level
in the pre-intervention period,
9
with a 95% condence
interval and a power of 80%. At least 72 medical records
of PPH at the district level and 219 at the referral level
were required for before and aer the intervention.
e data were entered in EpiData Entry, version
3.1 (Lauritsen JM and Bruus MA, a comprehensive tool
for the validated entry and documentation of data. e
EpiData Association, Odense Denmark, 2008), and
analyzed using R, version 2.10.0 (R Foundation for
Statistical Computing, Vienna, Austria). e acceptability
was analyzed descriptively by hospital level. e maternal
characteristics, adherence to practice of each indicator, and
maternal and neonatal clinical outcomes were compared
before and aer the intervention by hospital level using a
univariate analysis. P < 0.05 was considered as signicant
in the tests of statistical inference.
Fig 1. Diagram of the study ow and time frame
Liabsuetrakul et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
27
Original Article
SMJ
RESULTS
Acceptability of the indicators to healthcare providers
Of the 145 healthcare providers from both district
and referral hospitals, 124 were women (85.5%) and 38
were physicians (26.2%). eir mean age was 34 years
old, with working experience ranging from 1 to 35 years
(median 8 years; IQR 3 to 14 years).
In total, 805 medical records of women with PPH
were reviewed. During the pre- and post-intervention
periods, 132 and 142 were from district hospitals, while
228 and 303 were from referral hospitals, respectively. As
shown in Table 1, all the PPH indicators were accepted
by at least 80% of the participants, except for surgical
intervention. e main reason for the lower acceptability
of surgical intervention was because it requires special
training and high-level competency.
Use of indicators by healthcare providers
e characteristics of mothers during the pre- and
post-intervention periods are shown in Table 2. In district
hospitals, maternal age and parity were not signicantly
dierent, but signicantly higher proportions of preterm
birth and normal labor were found during the post-
intervention period. However, those 2 characteristics
were not signicantly dierent in the referral hospitals.
Table 3 shows the change of adherence to the
indicators for PPH. Signicant improvements, particularly
in the use of indwelling bladder catheter (p < 0.001 and
p = 0.007) and the monitoring of vital signs (p < 0.001
and p = 0.001), were found aer the intervention in
both contexts. Of all the women with PPH, uterine
atony was diagnosed in 78% of women attending district
hospitals and 69% of women attending referral hospitals.
TABLE 1. Acceptability of quality indicators on postpartum hemorrhage rated by healthcare providers.
Type of hospital
Indicators for postpartum hemorrhage District hospitals Referral hospitals
n (%) n (%)
Diagnosis
Registration of estimated blood loss 75 (88.2) 58 (96.7)
General management
Administrationofintravenousuid 85(100.0) 60(100.0)
Indwelling bladder catheter 74 (87.0) 54 (90.0)
Monitoring of blood pressure and pulse rate every 15 min 69 (81.2) 52 (86.7)
within 2 h of diagnosis and recording urine output every hour
Cross-matching blood NA 53 (88.3)
Specicmanagement
Uterine atony
Performing uterine massage 83 (97.6) 60 (100.0)
Administering oxytocin 77 (90.5) 56 (93.3)
Administering methylergometrine 75 (88.2) 57 (95.0)
Administering prostaglandin E2 NA 52 (86.7)
Surgical intervention NA 42 (70.0)
Retained placenta
Manual removal of placenta NA 53 (88.3)
Uterine massage and administration of uterotonic drugs 78 (91.7) 58 (96.7)
after removing the placenta
Abbreviation: NA = not applicable
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
28
District hospitals Referral hospitals
Intervention period Pre-intervention Post-intervention Pre-intervention Post-intervention
(n = 132) (n = 142) P* (n = 228) (n = 303) P*
n (%) n (%) n (%) n (%)
Age (years), mean (SD) 28.61 (6.69) 28.49 (6.72) 0.80 28.8 (6.8) 29.27 (7.15) 0.99
Age group (years) 0.49 0.39
≤20 10(7.6) 16(11.3) 13(5.7) 25(8.3)
20-34 92 (69.7) 100 (70.4) 149 (65.4) 183 (60.4)
≥35 29(22.0) 26(18.3) 66(28.9) 94(31.0)
Missing data 1 (0.7) 0 0 1 (0.3)
Parity 0.66 0.66
Nulliparity 36 (27.3) 36 (25.4) 75 (32.9) 106 (35)
Multiparity 90 (68.2) 96 (67.6) 145 (63.6) 190 (62.7)
Grand multiparity 6 (4.5) 10 (7.0) 8 (3.5) 7 (2.3)
Gestational age group 0.38 0.17
<37 weeks 9 (6.8) 15 (10.6) 46 (20.2) 63 (20.8)
≥37weeks 122(92.5) 127(89.4) 182(79.8) 236(77.9)
Missing data 1 (0.7) 0 0 4 (1.3)
Route of delivery 0.02 0.86
Normal labor 92 (69.7) 112 (78.9) 123 (53.9) 151 (49.8)
Cesarean section 25 (18.9) 26 (18.3) 82 (36.0) 107 (35.3)
Operative delivery 15 (11.4) 4 (2.8) 23 (10.1) 33 (10.9)
Missing data 0 0 0 12 (4.0)
TABLE 2. Characteristics of women with postpartum hemorrhage before and aer the intervention.
Number in the table are frequency and percent unless stated; SD, standard deviation
*Continuous data analyzed by unpaired t test; categorical data omitting missing data analyzed by Chi-square test.
e implementation of specic management for uterine
atony and for a retained placenta in district hospitals
was not signicantly dierent between the two periods,
while the implementation of uterine massage in uterine
atony (p < 0.001) and manual removal of a retained
placenta (p < 0.001) were signicantly improved in the
referral hospitals.
e clinical outcomes of women with PPH are shown
in Table 4. Neither maternal nor neonatal outcomes were
signicantly changed aer the intervention. More than
half of those who were diagnosed in the referral hospitals
had a blood loss of 1,000 mL or more and received a
blood transfusion. Hysterectomy was performed in
10% of cases. Few women diagnosed as PPH showed
an estimated blood loss of less than 500 mL; however,
the nurse records revealed that vaginal bleeding was
continuously observed within 24 h of parturition without
quantitative estimation.
DISCUSSION
e management of PPH as monitored by quality
indicators was improved aer multifaceted intervention
was implemented aimed at physicians and nurses. However,
the eect on clinical outcomes was not identied in
this study. A signicant improvement of active care,
particularly in the use of indwelling bladder catheters
Liabsuetrakul et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
29
Original Article
SMJ
TABLE 3. Percentage adherence to the quality indicators for the care of postpartum hemorrhage.
District hospitals Referral hospitals
Intervention period Pre- Post- Pre- Post-
intervention intervention intervention intervention
(n = 132) (n = 142) P* (n = 228) (n = 303) P*
n (%) n (%) n (%) n (%)
General management
Indwelling bladder
catheter
58 (43.9) 134 (94.4) < 0.001 173 (75.9) 259 (85.5) 0.007
Monitoring of blood pressure 88 (66.7) 141 (99.3) < 0.001 209 (91.7) 297 (98) 0.001
and pulse rate every 15 min
within 2 h of diagnosis and
recording urine output every hour
Administrationofintravenousuid 130(98.5) 142(100) 0.23 228(100) 302(99.7) 1
Cross matching NA NA NA 207 (90.8) 277 (91.4) 0.92
Specicmanagement
Uterine atony n = 104 n = 109 n = 158 n = 208
Performing uterine massage 103 (99) 102 (93.6) 0.07 105 (66.5) 174 (83.7) < 0.001
Administering uterotonic drugs 101 (97.1) 109 (100.0) 0.12 156 (98.7) 206 (99) 1
Surgical intervention 0.81
Vessel ligation NA NA 2 (1.3) 2 (1)
B-Lynch NA NA 3 (1.9) 6 (2.9)
TAH NA NA 20 (12.7) 21 (10.1)
Refer 8 (7.7) 11 (10.1) 0.71 NA NA
Retained placenta n = 27 n = 32 n = 54 n = 72
Performing manual removal of 16 (59.3) 25 (78.1) 0.20 42 (77.8) 71 (98.6) < 0.001
placenta
Performing uterine massage 16 (59.3) 25 (78.1) 0.20 54 (100.0) 70 (97.2) 0.51
and uterotonic drug given
after removal of placenta
Refer 11 (40.7) 10 (31.2) 0.627 NA NA
Numbers in table are frequency and percent. *Analyzed by Chi-square test or Fisher’s exact test as appropriate.
Abbreviations: NA = not applicable; TAH = total abdominal hysterectomy
and the monitoring of vital signs, was found aer the
interventions in both district and referral hospitals. e
adherence to the quality indicators was related to their
acceptability from the healthcare providers. However, a
signicant improvement of adherence to the indicators aer
the intervention could not be identied for all indicators
due to the low prevalence of certain specic conditions
(e.g., retained placenta at a district hospital) or a high
baseline adherence > 90% before the intervention. On
the other hand, some specic quality of care indicators
(uterine massage and manual removal of a retained
placenta) were signicantly improved at only referral
hospitals because of the low baseline adherence and
higher number of cases there than in district hospitals.
A multifaceted intervention with various approaches,
including audit and feedback, educational outreach,
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
30
TABLE 4. Clinical outcomes of women with postpartum hemorrhage.
District hospitals Referral hospitals
Intervention period Pre- Post- Pre- Post-
intervention intervention P* intervention intervention P*
(n = 132) (n = 142) (n = 228) (n = 303)
n (%) n (%) n (%) n (%)
Maternal complications
Amount of blood loss (mL) 0.50 0.51
<500 7 (5.3) 3 (2.1) 13 (5.8) 11 (3.6)
500-999 74 (56.1) 86 (60.6) 92 (41.1) 116 (38.4)
1,000-2,499 49 (37.1) 52 (36.6) 96 (42.9) 137 (45.4)
≥2,500 2(1.5) 1(0.7) 23(10.3) 38(12.6)
Blood transfusion 36 (27) 28 (34) 0.39 131 (57.5) 163 (53.8) 0.45
Shock 21 (15.9) 25 (17.6) 0.83 26 (11.4) 27 (8.9) 0.42
Hysterectomy 1 (1) 1 (0.7) 1 23 (10.1) 32 (10.6) 0.97
Length of hospital stay (days) 0.70 0.91
Median, IQR 3 (2,4) 3 (2,3) 4 (3,6) 4 (3,5)
Refer 19 (14.4) 22 (15.5) 0.93 NA NA
Neonatal complications
Apgar scores <7 at 5 min after birth 2 (1.6) 2 (1.4) 1 9 (4.1) 17 (6.2) 0.41
NICU admission NA NA 15 (6.6) 18 (5.9) 0.91
Numbers in table are frequency and percent except for length of hospital stay (median (IQR, interquartile range)). *Analyzed by Chi-square
test or Fisher’s exact test as appropriate.
Abbreviation: NA = not applicable
involvement of local opinion leaders, and reminders,
was chosen as the intervention in our study because
there is evidence that these interventions can change the
behavior of healthcare professionals.
17,21–23
A literature
search in PubMed performed on October 14, 2019, using
the terms “multifaceted intervention” and “postpartum
hemorrhage”, revealed 6 articles, including a controlled
trail
24
and 2 randomized-controlled trials
25–26
, as well
as 3 before-and-aer quasi-experimental trials.
20,27,28
Although these studies showed a variation in the outcomes
monitored of PPH, the use of multifaceted intervention
was a common strategy for preventing PPH and improving
its management.
29
In the present study, the healthcare providers who
worked in the participating hospitals contributed to the
design process to aid considering whether the management
strategies were feasible and practical in their contexts
and would be supported by policymakers before the
strategies were implemented. Details of the development
and assessment of the quality of care indicators were
published previously.
15
Apart from the importance of
the content disseminated, educational outreach assists
the provision of an audit of the relevant indicators in the
clinical setting, and contributes to an opportunity for
discussion between the members of the care team. e
involvement of opinion leaders and reminders are helpful
for encouraging the continuity of good practice. erefore,
multiple approaches as a multifaceted intervention have
been proven to be eective in changing the practice of
medical professionals.
17,20,21
e accuracy of estimating blood loss during delivery
and immediately postpartum is a challenge. Visual estimation
Liabsuetrakul et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
31
Original Article
SMJ
is common practice but has low sensitivity, leading to
the underestimation of PPH in 89% of cases.
30
Although
various techniques were used in previous studies, no
denite method is recommended in clinical practice.
31
Guidelines for PPH management were dierent in each
context.
18,20
Our management criteria were developed from
evidence-based guidelines and based on the acceptability
of healthcare providers as quality indicators.
15
Adherence
to practices such as cross-matching of blood and vital
sign monitoring both pre- and post-intervention periods
in our study were better than those reported in a study
in Malawi.
19
Signicant improvements in the use of
indwelling bladder catheters and the monitoring of vital
signs was found aer the interventions in this study.
ese practices are important for the early detection of
hemorrhagic shock.
8
e study has a few limitations to report. First, a
study using an interventional design may be subject
to possible selection and outcome bias. However, the
indicators of PPH management were applied to all
diagnosed cases, regardless of the patient age, parity,
gestational age, or route of delivery. Second, the results
of a quasi-experimental before-and-aer study may be
aected by temporal changes. However, the outcomes
in the present study were registered immediately aer
intervention, which minimizes the outcome change over
time. In addition, the benet of multifaceted intervention
for PPH has already been shown in randomized-controlled
trials, thus a quasi-experimental interventional study is a
reasonable choice for the study design. Finally, this study
could not identify the eect of the improvement of the
quality of care on maternal and neonatal clinical outcomes
because this was not the main objective of the present
study, therefore the sample size was not calculated for
improving clinical outcomes. A well-designed clinical trial
to determine the eect of implementing the indicators
on the quality of care and the incidence of complications
is thus warranted.
CONCLUSION
e use of indicators for the quality of care for
PPH was acceptable to healthcare practitioners and their
adherence to indicators was improved aer implementing
a multifaceted intervention targeting the practitioners.
e guideline for the development and implementation
of indicators can be applied to other hospitals with a
similar setting to the eligible criteria of the present study.
ACKNOWLEDGMENTS
is study is supported by the ailand Research
Fund and Oce of the Higher Education Commission via
the Royal Golden Jubilee PhD Program and a Research
Chair Grant from the National Science and Technology
Development Agency (NSTDA) [Grant number P-10-
10307 to TL]. We thank all participants for their support
and for providing fruitful opinions and comments on
the indicators developed.
Conicts of Interest Statement: e authors have each
completed and submitted an ICMJE Uniform Disclosure
Form for Potential Conicts of Interest. e authors
declare there are no conicts of interest to report.
REFERENCES
1. Acuin CS, Khor GL, Liabsuetrakul T, Achadi, EL, Htay TT,
Firestone R, et al. Maternal, neonatal, and child health in southeast
Asia: towards greater regional collaboration. Lancet 2011;377:
516-25.
2. Rossi AC, Mullin P. e etiology of maternal mortality in
developed countries: a systematic review of literature. Arch
Gynecol Obstet 2012;285:1499-503.
3. Khan KS, Wojdyla D, Say L, Gülmezoglu AM, Van Look PF.
WHO analysis of causes of maternal death: a systematic review.
Lancet 2006;367:1066-74.
4. World Health Organization. WHO recommendations for the
prevention and treatment of postpartum haemorrhage. Geneva:
WHO, 2012.
5. Grol R. Successes and failures in the implementation of evidence-
based guidelines for clinical practice. Med Care 2001;39
(8 Suppl 2):II46-54.
6. Haines A, Kuruvilla S, Borchert M. Bridging the implementation
gap between knowledge and action for health. Bull World
Health Organ 2004;82:724-31.
7. World Health Organization. Managing complications in
pregnancy and childbirth. A guide for midwives and doctors.
Geneva: WHO, 2007.
8. Lawton B, MacDonald EJ, Brown SA, Wilson L, Stanley J, Tait
JD, et al. Preventability of severe acute maternal morbidity.
Am J Obstet Gynecol 2014;210:557.e1-6.
9. Talungchit P, Liabsuetrakul T. Clinical audit of postpartum
hemorrhage at district-level and referral-level hospitals in
southern ailand. J Med Assoc ai 2012;95:1244-51.
10. Liabsuetrakul T, Promvijit T, Pattanapisalsak C, Silalai S,
Ampawa T. A criterion-based obstetric morbidity audit in
southern ailand. Int J Gynecol Obstet 2008;103:166-71.
11. Donabedian A. Explorations in quality assessment and monitoring.
Ann Arbor: Health Administration Press, 1980.
12. Wollersheim H, Hermens R, Hulscher M, Braspenning J,
Ouwens M, Schouten J, et al. Clinical indicators: development
and applications. Neth J Med 2007;65:15-22.
13. Campbell SM, Braspenning J, Hutchinson A, Marshall M.
Research methods used in developing and applying quality
indicators in primary care. Qual Saf Health Care 2002; 11:
358-64.
14. Campbell SM, Kontopantelis E, Hannon K, Burke M, Barber A,
Lester HE. Framework and indicator testing protocol for
developing and piloting quality indicators for the UK quality
and outcomes framework. BMC Fam Pract 2011;12:85.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
32
15. Talungchit P, Liabsuetrakul T, Lindmark G. Development and
assessment of indicators for quality of care in severe preeclampsia/
eclampsia and postpartum hemorrhage. J Healthc Qual. 2013;
35:22-34.
16. Smit M, Sindram SI, Woiski M, Middeldorp JM, van Roosmalen
J. e development of quality indicators for the prevention
and management of postpartum haemorrhage in primary
midwifery care in the Netherlands. BMC Pregnancy Childbirth
2013;13:194.
17. Chaillet N, Dubé E, Dugas M, Audibert F, Tourigny C, Fraser
WD, et al. Evidence-based strategies for implementing guidelines
in obstetrics: a systematic review. Obstet Gynecol. 2006;108:
1234-45.
18. Dupont C, Deneux-araux C, Touzet S, Colin C, Bouvier-Colle
MH, Lansac J, et al. Clinical audit: a useful tool for reducing
severe postpartum haemorrhages? Int J Qual Health Care 2011;
23:583-9.
19. Kongnyuy EJ, Mlava G, van den Broek N. Using criteria-based
audit to improve the management of postpartum haemorrhage
in resource limited countries: a case study of Malawi. Matern
Child Health J 2009;13:873-8.
20. Audureau E, Deneux-araux C, Lefèvre P, Brucato S, Morello R,
Dreyfus M, et al. Practices for prevention, diagnosis and
management of postpartum haemorrhage: impact of a regional
multifaceted intervention. Br J Obstet Gynaecol 2009;116:
1325-33.
21. Ivers N, Jamtvedt G, Flottorp S, Young JM, Odgaard-Jensen
J, French SD, et al. Audit and feedback: eects on professional
practice and healthcare outcomes. Cochrane Database Syst
Rev 2012;6:CD000259.
22. O’Brien MA, Rogers S, Jamtvedt G, Oxman AD, Odgaard-
Jensen J, Kristoersen DT, et al. Educational outreach visits:
eects on professional practice and health care outcomes.
Cochrane Database Syst Rev 2007;4:CD000409.
23. Flodgren G, Parmelli E, Doumit G, Gattellari M, O’Brien MA,
Grimshaw J, Eccles MP. Local opinion leaders: eects on
professional practice and health care outcomes. Cochrane
Database Syst Rev 2011; 8:CD000125.
24. Liabsuetrakul T, Palanukunwong K, Chinduereh A, Oumudee
N. Evaluation of a multifaceted postpartum hemorrhage-
management intervention in community hospitals in Southern
ailand. Int J Gynecol Obstet 2017;139:39-44.
25. Deneux-araux C, Dupont C, Colin C, Rabilloud M, Touzet S,
Lansac J, et al. Multifaceted intervention to decrease the rate
of severe postpartum haemorrhage: the PITHAGORE6 cluster-
randomised controlled trial. Br J Obstet Gynaecol 2010; 117:
1278-87.
26. Pirkle CM, Dumont A, Traoré M, Zunzunegui MV; QUARITE
group. Eect of a facility-based multifaceted intervention on
the quality of obstetrical care: a cluster randomized controlled
trial in Mali and Senegal. BMC Pregnancy Childbirth 2013;13:24.
27. Althabe F, Mazzoni A, Caerata ML, Gibbons L, Karolinski
A, Armbruster D, et al.; Oxytocin in Uniject Study Group. Using
Uniject to increase the use of prophylactic oxytocin for management
of the third stage of labor in Latin America. Int J Gynecol
Obstet 2011;114:184-9.
28. García-Elorrio E, Aleman A, Caerata ML, Colomar M, Tomasso
G, Lacayo Y, et al. A multifaceted intervention to increase
prophylactic oxytocin use during the third stage of labor and
to reduce routine episiotomies in Nicaragua. Int J Gynecol
Obstet 2014; 127:31-34.
29. Gaucher L, Occelli P, Deneux-araux C, Colin C, Gaucherand P,
Touzet S, Dupont C. Non-clinical interventions to prevent
postpartum haemorrhage and improve its management: A
systematic review. Eur J Obstet Gynecol Reprod Biol 2019;240:300-9.
30. Prasertcharoensuk W, Swadpanich U, Lumbiganon P. Accuracy
of the blood loss estimation in the third stage of labor. Int J
Gynecol Obstet 2000;71:69-70.
31. Schorn MN. Measurement of blood loss: review of the literature.
J Midwifery Womens Health 2010;55:20-27.
Liabsuetrakul et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
33
Original Article
SMJ
Wafaa Salih Rahmatullah, Diploma*, Manal Taha Al-Obaidi, Ph.D.*, Wasan I. AL-Saadi, FIBMS**, Mohammad
Oda Selman, FIBMS CAA*, Ghasak Ghazi Faisal, FIBMS***
*High Institute of Infertility Diagnosis and ARTs /Al-Nahrain University, **Medical College/Al-Nahrain University, ***Department of Fundamental
Dental and Medical sciences, International Islamic University Malaysia, Malaysia.
Role of Vascular Endothelial Growth Factor
(VEGF) and Doppler Sub-endometrial Parameters
as Predictors of Successful Implantation in
Intracytoplasmic Sperm Injection (ICSI) Patients
ABSTRACT
Objective: to investigate the expression of vascular endothelial growth factor (VEGF) in patients having infertility
due to low endometrial acceptance, and to correlate it to non-invasive ultrasound variables, endometrial thickness,
and sub-endometrial Doppler parameters (PI, RI, Vs/Vd).
Methods: 80 women all under the age of 40 underwent ICSI-ET; all patients were exposed to ovarian stimulation
protocols. e oocytes were retrieved using an ultrasound guide, and were fertilised via injection of sperm inside
the follicle (ICSI). VEGF serum level was analysed at day of embryo transfer by ELIZA test, and sub-endometrial
evaluation was conducted via two-dimension power Doppler ultrasound (2D PD-US), by measuring resistance
index (RI) and pulsatility index (PI) on the day of embryo transfer.
Results: ere was a signicantly higher VEGF level and endometrial thickness in pregnant (433± 207 and 9.72±
1.35) women, compared to non-pregnant (276± 165 and 8.95 ± 1.21) respectively as p-values were (0.001 and
0.01)). Additionally, there were signicantly lower RI and PI in pregnant (0.584 ± 0.124 and 0.829 ± 0.301) women
compared to non-pregnant (0.651±0.132 and 1.006±0.335) women, as p-values were (0.02 and 0.02, respectively).
e level of E2 was on the day of embryo transfer and Vs/Vd in pregnant women (1402± 524 and 3.14 ± 3.75) and
in the non-pregnant group (1296± 611 and 3.82 ± 3.07), as p-values were 0.41 and 0.38, respectively.
Conclusion: e combined analysis of endometrial receptivity was completed, and the serum level of VEGF and
sub-endometrial evaluation with 2D PD-US was dened by measuring resistance index (RI) and pulsatility index
(PI) on the day of embryo transfer. ese can serve as useful prognostic methods for the detection of endometrial
receptivity and pregnancy outcomes in infertile women undergoing ICSI protocols, and will be helpful for candidate
counselling about postponing embryo transfer and cryopreservation, which may serve as a better option, to be
recommended for the next cycle, when achieving better endometrial Doppler parameters.
Keywords: VEGF; Sub-endometrial doppler; infertility (Siriraj Med J 2020; 72: 33-40)
Corresponding author: Ghasak Ghazi Faisal
E-mail: drghassak@yahoo.com
Received 9 October 2019 Revised 28 October 2019 Accepted 7 November 2019
ORCID ID: http://orcid.org/0000-0002-4736-7630
http://dx.doi.org/10.33192/Smj.2020.05
INTRODUCTION
Infertility means the inability to have a baby despite
regular unprotected sex.
1
According to NICE guidelines,
two years of having regular sex without the presence of
a known abnormality in both partners is necessary for
establishing the diagnosis of infertility.
2
Management of
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
34
Rahmatullah et al.
these patients have changed and improved dramatically,
ranging from assistedreproduction technology to in vitro
fertilisation (IVF). Intracytoplasmic sperm injection
(ICSI), has helped tremendously in management of
infertility due to male factors.
3
e problem of IVF
arise in the failure of transfer of embryo. at’s way
research has focused on nding the best conditions in
the embryo and the endometrium that will allow proper
transfer. Discovering specic and accurate biomarkers
is a primary concern to many researchers. Additionally,
knowledge of the human implantation window length
on days (6-10) post-ovulation, the effect of steroid
hormones (oestrogen, progesterone, gonadotropins)
on genetic factors, the age of a woman, and functional
and morphological markers pertaining to endometrial
receptivity are of crucial signicance to all future studies
aimed at identifying endometrial receptivity
4
, studying
endometrial receptivity by non-invasive methods like
vaginal ultrasonography. Conventionally, pulsed Doppler
sonography is used to evaluate uterine and endometrial
blood ow, but non standard results have been reported
for the measurement of sub-endometrial resistance index
(RI) and pulsatility index (PI) at the second endometrial
zone (zone 2), with serum level of estradiol at day of
HCG and angiogenic factor VEGF at day of embryo
transfer. is variation is correlated with endometrial
receptivity and pregnancy outcome, and can contribute
to making a judgment as to whether an embryo should
be transferred, or whether to cryopreserve it until it
reaches an optimum level of endometrial receptivity.
Due to the limited mobility of the intravaginal
transducer in the small vaginal cavity, Normal ultrasound is
of limited value. Real-time U/S allows study of two factors
related to implantation: the thickness of endometrium
and the pattern of endometrial morphology.
5
Ultrasound
is crucial for predicting endometrial preparation before
transfer of embryo (ET), in both fresh IVF cycles and
frozen-thawed embryo transfer cycles. Frequently used
parameters for endometrial sonographic assessment
involve endometrial thickness, endometrial patterns,
and Doppler indices.
6
At the myometrial-endometrial
junction, there is a sub-endometrial region that is detected
on ultrasound examination as a thin hypoechoic layer
between the echogenic endometrium and myometrium; this
is referred to as the junctional zone, inner myometrium,
sub- endometrial halo, or sub- endometrial layer, and can
be identied by either ultrasound or MRI. Research has
conrmed that the sub-endometrial haloaround around
the endometrium is representative of the innermost layer
of the myometrium.
7
e vascularity zones are identied
as follows. Zone 1: vessels are detected in the myometrium
around the endometrium; zone 2: vessels penetrate the
hyperechogenic endometrial edge; zone 3: vessels reach
the internal hypo echogenic zone; zone 4: vessels reach
the endometrial cavity, when colour mapping of the
endometrial and sub-endometrial regions is absent; this
means a denite implantation failure or a considerable
decline in the implantation rate.
8
Doppler studies have
revealed that the resistance in arteries of the endometrium
is signicantly reduced during the mesoluteal phase,
which is the period when embryo implantation is possible.
e changes in blood vessels may have a vital role in
the implantation process, as they are present from the
start of the embryo implantation. Vascular endothelial
growth factor (VEGF), belongs to proteins binding
heparin that attach to endothelial cells, and which lead
to proliferation and new blood vessel formation. e
VEGF also causes release of cytokines by the endothelial
cells leading to dilatation of blood vessels. It acts as an
angiogenic factor to promote angiogenesis in various
tissues.
9
Consequently, this explains the crucial role of
angiogenesis in dierent female reproductive processes, e.g.,
development of a dominant follicle, formation of a corpus
luteum, endometrial growth, as well as implantation. As
such, local angiogenesis is considered a major prerequisite
for implantation and subsequent conception.
10
Angiogenesis
was reported as it is least occurring during the menstrual
phase, then there will be an increase during the early
proliferative phase, reaching maximum in mid-cycle
then decrease near the end of the cycle.
10,11
MATERIALS AND METHODS
is is a prospective study was conducted in the High
Institute of Infertility Diagnosis and Assisted Reproductive
Technologies, Al-Nahrain University. e study was
approved by the Local Medical Ethical Committee of
the High Institute of Infertility Diagnosis and Assisted
Reproductive Technologies, Al-Nahrain University. Eighty
women undergoing an IVF/ICSI cycle were included in
this study. At the day of embryo transfer, two dimensional
Power Doppler (2D-PD) ultrasound was conducted for
sub-endometrial blood ow zone 2, with serum VEGF
level analysis.
IVF/ICSI procedures
Step-by-step descriptions of IVF procedures:
1. Controlled ovarian hyper stimulation (COH)
2. Oocyte retrieval (OCR)
3. Fertilisation and embryo culture
4. Embryo quality
5. Embryo transfer (ET)
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
35
Original Article
SMJ
ree dierent types of controlled ovarian hyper-
stimulation (COH) were used according to the demographic
parameters of patients, starting a long agonist protocol,
from day 21 of the preceding cycle and using a GnRH
agonist (Decapeptyl® 0.1 mg, Ferring Co., Germany); then,
on the second day of the menstrual cycle, stimulation was
started with a daily dose of 150-450 IU rFSH (follitropin
alfa, Gonal F®, Merck Serono). Secondly, a short agonist
protocol was initiated, beginning with a GnRH agonist
from cycle day one-to-two, then starting a daily dose
of 150-450 IU rFSH (follitropin alfa, Gonal F®, Merck
Serono). e third option used an antagonist protocol
by starting with a daily dose of 150-450 IU rFSH
(follitropin alfa, Gonal F®, Merck Serono), and when
follicles reached 13-14 mm in size, a GnGH antagonist
was started (Cetrotide®, Merck Serono). Patients were
monitored by transvaginal sonography (TVS). HCG
(Ovitrelle® 250 microgram, Merck Serono) was given
when three or more follicles reached a diameter of 18 mm.
Oocytes retrieval was performed using a transvaginal probe
34-36 hours following the HCG injection, immediately
prior to the rupture of follicles. Oocytes were aspirated by
transvaginal ultrasound guided oocyte retrieval (TUGOR);
oocytes at retrieval can be either the germinal vesicle (GV),
which is the most immature, or from metaphase I (MI),
at which the oocyte is an immature egg. e absence of
a polar body or germinal vesicle indicates it to be at the
MI stage, which is an intermediate stage between the
GV and MII (mature) stages, or metaphase II oocyte
(MII), which is mature. Generally, prior to OCR, a semen
sample is prepared for sperm extraction, following a
minimum of two days and a maximum of seven days’
sexual abstinence. Testicular sperm extraction (TESE)
is a surgical sperm retrieval procedure used in infertility
treatment for men with azoospermia.
At the IVF laboratory, aspirated follicles were
examined. Flushing was performed, and the follicles
are kept one-to-two hours in a 37°C/CO
2
incubator.
Later, all oocytes underwent denudation and grading in
a laminar ow cabinet. ereaer, a needle was carefully
inserted through the shell of the egg into its cytoplasm,
then kept in the CO
2
incubator while waiting for the
results of cell division, which would be detected with the
aid of a Nikon ICSI microscope. Following insemination,
zygotes were observed for 18-20 hours to check for the
presence of two pronuclei, and for 25-29 hours to conrm
the existence of early cleavage, which was correlated with
higher implantation rates. At day one, the presence of
two pronuclei was considered a good prognostic sign.
At day two-or-three post-OCR, and according
to the number and grading of the embryos, patients
were prepared for the transfer. The patient's serum
was obtained for VEGF serum level, two dimensional
transvaginal ultrasound scans were done to measure
endometrial thickness, regularity, and echogenicity, and
sub-endometrial blood ow colour Doppler indices (PI, RI
and Vs/Vd) were measured. e measurement involved
both endometrial layers, excluding the surrounding low
amplitude echo layer; three measurements were taken
and the average value was recorded. A pulsed Doppler
system was used for blood ow analysis. Sub-endometrial
vessels were visualised at the endometrial periphery,
sometimes penetrating the hyperechogenic endometrial
edge, or even reaching the endometrial cavity.
We measured the endometrial thickness on the day
of ET as we want it to be more accurately correlated with
the level of VEGF which was measured on the same day.
Fig 1. Ultrasound review reveals myometrial, endometrial, and sub-
endometrial zones.
Blood flow velocity waveforms from the sub-endometrial
vessels were obtained by placing the Doppler gate over
the coloured area at zone 2 and activating the pulsed
Doppler function.
RI =
PSV−EDV
PI =
PSV−EDV
mv =
PSV+EDV
PSV mv 2
Abbreviations: RI = resistance index; PSV = peak systolic velocity;
EDV = end-diastolic velocity; PI = pulsatility index; mv = mean
velocity
RESULTS
Demographic parameters and hormonal prole
Regarding the demographic parameters (age,
BMI of patients, and duration of infertility) and the
basal hormonal prole level (FSH, LH, PRL, E2, and
testosterone), the statistical analysis showed no signicant
dierences between pregnant and non-pregnant ICSI
patient groups, and p-values of 0.516, 0.06, and 0.622, and
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
36
Rahmatullah et al.
0.289, 0.495, 0.334, 0.782, 0.568, and 0.342, respectively
(Table 1).
Vascular endothelial growth factor (VEGF) and Doppler
parameters
Table 2 demonstrates the dierences between some
measurements in pregnant and non-pregnant women
in this study. Endometrial thickness, RI, PI, and VEGF
showed statistically signicant dierences, with p-values
equaling 0.01, 0.021, 0.015, and 0.001, while Vs/Vd
showed no statistically signicant dierences, with a
p-value of 0.385.
Correlation of VEGF with other endometrial receptivity
parameters
Table 3 shows that the Pearson correlation for VEGF,
and E2 at day of HCG administration and endometrial
thickness at day of embryo transfer, showed a positive
association above zero (r= 0.325, 0.227), with a statistically
signicant value of association (p-value equal to 0.003
and 0.043). Pearson correlation for VEGF and RI, PI, and
Vs /Vd showed a weak negative association below zero
(r= -0.168, -0.11, -0.088), with no statistically signicant
value of association (p-value equal to 0.137, 0.331, and
0.436).
Correlation of endometrial thickness with other
endometrial receptivity parameters
e Pearson correlation of endometrial thickness
with Vs/Vd showed a positive association above zero
(r=0.23), with a statistically signicant association value
(p-value 0.04), while Pearson correlation for endometrial
thickness with RI, PI, and E2 at D. of HCG administration
showed weak association above zero (r=0.062, 0.079, and
0.129), with no statistically signicant association value
(p-value=0.586, 0.489, and 0.253), as shown in Table 4.
Correlation of E2 at day of embryo transfer with
endometrial receptivity parameters
Regarding E2 level at day of HCG administration,
its Pearson correlation with RI, PI, and Vs/Vd showed
weak negative association below zero (r=-0.01, -0.031,
and -0.107, respectively), with no statistically signicant
association value (p-value=0.933, 0.786, and 0.346,
respectively), as shown in Table 5.
DISCUSSION
The data for the present study show that the
pregnant group had higher levels of VEGF compared
to the non-pregnant group, as shown in Table 2. e
rate of endometrial blood ow during the normal female
reproductive cycle has been correlated with increased
expression of angiogenic factors (e.g., VEGF). ese are
members of a family of heparin binding proteins that act
directly on the endothelial cells, and induce proliferation
and angiogenesis.
12
Zenneni et al. found that patients with
primary infertility had a much less immunohistochemical
expression of VEGF in the secretory endometrium.
13
In another study, Hannan et al conrmed the presence
of thirty dierent types of mediators present in the
myometrial cavity during the menstrual cycle, of which
VEGF was much less in uterine uid obtained during
implantation from patients with infertility.
11
Schild et
al. state that it is important to have good emdometrial
vascularization inorder to implant.
14
In the current study,
there was positive correlation between VEGF and E2 at
day of HCG administration (Table 3), which is supported
by authors who have noted a maximum expression of
VEGF in the stroma during proliferative phase, with a
peak glandular VEGF expression during the secretary
phase.
15
Additionally, in the present study, we noticed a clear
positive association between VEGF and endometrial
thickness at day of embryo transfer (Table 3), which
is also concluded in a study conducted by Miwa et al.,
which indicates a ‘thin’ endometrium characterised by
high blood ow impedance of the uterine radial artery,
poor epithelial growth, decreased VEGF expression, and
poor vascular development.
16
us, there is a positive
association between endometrial thickness and VEGF; the
angiogenic factor plays a role in endometrial receptivity,
and this elucidates higher endometrial thickness for
pregnant women compared to non-pregnant women,
as well as a signicant value association with Vs/Vd in
our study (Table 4).
In the current study, pregnant women had a thicker
endometrium than non-pregnant women. Researchers
found that there is a reduced success rate of IVF in
patients with a thin endometrium, even if there was no
previous intrauterine surgery or infection. Some studies
found that a thin EMT negatively aects pregnancy
rates following fertility treatment
17
, while other studies
were unable to conrm this, indicating that the use of
endometrial thickness as a tool for deciding on cycle
cancellation or the freezing of all embryos, or cessation
from further IVF treatment, appears unjustied.
18,19
Some groups did not nd a correlation between age,
the number of follicles, and gonadotropin ampoules with
endometrial thickness; however, in all age ranges, the chance
of pregnancy was higher with an endometrial thickness
of 6<ET≤10 mm.
20
Higher conception ratesoccured in
patients with endometrial thickness of 10 mm and above.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
37
Original Article
SMJ
TABLE 1. Comparison of demographic parameters and hormonal prole between pregnant and non-pregnant
ICSI patient groups by unpaired t-test.
Pregnant Non-pregnant
Parameters N = (36) N = (44) P-value
Mean ± SD Mean ± SD
Age (yrs) 28.89 ± 6.07 29.77 ± 5.97 0.52
BMI (kg/m
2
) 26.05 ± 2.85 27.31 ± 2.94 0.06
Duration of infertility (yrs) 5.64 ± 3.79 6.07 ± 3.95 0.62
FSH (mIU/ml) 7.59 ± 1.94 7.02 ± 2.80 0.29
LH (mIU/ml) 4.89 ± 2.06 5.31 ± 3.36 0.49
E
2
(pg/ml) 37.3 ± 13.6 34.3 ± 13.4 0.33
PRL (ng/ml) 15.87 ± 6.18 15.46 ± 7.03 0.78
TSH (mIU/ml) 2.275 ± 0.625 2.176 ± 0.920 0.57
Testosterone (ng/dl) 1.039 ± 0.682 1.50 ± 3.06 0.34
Abbreviations: n=number, SD=standard deviation, yrs=years, BMI=body mass index, FSH=follicle stimulating hormone, LH=luteinized
hormone, E
2
=estradiol hormone, PRL=prolactin hormone, TSH=thyroid stimulating hormone
TABLE 2. Comparison between pregnant and non-pregnant ICSI patient groups with ultrasound Doppler parameters
and VEGF level by unpaired t-test.
Pregnant Non-pregnant
Parameters N = (36) N = (44) P-value
Mean ± SD Mean ± SD
Endometrial thickness (mm) 9.72 ± 1.35 8.95 ± 1.21 0.0
RI 0.584 ± 0.124 0.651± 0.132 0.0
PI 0.829 ± 0.301 1.006± 0.335 0.02
V
s
/V
d
3.14 ± 3.75 3.82 ± 3.07 0.38
VEGF (pg/ml) 433 ± 207 276 ± 165 0.01
Abbreviations: n=number, SD=standard deviation, RI=resistance index, PI=pulsatility index, V
s
=peak systolic velocity, V
d
=diastolic
velocimetry, VEGF=vascular endothelial growth factor
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
38
Rahmatullah et al.
TABLE 3. Correlation between VEGF with E2 at day of HCG administration, and ultrasound Doppler parameters
at day of embryo transfer by Pearson correlation test.
E
2
at Day of HCG RI PI V
s
/V
d
Endometrial
administration(pg/ml) thickness(mm)
VEGF (pg/ml)
r 0.325 -0.168 -0.11 -0.088 0.23
P 0.003 0.137 0.331 0.436 0.04
Abbreviations: r=Pearson correlation, P=p-value, E
2
==estradiol hormone, HCG=human chorionic gonadotropin hormone, RI=resistance
index, PI=pulsatility index, V
s
=peak systolic velocity, V
d
=diastolic velocimetry, VEGF=vascular endothelial growth factor.
TABLE 4. Correlation between endometrial thickness with E2 at day of HCG administration, and ultrasound
Doppler parameters at day of embryo transfer by Pearson correlation test.
TABLE 5. Correlation between E
2
at day of HCG administration with ultrasound Doppler parameters at day of
embryo transfer by Pearson correlation test.
E
2
at day of HCG RI PI V
s
/V
d
administration(pg/ml)
Endometrial thickness
r 0.129 0.062 0.079 0.23
(mm)
P 0.253 0.586 0.489 0.04
Abbreviations: r=Pearson correlation, P=p-value, E
2
=estradiol hormone, HCG=human chorionic gonadotropin hormone, RI=resistance
index, PI=pulsatility index, V
s
=peak systolic velocity, V
d
=diastolic velocimetry
PI RI V
s
/V
d
E
2
at day of HCG
r -0.01 -0.031 -0.107
administration (pg/ml)
P 0.933 0.786 0.346
Abbreviations: r=Pearson correlation, P=p-value, E
2
=estradiol hormone, HCG=human chorionic gonadotropin hormone, RI=resistance
index, PI=pulsatility index, V
s
=peak systolic velocity, V
d
=diastolic velocimetry
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
39
Original Article
SMJ
Another study found that when the is an endometrium
of less than 7 mm it is better to do cryopreservation,
however if the endometrium is thin but with a good
texture (triple-line pattern), other factors should be
considered such as the quality of the embryo. is is in
agreement with the results of the current study.
21
e sub-endometrial RI and PI in the present
study were signicantly lower in pregnant patients than
in non-pregnant patients. is is supported by other
researchers, who found patients that became pregnant
were characterised by a signicantly lower resistance
index; these results were obtained from sub-endometrial
vessels by transvaginal colour Doppler ultrasonography.
22
e reason for this is because sub-endometrial vascularity
increased signicantly at day of embryo transfer, due
to the eect of stimulated hormones on endometrial
angiogenesis.
23
Jain et al. (2015) found that serum VEGF
levels rose alongside with an increase in Doppler vascular
penetration zones (zone 2, intermediate vascularity), which
implies that serum VEGF concentrations can be used as
a marker of endometrial receptivity. No conception was
noted in patients with poor or intermediate vascularity,
as identied in Doppler vascular zone 2.
is means that vascular endothelial growth factor
(VEGF) is a major regulator of endothelial cell proliferation,
angiogenesis, vasculogenesis, and capillary permeability.
A concurrent rise in serum VEGF level was observed
alongside an increase in Doppler vascular penetration
zones. A receptive endometrium is a reection of good
endometrial vascularity, which signies serum VEGF as
a marker of endometrial receptivity.
24,25
is is supported
by our results as shown in Table 2, where a high level of
VEGF is associated with low RI, PI, and VS/Vd, and an
elevated level of E2 at day of embryo transfer, as well as
a thick endometrium.
CONCLUSION
Combined analysis of endometrial receptivity was
conducted for the present research. e serum level of
VEGF, and sub-endometrial evaluation with 2D PD-
US, by measuring resistance index (RI) and pulsatility
index (PI) at day of embryo transfer, can serve as useful
prognostic methods for the detection of endometrial
receptivity and pregnancy outcome in infertile women
undergoing ICSI protocols, and will be helpful for candidate
counselling with regard to postponing embryo transfer
and cryopreservation, which may be a better option
for a future cycle, when better endometrial Doppler
parameters can be achieved.
ACKNOWLEDGMENTS
The authors would like to thank International
Islamic University Malaysia for funding publication of
this project under grant number PRIGS 18-03-0030.
REFERENCES
1. Nordqvist C. Infertility: Causes, Diagnosis, Risks, Treatments.
Medical News Today, 21 January 2016. Available from: www.
medicalnewstoday.com
2. National Institute for Health and Clinical Excellence. Fertility:
assessment and treatment for people with fertility problems.
NICE Clinical Guidelines, No. 11, Feb.2004.
3. Balen AH, Rutherford AJ. Management of infertility. BMJ
2007;335:608-11.
4. Padmanabhan RA, Laloraya M. Estrogen-Initiated Protein
Interactomes During Embryo Implantation. Am J Reprod
Immunol 2016;75:256-62.
5. Kim A, Han JE, Yoon TK, Lyu SW, Seok HH, Won HJ. Relationship
between endometrial and subendometrial blood ow measured
by three-dimensional power Doppler ultrasound and pregnancy
aer intrauterine insemination. Fertil Steril 2010;94:747-52.
6. Ivanovski M. e Role of Ultrasound in the Evaluation of
Endometrial Receptivity Following Assisted Reproductive
Treatments. In: Friendler S, ed. In Vitro Fertilization. IntechOpen
Ltd., London, UK. Published: April 11th, 2012. DOI: 10.5772/2323.
7. Naalin J. Ultrasound studies of the endometrial-myometrial
junction for the diagnosis of adenomyosis and endometrial
cancer. e Gynaecology Diagnostic and Outpatient Treatment
Unit, Department of Obstetrics and Gynaecology, Elizabeth
Garrett Anderson Wing, University College Hospital, Graon
Way, London. WC1H. esis submitted for the degree of
MD(Res), University College London, March 2014.
8. Zaidi J, Campbell S, Pittrof R, Tan SL. Endometrial thickness,
morphology, vascular penetration and velocimetry in predicting
implantation in an in vitro fertilization program. Ultrasound
Obstet Gynecol 1995;6:191-8.
9. Smith SK. Regulation of angiogenesis in the endometrium.
Trends Endocrinol Metab 2001;12:147-51.
10. Nardo LG. Vascular endothelial growth factor expression in
the endometrium during the menstrual cycle, implantation
window and early pregnancy. Curr Opin Obstet Gynecol 2005;
17:419-23.
11. Hannan NJ, Paiva P, Meehan KL, Rombauts LJ, Gardner DK,
Salamonsen LA. Analysis of fertility-related soluble mediators
in human uterine uid identies VEGF as a key regulator of
embryo implantation. Endocrinology 2011;152:4948-56.
12. Yang JH, Wu MY, Chen CD, Jiang MC, Ho HN, Yang YS.
Association of endometrial blood ow as determined by a
modied color doppler technique with subsequent outcome
of in-vitro fertilization. Hum Reprod 1999;14:1606-10.
13. El-Zenneni H, Moustafa R, Abdel-Hafeez M, El-Salally H,
Abdel-Kader A, Elnaggar A. Assessment of uterine, subendometrial
blood ows and endometrial gland vascular endothelial growth
factor (EG-VEGF) in women with unexplained infertility.
Middle East Fertility Society Journal 2015;20:119-26.
14. Schild RL, Holthaus S, d'Alquen J, Fimmers R, Dorn C, van Der
Ven H, Hansmann M. Quantitative assessment of subendometrial
blood ow by three-dimensional- ultrasound is an important
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
40
predictive factor in an in-vitro fertilization programme. Hum
Reprod 2000;15:89-94.
15. Nayak NR, Brenner RM. Vascular proliferation and vascular
endothelial growth factor expression in the rhesus macaque
endometrium. J Clin Endocrinol Metab 2002;87:1845-55.
16. Miwa I, Tamura H, Takasaki A, Yamagata Y, Shimamura K,
Sugino N. Pathophysiologic features of “thin” endometrium.
Fertil Steril 2009;91:998-1004.
17. El-Toukhy T, Coomarasamy A, Khairy M, Sunkara K, Seed P,
Khalaf Y, Braude P. e relationship between endometrial
thickness and outcome of medicated frozen embryo replacement
cycles. Fertil Steril 2008;89:832-9.
18. Schild RL, Knobloch C, Dorn C, Fimmers R, van der Ven H,
Hansmann M. Endometrial receptivity in an in vitro fertilization
program as assessed by spiral artery blood ow, endometrial
thickness, endometrial volume, and uterine artery blood ow.
Fertil Steril 2001;75:361-6.
19. Kasius A, Smit JG, Torrance HL, Eijkemans MJ, Mol BW,
Opmeer BC, Broekmans FJ. Endometrial thickness and pregnancy
rates aer IVF: a systematic review and meta-analysis. Hum
Reprod Update 2014;20:530-41.
20. Habibzadeh V, Nematolahi Mahani SN, Kamyab H. The
correlation of factors aecting the endometrial thickness with
pregnancy outcome in the IUI cycles. Iran J Reprod Med 2011;
9:41-6.
21. Matyas S, Boda K, Kaali SG. e eect of endometrial thickness
on IVF/ICSI outcome. Hum Reprod 2003;18:2337-41.
22. Zhao J, Zhang Q, Li Y. e eect of endometrial thickness
and pattern measured by ultrasonography on pregnancy
outcomes during IVF-ET cycles. Reprod Biol Endocrinol 2012
28;10:100.
23. Raine-Fenning NJ, Campbell BK, Kendall NR, Clewes JS,
Johnson IR. Endometrial and subendometrial perfusion are
impaired in women with unexplained subfertility. Hum Reprod
2004;19:2605-14.
24. El-Zenneni H, Moustafa R, Abdel-Hafeez M, El-Salally H,
Abdel-Kader A, Elnaggar A. Assessment of uterine, subendometrial
blood ows and endometrial gland vascular endothelial growth
factor (EG-VEGF) in women with unexplained infertility.
Middle East Fertility Society Journal 2015;20:119-26.
25. Jain P, Paul S, Gupta U, Tuli A, Jain M. Doppler Blood Flow
Studies of the Endometrium and Its Relation with Serum VEGF
in Women with Unexplained Infertility. IOSR Journal of
Dental and Medical Sciences (IOSR-JDMS) 2015;14:74-78.
Rahmatullah et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
41
Original Article
SMJ
Pajaree itthiwong, M.D.,* antida Koopitakkajorn, M.D.**
*Division of Pediatric Dermatology, **Department of Pediatrics, Phramongkutklao Hospital, Phramongkutklao College of Medicine, Bangkok, ailand.
TheGoodSkinCarePracticesandEmollientUse
since Early Infancy as the Primary Prevention of
InfantileAtopicDermatitisamongInfantsatRisk:
a Randomized Controlled Trial
ABSTRACT
Objective: e study aimed to determine whether enhancing the skin barrier since early infancy could reduce the
incidence of infantile atopic dermatitis among high risk infants.
Methods: We conducted a prospective, randomized, controlled trial at the Pediatric Clinic, Phramongkutklao
Hospital in Bangkok. Eligible infants aged less than 10 weeks with family history of atopy were enrolled and
randomly allocated to one of the two groups. e intervention group applied emollient at least once daily all over
the body together with receiving good skin care practice advice, whereas the control group received only good skin
care practice advice. All infants were followed up and assessed at 2, 4, 6 and 9 months old.
Results: Fiy-two infants were enrolled, 25 in the intervention group and 27 in the control group. At 9 months
old follow-up, none in the intervention group had infantile atopic dermatitis, whereas 14.8% in the control group
developed infantile atopic dermatitis (p-value < 0.05). e mean age at diagnosis of infantile atopic dermatitis was
5.5 months.
Conclusion: Regular emollient application together with good skin care practice since early infancy could reduce
the incidence of infantile atopic dermatitis among high risk infants.
Keywords: Infantile atopic dermatitis; primary prevention; emollient (Siriraj Med J 2020; 72: 41-46)
Corresponding author: Pajaree itthiwong
E-mail: pajareepmk@gmail.com
Received 15 May 2019 Revised 11 September 2019 Accepted 15 October 2019
ORCID ID: http://orcid.org/0000-0003-4236-450X
http://dx.doi.org/10.33192/Smj.2020.06
INTRODUCTION
Infantile atopic dermatitis (IAD) is chronic
inammatory dermatosis with increasing prevalence
worldwide. In 2009, the International Study of Asthma
and Allergies in Childhood (ISAAC) Phase ree provided
comprehensive data on the prevalence of atopic dermatitis
stressing the global concern of the increasing trends
both in developing and developed countries.
1
is study
proposed the prevalence of atopic dermatitis in the age
group 6 to 7 years ranging from 0.9% to 22.5% and 0.2%
to 24.6% for the group of 13 to 14 years old. Interestingly,
the high prevalence of 15% or more in the age group
6 to 7 years has been found in many countries across
the globe including the Asia-Pacic region. One study
conducted in Bangkok reported the prevalence of atopic
dermatitis at 9.4% in the older age group.
2
To our knowledge, factors responsible for developing
of IAD are multifactorial including genetic, environment,
immune dysregulation and dysfunctional skin barrier.
3
For the genetic aspect, when one or both of parents are
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
42
Thitthiwong et al.
diagnosed with any allergic disease, their children could
have a 40 to 50% chance to develop atopic dermatitis.
Infants who develop IAD before 2 years of age with
persisting disease would have the risk to develop other
atopic diseases as a part of the atopic march.
4-5
e
chronic relapsing itchy eczematous skin with typical
locations for particular age is known to be the hallmarks
of atopic dermatitis, aecting the quality of life not only
of the patients but also their family members.
6-7
Various modalities for the primary prevention of IAD
have been studied, for example, exclusive breastfeeding,
hydrolyzed protein formula, maternal antigen avoidance,
prebiotics and probiotics but without promising evidence.
8-13
e skin barrier dysfunction and immune dysregulation
are known as the key pathogenesis of IAD.
14-15
Hence,
the hypothesis of enhancing the skin barrier since birth
or early infancy to impede penetration of allergens and
consequently prevent inammatory cytokines release may
play an important role to prevent the disease. Simpson
et al. investigated the eect of emollient on the infant
skin barrier for the primary prevention of IAD among
high risk infants with positive outcomes.
16-17
Horimukai
et al. also studied the primary prevention of IAD among
high risk infants in Japan and reported reduced IAD
incidence among infants regularly using emollient.
18
We conducted a study to determine whether good
skin care practice together with regular emollient use
since early infancy could reduce the incidence of IAD
among high risk infants.
MATERIALS AND METHODS
e study protocol was approved by the Institutional
Review Board of the Royal ai Army Medical Department in
Bangkok (IRB RTA; Ro51h/58). A prospective, randomized,
controlled trial among healthy high risk infants for
whom at least one of their parents or siblings presented
any allergic diseases was conducted at the Pediatric
Outpatient Department of Phramongkutklao Hospital
in Bangkok from 2016 to 2017. Funding was granted by
Phramongkutklao Hospital, and investigators declared
they had no conict of interest. ailand’s clinical trial
registration (TCTR20161208001).
Study population
Eligible infants were defined as healthy, term
infants, aged less than 10 weeks old whose parent(s)
or sibling(s) had a history of any allergic disease such
as atopic dermatitis, asthma, allergic rhinitis, allergic
conjunctivitis, food allergy or other allergic conditions.
Infants known to have major congenital anomalies,
immunodeciency syndrome, any skin disease other
than infantile seborrheic dermatitis or neonatal acne
were excluded from the study. Infants whose parents
reported regular emollient use before enrollment were
also excluded. Written informed consent forms were
obtained from all enrolled families.
Study design and intervention
e enrolled infants were allocated to either the control
or intervention group by block of 4 randomization with
allocation concealment using opaque envelopes. Infants
in the intervention group were assigned to regularly
apply the hospital formulated emollient containing white
petrolatum, stearyl alcohol, propylene glycol and glycerin,
named “Cold Cream” all over the body except periorbital
and perioral areas at least once daily shortly within 3 to
5 minutes aer bathing and padding dry the baby skin.
However, those parents or caregivers of infants in the
control group were asked not to apply any skin care
products on the baby skin except using the gentle liquid
baby cleansers during bathing and the barrier ointment
or cream on diaper areas as needed. Caregivers in both
groups were given verbal advice for good skin care
practice repeatedly during every visit. e good skin care
practice comprised the proper duration of bathing of 5
to 10 minutes with tap or lukewarm water, bathing not
more than twice daily and using only a minimal amount
of gentle liquid baby cleansers of any manufactures. Bath
oil, bubble bath or any bath additives were not allowed
to be used in both groups.
Appointments with investigators were set up during
their regular well baby clinic visits at 2, 4, 6 and 9 months
old at the Pediatric Outpatient Clinic. One unblinded
investigator would perform a general physical examination,
growth and developmental milestone evaluation, general
health supervision, routine vaccination and advice regarding
good skin care practice to every caregiver of the infants
in both groups. e intervention group infants aged less
than 6 months received Cold Cream, 180 grams and the
infants aged 6 months or older received 240 grams of
Cold Cream per visit. e parents in the intervention
group were asked to bring back the empty bottles of Cold
Cream to conrm their compliance. During each visit,
the infants in both groups would be sent to another room
and receive only the skin examination and evaluation for
the diagnosis of infantile atopic dermatitis by a pediatric
dermatologist who was blinded as to group allocation.
Neither blood test nor equipment used for transepidermal
water loss or stratum corneum hydration measurements
were used in this study.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
43
Original Article
SMJ
Outcomes
e primary outcome was the cumulative incidence
of IAD in both groups. e diagnostic criteria for atopic
dermatitis was based on the atopic dermatitis guidelines
by Eicheneld et al. in 2014.
6
e secondary outcomes
were mean age of onset of IAD, adverse reaction of
Cold Cream application and the factors associated with
developing IAD. e study end points were dened when
the enrolled infants developed IAD or when the infants
were 9 months old.
Statistical analysis
e sample size needed to compare two proportions
was calculated between groups with an expected 60%
reduction in IAD incidence based on related studies
using α= 0.05 and β= 0.20.
17-18
A sample size of 70 infants
(35 in each group) was required.
Statistical analysis was performed by comparing
between the two groups by two-sample test proportion.
Descriptive data were presented as means (with standard
deviation), medians (with inter-quartile range), and
percentages.
RESULTS
is study was conducted from January 2016 to
April 2017. Initially, 70 eligible families were informed to
recruit in this study. Of these, 17 families were unwilling
to participate. One of investigators had to move to work
in another hospital, then due to time limitation, the
enrollment had to stop before reaching the calculated
sample size of 70 infants. en 53 infants were enrolled
and randomly allocated to one of the two groups. In all,
26 infants were placed in the intervention group and 27 in
the control group. One infant in the intervention group
had to leave the study before 9 months old because the
family had to move outside Bangkok. A diagram of the
participants is shown in Fig 1. e baseline demographic
data between groups were comparable (Table 1).
Regarding the primary and secondary outcomes,
none of the infants in the intervention group received a
diagnosis of IAD which was repeatedly evaluated during
each visit up to 9 months old, whereas 4 (14.8%) infants
in the control group developed IAD (p = 0.045; Table 2).
e mean age of the 4 infants at the onset of IAD was 5.5
± 0.55 months. e dryness of skin or xerosis, assessed
by skin examination during the last follow-up visits,
was comparable between groups (p = 0.120; Table 2).
However, the skin dryness was evaluated clinically without
transepidermal water loss or skin moisture measurements.
None of the 4 IAD infants developed cow’s milk protein
allergy or any other food allergy.
e exact volume of the Cold Cream used in each
infant in the intervention group was not recorded in
this study. Most of the parents carried only the empty
bottles of Cold Cream to present to the investigator to
conrm their compliance. However, the average Cold
Cream used monthly for infants less than 6 months was
90 gm and 120 gm for infants 6 months or older. No
interventions related to adverse events were reported
by the caregivers.
We also studied those factors assumed to have
eects on developing IAD including area of residence
(suburb or intown residence), feeding type (exclusive
or nonexclusive breast feeding), inhouse pets (dogs or
cats) and inhouse smoking and found no signicant
dierence between the IAD and non IAD groups (Table
3). e environmental data around the time of the study
in Bangkok are presented. e average temperature
in Bangkok was 31°C with 66% relative humidity. e
mean level of PM
2.5
(ne particulate matter) was 58.69
µg/m
3
at the open roadside areas.
19
DISCUSSION
This comprised a randomized controlled trial
conducting to demonstrate a modality for the primary
prevention of infantile atopic dermatitis among high risk
infants by enhancing the skin barrier with regular use
of emollient and educating concerning good skin care
practice to caregivers. We found that using Cold Cream or
a petrolatum containing moisturizer at least once daily to
the whole body of the high risk infants showed a decrease
in the cumulative incidence of infantile atopic dermatitis
compared with the infants in the control group receiving
only the good skin care advice. We designed to enroll
only term, healthy high risk infants, because manipulating
the preterm or sick infant skin might increase rates of
infection. Our results were in the same direction as in
the related data from the studies of Simpson EL in the
US and Horimukai K in Japan, although our study did
not show a strongly signicant dierence.
17-18
is might
have been a result of the small sample size in this study.
e related studies demonstrated that daily emollient
use among high risk infants signicantly decreased the
incidence of infantile atopic dermatitis.
Which emollients would be the best to use in this
setting? For example, in the hot and humid geographic
regions, we would not recommend using pure petrolatum
to the infant skin especially during the summer months.
Heat rash would be one concern. Instead we would
recommend using the oil-in-water emulsions for the
normal skin type and the water-in-oil emulsions such
as the “Cold Cream” for the dry skin type. In this study,
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
44
Thitthiwong et al.
Fig 1. Study population diagram.
TABLE 1. Demographic characteristics.
*Data presented as mean ± standard deviation. ** By dermatologic examination.
Characteristics
Intervention Control
(n = 25) (n = 27)
Sex, n (%)
Male 12 (48) 13 (48)
Female 13 (52) 14 (52)
Birthweight (g)* 2,270 ± 0.4 2,480 ± 0.7
Age at enrollment (weeks)* 4.19 ± 2.9 4.00 ± 2.9
Both parental atopy (%) 42.3 18.5
Skin at enrollment ** (%)
Moderate to marked skin dryness 20.0 7.4
Feeding (%)
Exclusive breast feeding 56.0 80.5
Non-exclusive breast feeding 44.0 19.5
Bathingfrequency(%)
Once daily 48.0 66.7
More than once daily 52.0 33.3
Bathing time (%)
Up to 10 minutes 92.0 96.3
More than 10 minutes 8.0 3.7
In house pets (%)
Pets 32.0 37.0
No pets 68.0 63.0
In house smoking (%)
Smoking 68.0 63.0
No smoking 32.0 37.0
Area of residences (%)
Urban areas 72.0 48.1
Suburb areas 28.0 51.9
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
45
Original Article
SMJ
we did not measure neither the stratum corneum hydration
nor the transepidermal water loss (TEWL). us, we
cannot conclude using scientic data to explain the
positive results in the emollient group. ese were our
limitations.
Scientists have unveiled many associated genes
linked to atopic dermatitis such as the Filaggrin gene
(FLG) for skin barrier dysfunction, β-defensin 1 (DEFB1)
for the susceptibility to infections, nucleotide-binding
oligomerization domain 1 (NOD1) and Toll-like receptor
2 (TLR2) genes for immune dysregulation.
15
ese provide
evidence stressing the importance of genetic background
for developing atopic dermatitis. In this study, we enrolled
high risk infants with any atopic problems among the
families, even though the infants from atopic dermatitis
families would have higher risk to develop IAD than
infants from families with asthma or allergic rhinitis.
20
We realized that some parents could not recall having
atopic dermatitis as a part of their atopic march during
their childhood.
Allergic sensitization can occur through an impaired
skin barrier leading to inammatory responses and
consequently developing atopic rashes. To close the critical
gateways for microbes and allergens is to strengthen or
to enhance the skin barrier. ese strategies of enhancing
the skin barrier as early as possible might constitute a
convenient and eective method to implement to high
risk infants to prevent infantile atopic dermatitis.
According to the ISAAC Phase ree survey, the
number of patients with IAD still dramatically and
continuously increases over time with higher prevalence
(>15%) in many parts of big cities around the world
including Asia-Pacic.
1
Because of the small sample
size in this study, we could not conrm any correlation
between environmental or biological factors such as
feeding type, area of residence, smoking or inhouse pets
and IAD incidence.
Atopic dermatitis is a public health problem for
children across the globe. On the other hand, IAD is one
of the noncommunicable diseases (NCDs) of childhood
as we claimed metabolic syndrome for adulthood, and
atopic dermatitis causes burdens in health care systems.
21
One study estimated the direct and indirect costs of
atopic dermatitis to total over 5 billion dollars annually
in the US.
22
e primary prevention of IAD remains a
major concern.
Limitations of this study included the short time
of follow-up visit, small number of participants, lack
TABLE 2. Cumulative incidence of IAD and moderate to marked skin dryness (xerosis).
TABLE 3. e environmental or biological factors and IAD developing group.
Outcomes Intervention Control P-value*
(n = 25) (n = 27)
Infantile atopic dermatitis 0 (0%) 4 (14.8%) 0.045
Xerosis** 3 (12.0%) 8 (29.6%) 0.120
*Two-sample test of proportion, p < 0.05 = statistical signicant.
**Evaluated by Pediatric dermatologist at the end of study.
Abbreviation: IAD = infantile atopic dermatitis
Factors n in group IAD P-value*
(N = 52) n (%)
In town residential areas 31 2 (6.5) 0.683
Exclusive breast feeding 36 3 (8.3) 0.791
In house pets (cats or dogs) 14 3 (21.4) 0.162
In house smoking 34 4 (11.8) 0.130
*Multivariate analysis
Abbreviation: IAD = infantile atopic dermatitis
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
46
of measurement of skin moisture or TEWL and lack of
sensitization tests for allergens. More evidence for the
primary prevention of IAD still needs further investigations.
CONCLUSION
is study presented evidence that regular emollient
use concomitant with good skin care practice since early
infancy could be an eective modality for the primary
prevention of infantile atopic dermatitis among high
risk infants.
What this study adds
Enhancing the skin barrier by regular emollient use
together with implementing good skin care practice for
caregivers of infants born in atopic families could serve
as a modality for the primary prevention of infantile
atopic dermatitis.
ACKNOWLEDGMENTS
e authors would like to thank all infants and their
families for participating in this study and dedicating in the
disease prevention project. We also thank Phramongkutklao
Hospital for granting funds. We feel grateful to our
colleagues in Pediatric out-patient clinic and also the
pharmacists in Phramongkutklao Hospital for their
supports to our works.
REFERENCES
1. Odhiambo JA, Williams HC, Clayton TO, Robertson CF,
Asher MI; ISAAC Phase ree Study Group. Global variations
in prevalence of eczema symptoms in children from ISAAC
Phase ree. J Allergy Clin Immunol 2009;124:1251-8.
2. Vichyanond P, Sunthornchart S, Singhirannusorn V, Ruangrat
S, Kaewsomboon S, Visitsunthorn N. Prevalence of asthma,
allergic rhinitis and eczema among university students in
Bangkok. Respir Med 2002;96:34-8.
3. Bieber T. Atopic dermatitis. Ann Dermatol 2010;22:125-37.
4. Carlsten C, Dimich-Ward H, Ferguson A, Watson W, Rousseau
R, Dybuncio A, et al. Atopic dermatitis in a high-risk cohort:
natural history, associated allergic outcomes, and risk factors.
Ann Allergy Asthma Immunol 2013;110:24-8.
5. Spergel JM. From atopic dermatitis to asthma: the atopic
march. Ann Allergy Asthma Immunol 2010;105:99-106.
6. Eicheneld LF, Tom WL, Chamlin SL, Feldman SR, Hanin
JM, Simpson EL, et al. Guidelines of care for the management
of atopic dermatitis: section 1. Diagnosis and assessment of
atopic dermatitis. J Am Acad Dermatol 2014;70:338-51.
7. Yaghmaie P, Koudelka CW, Simpson EL. Mental health
comorbidity in patients with atopic dermatitis. J Allergy Clin
Immunol 2013;131:428-33.
8. Foisy M, Boyle RJ, Chalmers JR, Simpson EL, Williams HC.
Overview of Reviews e prevention of eczema in infants and
children: an overview of Cochrane and non-Cochrane reviews.
Evid Based Child Health 2011;6:1322-39.
9. Fleischer DM, Spergel JM, Assa’ad AH, Pongracic JA. Primary
prevention of allergic disease through nutritional interventions.
J Allergy Clin Immunol Pract 2013;1:29-36.
10. Oszukowska M, Michalak I, Gutfreund K, Bienias W, Matych
M, Szewczyk A, et al. Role of primary and secondary prevention
in atopic dermatitis. Postepy Dermatol Alergol 2015;32:409-20.
11. Ludvigsson JF, Mostrom M, Ludvigsson J, Duchen K. Exclusive
breastfeeding and risk of atopic dermatitis in some 8300 infants.
Pediatr Allerg Immunol 2005;16:201-8.
12. Boyle RJ, Ierodiakonou D, Khan T, Chivinge J, Robinson
Z, Geoghegan N, et al. Hydrolysed formula and risk of allergic
or autoimmune disease: systematic review and meta-analysis.
BMJ 2016;352:i974.
13. Allen SJ, Jordan S, Storey M, ornton CA, Gravenor MB,
Garaiova I, et al. Probiotics in the prevention of eczema: a
randomized controlled trial. Arch Dis Child 2014;99:1014-9.
14. Boguniewicz M, Leung DY. Atopic dermatitis: a disease of altered
skin barrier and immune dysregulation. Immunol Rev
2011;242:233-46.
15. Leung DY. New insights into atopic dermatitis: role of skin
barrier and immune dysregulation. Allergol Int 2013;62:151-
61.
16. Simpson EL, Berry TM, Brown PA, Hanin JM. A pilot study
of emollient therapy for the primary prevention of atopic
dermatitis. J Am Acad Dermatol 2010;63:587-93.
17. Simpson EL, Chalmers JR, Hanin JM, omas KS, Cork MJ,
McLean WH, et al. Emollient enhancement of the skin barrier
from birth oers eective atopic dermatitis prevention. J
Allergy Clin Immunol 2014;134:818-23.
18. Horimukai K, Morita K, Narita M, Kondo M, Kitazawa H, Nozaki
M, et al. Application of moisturizer to neonates prevents
development of atopic dermatitis. J Allergy Clin Immunol 2014;
134:824-30.
19. Sahanavin N, Tantrakarnapa K, Prueksasit T. Ambient PM10
and PM2.5 concentrations at dierent high trac related street
congurations in Bangkok, ailand. Southeast Asian J Trop
Med Public Health 2016;47:528-34.
20. Dold S, Wjst M, von Mutius E, Reitmeir P, Stiepel E. Genetic
risk for asthma, allergic rhinitis, and atopic dermatitis. Arch
Dis Child 1992;67:1018-22.
21. Silverberg JI. Public Health Burden and Epidemiology of
Atopic Dermatitis. Dermatol Clin 2017;35:283-9.
22. Adamson AS. e Economics Burden of Atopic Dermatitis.
Adv Exp Med Biol 2017;1027:79-92.
Thitthiwong et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
47
Original Article
SMJ
Vutisiri Veerasarn, M.D.*, Nattapatch Janhom, M.D.**, Yaowalak Chansilpa, M.D.*, Nan Suntornpong, M.D.*,
Kullathorn ephamongkhol, M.D.*, Nantakan Apiwarodom, M.D.*, Janjira Petsuksiri, M.D.*, Pittaya Dankulchai,
M.D.*, Jiraporn Setakornnukul, M.D.*, Achiraya Teyateeti, M.D.*, Warissara Rongthong, M.D.*, Panid Chaysiri, M.D.*
*Department of Radiology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, **Department of Radiology, Rajavithi Hospital,
Bangkok 10400, ailand.
The Impact of Active Nutritional Support for Head
andNeckCancerPatientsReceivingConcurrent
Chemoradiotherapy
ABSTRACT
Objective: Malnutrition is the most common problem in head and neck cancer (HNC) patients receiving concurrent
chemoradiotherapy. e radiation toxicities cause decreased food intake, with resultant severe weight loss and
malnutrition. is study sought to determine whether an active nutrition improvement counseling program before
and during concurrent chemoradiotherapy for HNC patients could increase the treatment completion rate without
the interruptions caused by the side eects of chemoradiotherapy.
Methods: e ndings of a prospective study of the eects of an active nutrition improvement program before and
during concurrent chemoradiotherapy (study, n = 32) was compared with those of a retrospective chart review of
HNC patients who had received denite or postoperative concurrent chemoradiotherapy (control, n = 80). e
correlations between nutritional status and the number of treatment completions, number of tube feeding insertions
during treatment, RTOG toxicity, nutritional status, and quality of life were obtained.
Results: ere was no statistically signicant dierence between the concurrent chemoradiotherapy completion
rates of both groups (p = 0.121; 95% CI, 0.226-1.188). e major cause of delayed or discontinued chemotherapy
was oral mucositis. No signicant dierences were found in the tube feeding insertion rates and RTOG toxicities
of both groups. However, the data showed a clinically signicant dierence in the concurrent chemoradiotherapy
completion rate for the study group (56%), more than 15 percentage points higher than the control group’s rate (40%).
Conclusion: An active nutrition improvement program before and during concurrent chemoradiotherapy is
clinically benecial for HNC patients, providing a higher treatment completion rate than otherwise.
Keywords: Head and neck cancer; radiotherapy; chemotherapy; chemoradiotherapy; nutrition. (Siriraj Med J 2020;
72: 47-58)
Corresponding author: Vutisiri Veerasarn
E-mail: vutisiriv@hotmail.co
Received 12 June 2019 Revised 21 October 2019 Accepted 30 October 2019
ORCID ID: http://orcid.org/0000-0002-9601-994X
http://dx.doi.org/10.33192/Smj.2020.07
INTRODUCTION
Head and neck cancer (HNC) is one of the most
common malignancies in the world, with high mortality
rates in developing countries.
1
In 2016, Siriraj Cancer
Registry reported that HNC accounted for 6.9% of newly
diagnosed cancers.
2
Malnutrition as a comorbidity of
cancer has been recognized. A study by Unsai et al. that
evaluated the nutritional status of HNC patients receiving
radiotherapy found that a quarter of the patients had
malnutrition at presentation. Although the number with
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
48
malnutrition had increased by more than three quarters
by the end of the treatment, the patients subsequently
demonstrated nutritional improvement by their 3- to
6-month follow-ups.
3
Nutritional status is a modiable factor that impacts
on disease prognosis and treatment compliance. In the case
of HNC, Langius et al. found that weight loss both before
and during radiotherapy was an important prognostic
indicator for 5-year disease-specic survival.
4
Another
study found that less weight loss was experienced by
radiotherapy-receiving patients with HNC or gastrointestinal
cancer who had attended intensive nutrition counseling
sessions than those just given standard care
5
e aim of this study was to establish the impact
of the nutritional status of HNC patients on treatment
compliance, radiation toxicities, rate of any tube feeding,
and quality of life.
MATERIALS AND METHODS
is study was performed at the Division of Radiation
Oncology, Department of Radiology, Faculty of Medicine
Siriraj Hospital, Mahidol University, during 2016-2017.
A quasi-experimental study design was used to compare
patients receiving routine pretreatment counseling with
those participating in an active nutrition improvement
program conducted by a physician. e study was approved
by the Ethics Committee of the Siriraj Institutional
Review Board (Si 797/2016).
e candidates in this study were newly diagnosed
HNC patients (nasopharyngeal, oral cavity, oropharyngeal,
laryngeal, and hypopharyngeal cancers) aged 18-80
years with an Eastern Cooperative Oncology Group
(ECOG) performance status of 0-2. Those patients
treated with a curative aim received either prescribed
concurrent chemoradiotherapy or postoperative
concurrent chemoradiotherapy. We excluded patients
who had early stage (T1N0M0) or distant metastatic (M1)
diseases. Candidates were also excluded if they had other
malignancies; had received neoadjuvant chemotherapy
or previous radiotherapy; or had medical illnesses that
would compromise their ability to complete the study,
such as a systemic infection. Data were collected during
the chemoradiotherapy treatment and through to the
3-month follow-up.
Study population
Control group (retrospective study)
A retrospective chart review was conducted of
newly diagnosed HNC patients who had received denite
concurrent chemoradiotherapy or postoperative concurrent
chemoradiotherapy for curative intent January-December
2016. e information was collected from the records
relating to the patients’ rst visits to the Division of
Radiation Oncology, their entire treatment schedule,
and the follow-up conducted 3 months aer treatment
completion. ose patients were all scheduled for routine,
pretreatment counseling.
Study group (prospective study)
A prospective intervention was undertaken of newly
diagnosed HNC patients with a curative aim who had been
scheduled for denite concurrent chemoradiotherapy or
postoperative concurrent chemoradiotherapy January-
December 2017. At least one month before starting the
radiotherapy, the patients began to participate in an
active nutrition improvement program conducted by
their physician. e program sessions were held at the
date of enrollment, every 2 weeks before radiotherapy
started, and every week during their treatment. All patients
gave written, informed consent before commencing the
program.
Denitions and terms
Head and neck cancer (HNC) was one whose
primary site was the nasopharynx, oral cavity,
oropharynx, hypopharynx, or larynx. e tumor
staging was determined in accordance with the
American Joint Committee on Cancer (AJCC)
Cancer Staging Manual, Seventh Edition (2010).
e active nutrition improvement program for
the study group was the nutrition counseling
provided by the physician, comprising nutrition
recommendations, guidance on the calculation of
a proper calorie-intake (30-35 kcal/kg), and a
notebook for the patients’ daily dietary records.
A serial assessment was performed at the rst
visit; the rst, fourth, and last weeks of the radiation
treatment; and then at the rst, second, and third
months aer treatment completion. Two tools
were used for this purpose: the Patient-Generated
Subjective Global Assessment (PG-SGA), and
the Functional Assessment of Cancer erapy–
Head and Neck (FACT–H&N, version 4).
e radiotherapy techniques in this study were
either the three-dimensional conformal radiation
therapy or volumetric modulated arc therapy.
e concurrent chemotherapy regimen (Cisplatin
or Carboplatin) selected for each patient was
based on medical oncologists’ judgments.
Treatment completion was deemed to have
occurred if an HNC patient attended and nished
the denite concurrent chemoradiotherapy or
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
49
Original Article
SMJ
postoperative concurrent chemoradiotherapy
treatment sessions as scheduled, without any
interruptions or postponements.
e study hypothesis and objectives
It was hypothesized that the active nutrition
improvement program would lessen the incidences of
treatment interruption or postponement.
e primary objective was to study the impact of
nutritional status before and during the radiotherapy and
chemotherapy treatment on patients’ rates of treatment
completion, without any interruptions or postponements
resulting from treatment-related side eects.
e secondary objectives were to review the relationships
between the patients’ nutritional statuses before and
during treatment and the feeding tube insertion rate,
Radiation erapy Oncology Group (RTOG) radiation
toxicity grading, nutritional statuses (PG-SGA), and
quality of life (FACT–H&N, version 4).
Statistical analysis
From the Siriraj Hospital medical record reviewed
of the 45 patients with head and neck cancer, 17 (37%) of
the patients experienced an interruption in concurrent
chemoradiotherapy. We considered a 15% reduction in
treatment interruption, so, the sample size should be 62
evaluable patients per treatment group, the study had
an 80% power to detect an absolute dierence of 15% in
the treatment interruption rate, assuming a two-sided
test and an overall signicance level of 0.05.
For comparisons between the two groups, continuous
variables were analyzed using Mann-Whitney U tests,
whereas chi-squared or Fisher’s exact tests and univariate
analysis by logistic regression were used for categorical
variables. A p-value of < 0.05 was considered statistically
signicant. e statistical analyses were conducted using
IBM SPSS Statistics for Windows, version 21 (IBM Corp.,
Armonk, NY, USA).
RESULTS
A total of 112 HNC patients were included in
the study. Of those, 80 were in the control group (the
retrospective chart review of patients receiving treatment
January-December 2016), but only 32 were in the study
group (the prospective study of patients undergoing
treatment January-December 2017). e patients in
the study group were not reach the target but we had
desired to stop enrolled the patients because it was at
the end of the year 2017. e patients received denite
concurrent chemoradiotherapy or postoperative concurrent
chemoradiotherapy. Seventy-three patients in the control
group and 27 in the study group remained throughout
the study. In control group, 1 patient had refused the
radiotherapy during the treatment period and after
complete treatment, 4 patients had disease progression,
2 patients had lost to follow-up. In study group, 1 patient
had refused the radiotherapy during the treatment period
and 1 patient had died from hemorrhagic stroke, 2
patients had disease progression and 1 patient had lost
to follow-up.
For both groups, the demographic data and patient
characteristics were similar. e majority of patients
were male (77.7%), and the mean age was 56.11 ± 9.94
years. Almost half of the patients were diagnosed with
nasopharyngeal cancer (45.5%). Most of the patients had
lymph node involvement (84.9%), and half had stage IV
disease (55.4%). ese patients received radiotherapy via
the volumetric modulated arc therapy technique (83%),
concurrent with Cisplatin (88.4%). About 18% of the
patients had tube feeding before enrollment, and this was
the same proportion as those who experienced signicant
weight loss following their diagnosis with cancer. In
the case of the study group, none of the patients who
enrolled in the active nutrition improvement program
had severe malnutrition at the time of diagnosis (Tables
1 and 2).
A comparison of the two groups did not reveal
any statistically signicant dierence in their treatment
completion rates (p = 0.121; 95% CI, 0.226-1.188).
Nevertheless, the completion rate for concurrent
chemoradiotherapy was clinically signicant for the
study group (56%), being 16 percentage points better
than that of the control group (40%). Similarly, the
chemotherapy compliance of the study group (56.3%)
was 15 percentage points better than that of the control
group (41.3%). Most patients in both groups were able to
attend their radiotherapy sessions without interruption,
with 81.3% and 77.5% attendance rates for the patients
in the study and control groups, respectively (Table 3).
Even in the multivariate analysis, adjusted by dening
the prognostic factors for age ≤ 60 years versus > 60
years, nasopharyngeal cancer versus non-nasopharyngeal
cancer, and stages I–III disease versus stage IV disease,
the dierences showed no statistical signicance for
both groups (p = 0.118; 95% CI, 0.288-1.183; Table 4).
As to patients who had their treatment delayed,
25% of the control group patients experienced oral
mucositis, with half of those (55.3%) being prescribed
narcotic drugs. However, just 3.1% of the patients in
the study group experienced the same treatment side
eect, with only 22.2% of the ones aected being given
narcotic drugs (Table 5).
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
50
TABLE 1. Baseline patient characteristics: demographics.
Characteristics Control (n = 80) Study (n = 32) Total (n = 112) P-value
Sex
Male 59 (73.8%) 28 (87.5%) 87 (77.7%)
Female 21 (26.3%) 4 (12.5%) 25 (22.3%) 0.137
Age (mean ± SD) 56.11 (± 9.94) 55.88 (± 12.52) 56.04 (± 10.69) 0.916
Site
Nasopharynx 36 (45%) 15 (46.9%) 51 (45.5%)
Oropharynx 18 (22.5%) 4 (12.5%) 22 (19.6%)
Oral cavity 14 (17.5%) 3 (9.4%) 17 (15.2%)
Larynx 7 (8.8%) 5 (15.6%) 12 (10.7%)
Hypopharynx 5 (6.3%) 5 (15.6%) 10 (8.9%) 0.267
T stage
1 18 (22.5%) 4 (12.5%) 22 (19.6%)
2 23 (28.8%) 5 (15.6%) 28 (25%)
3 16 (20%) 11 (34.4%) 27 (24.1%)
4 23 (28.8%) 12 (37.5%) 35 (31.3%) 0.160
N stage
0 14 (17.5%) 3 (9.4%) 17 (15.2%)
1 14 (17.5%) 6 (18.8%) 20 (17.9%)
2 44 (55%) 19 (59.4%) 63 (56.3%)
3 8 (10%) 4 (12.5%) 12 (10.7%) 0.768
Stage
I 2 (2.5%) 0 (0%) 2 (1.8%)
II 7 (8.8%) 3 (9.4%) 10 (8.9%)
III 27 (33.8%) 11 (34.4%) 38 (43.8%)
IVA 35 (43.8%) 14 (43.8%) 49 (43.8%)
IVB 9 (11.3%) 4 (12.5%) 13 (11.6%) 1.000
During the treatment, about 30% of the patients in
both groups required enteral tube feeding. As shown in
Table 6, most of them preferred nasogastric (NG) feeding
(94.1%), but 5.9% elected to be fed via percutaneous
endoscopic gastrostomy (PEG).
Serial assessments of the patients’ nutritional statuses
were performed during the treatment and throughout
the study. Looking at the body-weight changes between
before-receiving the treatment and at the end of treatment,
patients suering more than 10% weight loss represented
34.2% and 16.1% of those in the control and study groups,
respectively. During the rst month following treatment,
weight was regained, but the gain was better in the study
group than in the control group. However, by 3 months,
weight gain had improved equally in both groups (Table 7).
e percentage of body-weight change was calculated
at each patient assessment to monitor the progression
of weight loss. We used the median percentage of the
body-weight change of each group as a comparable value
measurement, and the dynamic weight changes were
plotted on a line graph (Fig 1). Once the concurrent
chemoradiotherapy started, patients experienced gradual
weight losses throughout the study. ere was a statistically
signicant dierence 1 month aer treatment completion:
at that time, the study group patients were likely to have
less weight loss than those in the control group. However,
at the end of the study, there was no statistical dierence
in the weight changes of the two groups.
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
51
Original Article
SMJ
TABLE 2. Baseline patient characteristics: treatment and nutritional status.
Characteristics Control (n = 80) Study (n = 32) Total (n = 112) P-value
Radiationtechnique
3D
1
15 (18.8%) 4 (12.5%) 19 (17%)
VMAT
2
65 (81.3%) 28 (87.5%) 93 (83%) 0.580
Type of treatment
CCRT 62 (77.5%) 26 (81.3%) 88 (78.6%)
Postop CCRT 18 (22.5%) 6 (18.8%) 24 (21.4%) 0.801
Chemotherapy regimen
Cisplatin 71 (88.8%) 28 (87.5%) 99 (88.4%)
Carboplatin 9 (11.3%) 4 (12.5%) 13 (11.6%) 1.000
Tube feeding (before treatment)
No 65 (81.3%) 26 (81.3%) 91 (81.3%)
Yes 15 (18.8%) 6 (18.8%) 21 (18.8%) 1.000
NG
4
(n = 21) – 1 (14.3%) 1 (4.5%)
PEG
5
(n = 21) 15 (100%) 5 (83.3%) 20 (95.2%) 0.286
% Baseline body-weight loss
Nil 46 (57.5%) 17 (53.1%) 63 (56.3%)
< 5% 23 (28.8%) 9 (28.1%) 32 (28.6%)
5%-10% 9 (11.3%) 5 (15.6%) 14 (12.5%)
> 10% 2 (2.5%) 1 (3.1%) 3 (2.7%) 0.889
Nutritional status (PG-SGA
6
)
A – 19 (59.4%) 19 (59.4%)
B – 13 (40.6%) 13 (40.6%)
C – – –
QOL mean score (± SD) – 71.97 (± 9.74) –
Abbreviations: 3D = ree-dimensional conformal radiotherapy; VMAT = Volumetric modulated arc therapy; CCRT = Concurrent
chemoradiotherapy; NG = Nasogastric tube; PEG = Percutaneous endoscopic gastrostomy; PG-SGA = Patient-Generated Subjective Global
Assessment
Radiation oral mucositis and radiation dermatitis
were the common acute complications of the radiotherapy.
e descriptive data showed that in the control group,
those complications developed at the third week of
treatment, which was 1 week earlier than for the study
group (Figs 2 and 3).
With patients who had participated in the active
nutrition improvement program (the study group), serial
assessments using their nutritional status classied by
PG-SGA were performed at the rst visit; the rst, fourth,
and last week of treatment; and the rst, second, and
third month aer nishing the treatment. e patients’
nutritional statuses were divided into 3 classes: Class A, well-
nourished (scores 1-8); Class B, moderately malnourished
(scores 9-18); and Class C, severely malnourished (scores
> 18). At the rst visit, most of the patients had good
nutritional status, and none had severe malnutrition.
However, during the treatment, almost all of the patients
became malnourished. Fortunately, their nutritional
statuses improved during the three months aer the
completion of the treatment (Fig 4). In addition, the
quality of life by functional assessment of cancer therapy
(FACT-H&N, version 4) was performed. Although the
mean of the quality of life score reduced gradually, the
mean score fell by less than 10 points (Fig 5).
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
52
TABLE 3. Results of treatment.
Treatment Control Study Total P-value Odds 95% CI
Ratio*
Concurrent Chemoradiotherapy
Complete 32 (40%) 18 (56.3%) 50 (44.6%)
Incomplete 48 (60%) 14 (43.8%) 62 (55.4%) 0.121 0.519 0.226–1.188
Radiotherapy
Complete 76 (95%) 31 (96.9%) 107 (95.5%)
Incomplete 4 (5%) 1 (3.1%) 5 (4.5%) 0.515 0.484 0.54–4.312
Radiotherapy: interruption
No 62 (77.5%) 26 (81.3%) 91 (79.1%)
Yes 18 (22.5%) 6 (18.8%) 24 (21.4%) 0.663 0.795 0.283–2.229
≤7days 8(10%) 3(9.4%) 11(9.8%) 0.632 0.715 0.182–2.813
> 7 days 10 (12.5%) 3 (9.4%) 13 (11.6%) 0.813 0.8 0.126–5.092
No 63 (78.8%) 25 (78.1%) 88 (78.6%)
Yes 17 (21.3%) 7 (21.9%) 24 (21.4%) 0.942 1.038 0.384–2.806
Chemotherapy
Complete 33 (41.3%) 18 (56.3%) 51 (45.5%)
Incomplete 47 (58.8%) 14 (43.8%) 61 (54.5%) 0.152 0.546 0.239–1.250
Chemotherapy compliance
60%–100%** 73 (91.3%) 30 (93.8%) 103 (92%)
< 60% 7 (8.8%) 2 (6.3%) 9 (8%) 0.662 0.695 0.136–3.541
Compliance of Cisplatin
100% 27 (37.5%) 16 (57.1%) 43 (43%) 0.26 0.281 0.031–2.552
66% 39 (54.2%) 11 (39.9%) 50 (50%) 0.642 0.591 0.064–5.442
33% 6 (8.3%) 1 (3.6%) 7 (7%)
Compliance of Carboplatin
100% 6 (75%) 2 (50%) 8 (66.7%) 0.501 3 0.122–73.64
66%–83% 1 (12.5%) 1 (25%) 2 (16.7%) 1 1 0.020–50.39
< 50% 1 (12.5%) 1 (25%) 2 (16.7%)
*e reference group was the control group; **Cisplatin 2/3 cycles or Carboplatin > 4/6 cycles
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
53
Original Article
SMJ
TABLE 4. Results of treatment: multivariate analysis by logistic regression.
Treatment Control Study Total P-value Odds 95% CI
Ratio*
Concurrent Chemoradiotherapy
Complete 32 (40%) 18 (56.3%) 50 (44.6%)
Incomplete 48 (60%) 14 (43.8%) 62 (55.4%) 0.118 0.514 0.224–1.183
*e reference group was the control group; adjusted by age ≤ 60 versus > 60 years, nasopharyngeal versus non-nasopharyngeal, and stages
I, II, III versus stage IV
TABLE 5. Cause of treatment delays, and the treatments for oral mucositis.
Treatment Control Study Total P-value Odds 95% CI
Ratio*
No treatment delay 33 (41.3%) 18 (56.3%) 51 (45.5%)
Cause of the treatment delay
Oral mucositis 20 (25%) 1 (3.1%) 21 (18.8%)
Dermatitis 1 (1.3%) 1 (3.1%) 2 (1.8%)
Hematologic toxicity 2 (2.5%) 2 (6.3%) 4 (3.6%)
Renal toxicity 7 (8.8%) 1 (3.1%) 8 (7.1%)
Infection 4 (5%) – 4 (3.6%)
Body-weight loss 1 (1.3%) 1 (3.1%) 2 (1.8%)
N/A 12 (15%) 8 (25%) 20 (17.9%) 0.071
Oral mucositis treatment
Non-narcotic drug 17 (44.7%) 7 (77.8%) 24 (51.1%)
Narcotic drug 21 (55.3%) 2 (22.2%) 38 (48.9%) 0.091 0.231 0.042–1.262
*e reference group was the control group
TABLE 6. Tube feeding during treatment.
Results Control Study Total P-value Odds 95% CI
Ratio*
Tube feeding 25 (31.3%) 9 (28.1%) 34 (30.4%) 0.745 0.861 0.349–2.126
NG
1
23 (92%) 9 (100%) 32 (94.1%)
PEG
2
2 (8%) – 2 (5.9%) 1.000
*e reference group was the control group
Abbreviations: NG = Nasogastric tube; PEG = Percutaneous endoscopic gastrostomy
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
54
TABLE 7. Percentage of body-weight loss: during treatment and follow-up
Body-weight loss Control Study Total P-value
Week 1
N 80 32 112
Nil 41 (51.3%) 22 (66.8%) 63 (56.3%)
< 5% 25 (31.3%) 7 (21.9%) 32 (28.6%)
5%–10% 11 (13.8%) 2 (6.3%) 13 (11.6%)
> 10% 3 (3.8%) 1 (3.1%) 4 (3.6%) 0.42
Week 2
N 80 31 111
Nil 22 (27.5%) 9 (29%) 31 (27.9%)
< 5% 44 (55%) 20 (64.5%) 64 (57.7%)
5%–10% 13 (16.3%) 2 (6.5%) 15 (13.5%)
> 10% 1 (1.3%) – 1 (0.9%) 0.634
Week 3
N 79 31 110
Nil 8 (10.1%) 6 (19.4%) 14 (12.7%)
< 5% 46 (58.2%) 15 (48.4%) 61 (55.5%)
5%–10% 25 (31.6%) 9 (29%) 34 (30.9%)
> 10% – 1 (3.2%) 1 (0.9%) 0.143
Week 4
N 79 31 110
Nil 10 (12.7%) 5 (16.1%) 15 (13.6%)
< 5% 29 (36.7%) 13 (41.9%) 42 (38.2%)
5%–10% 36 (45.6%) 12 (38.7%) 48 (43.6%)
> 10% 4 (5.1%) 1 (3.2%) 5 (4.5%) 0.866
Week 5
N 77 31 108
Nil 5 (6.5%) 4 (12.9%) 9 (8.3%)
< 5% 27 (35.1%) 12 (38.7%) 39 (36.1%)
5%–10% 31 (40.3%) 13 (41.9%) 44 (40.7%)
> 10% 14 (18.2%) 2 (6.5%) 16 (14.8%) 0.330
End of treatment
N 79 31 110
Nil 7 (8.9%) 5 (16.1%) 12 (10.9%)
< 5% 16 (20.3%) 6 (19.4%) 22 (20%)
5%–10% 29 (36.7%) 15 (48.4%) 44 (40%)
> 10% 27 (34.2%) 5 (16.1%) 32 (29.1%) 0.170
1
st
FU (1 month after RT)
N 79 31 110
Nil 4 (5.1%) 6 (19.4%) 10 (9.2%)
< 5% 11 (14.1%) 10 (32.3%) 21 (19.3%)
5%–10% 30 (38.5%) 5 (16.1%) 35 (32.1%)
> 10% 33 (42.3%) 10 (32.3%) 43 (39.4%) 0.005
2
nd
FU (2 months after RT)
N 76 28 104
Nil 9 (11.8%) 6 (21.4%) 15 (14.4%)
< 5% 14 (18.4%) 7 (25%) 21 (20.2%)
5%–10% 16 (21.1%) 7 (25%) 23 (22.1%)
> 10% 37 (48.7%) 8 (28.6%) 45 (43.3%) 0.284
3
th
FU (3 months after RT)
N 73 27 100
Nil 11 (15.1%) 5 (18.5%) 16 (16%)
< 5% 12 (16.4%) 9 (33.3%) 21 (21%)
5%–10% 11 (15.1%) 2 (7.4%) 13 (13%)
> 10% 39 (53.4%) 11 (40.7%) 50 (50%) 0.226
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
55
Original Article
SMJ
Fig 1. Median percentages of the body-weight changes.
P-value from Mann-Whitney U test
Fig 2. Acute complications: grand 2 or grade 3 radiation oral mucositis.
Fig 3. Acute complications: grand 2 or grade 3 radiation dermatitis.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
56
DISCUSSION
Patients who have been diagnosed with cancer usually
experience weight loss and/or malnutrition, especially in
the case of those with HNC. White et al. identied that
malnutrition could aect cancer patients’ functional and
quality of life outcomes.
6
ey also established easy criteria
to identify severe malnutrition by using the percentage
of weight loss from baseline: > 5% within 1 month, >
7.5% within 3 months, > 10% within 6 months, and >
20% within 1 year. In another study, O’Neill and Shaha
assessed malnutrition by using either the percentage of
weight loss during the preceding 6 months (the dierence
between the usual weight and the actual, current weight)
or the gold standard method, PG-SGA, to monitor the
nutritional status of patients.
7
ey recommended that
malnutrition be corrected in order to reduce the morbidity
and mortality rates resulting from treatment eects. Several
other studies have found that HNC patients experience
a reduced calorie intake and a consequential weight loss
before and during treatment. e studies suggested that
Fig 4. Nutritional status by PG-SGA.
Fig 5. Mean score of quality of lfe by functional assessment of cancer therapy: mean score by FACT-H&N version 4.
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
57
Original Article
SMJ
early intervention (active dietary counseling, the addition
of nutrition supplements, or the use of a prophylactic
feeding tube) may improve patients’ nutritional statuses.
8-11
ere are a number of ways to manage the nutrition
of HNC patients. Many studies have explored eective
ways to improve nutrition before, during, and aer cancer
treatment, especially concurrent chemoradiotherapy.
Prophylactic enteral feeding using a PEG or NG tube,
and which was done with and without a nutrition
supplement before cancer treatment, demonstrated
no eects on weight loss, quality of life, or nutritional
improvement.
12–13
However, the prophylactic enteral
feeding revealed the benet of overall clinical outcome
improvements and a decline in the incidences of serious
treatment side eects. Prophylactic enteral tube feeding
was indicated in those patients who were predicted to
have severe malnutrition, were curative aim patients
with old age and locally advanced disease (T3/T4 and/
or lymph node involvement), and were undergoing
post-radical surgery with a reconstruction procedure
and concurrent chemoradiotherapy.
14-16
Pretreatment weight loss is an important prognostic
indicator for the overall survival of HNC patients. e
benet of individual dietary counseling is that it helps
maintain good nutrition in HNC patients. Isenring et al.
demonstrated that early and intense nutrition intervention
in the form of individual nutrition counseling and oral
supplements not only minimized weight loss and the
ensuing deterioration in nutritional status, quality of
life and physical function, but it also reduced the risks
of infection and treatment toxicities.
5
In this study, the primary outcome was to explore
the correlation between nutritional improvement and the
rate of treatment completion. ere were no signicant
dierences in the patient characteristics of the study and
control groups. Most of the patients were middle-aged
men who had been diagnosed with locally advanced
HNC and had received volumetric modulated arc therapy
concurrent with a cisplatin regimen. e active nutrition
improvement program with serial nutritional assessments
was introduced to the study group. ere was clinical
signicance in the higher rate of treatments that were
completed as scheduled by these study group patients.
e active nutritional counseling also provided a clinical
benet by way of improved compliance with chemotherapy.
e number of chemotherapy completions by the study
group, 56%, was 15 percentage points higher than the
corresponding gure for patients in the control group
(40%), who had only been given routine counseling. We
also examined whether there was a correlation between the
nutrition improvement program and serial assessments
and chemotherapy tolerance. To this end, the causes
of treatment interruptions and postponements were
studied. We found that common problems aer receiving
concurrent chemoradiotherapy were oral mucositis and
dermatitis, mainly occurring 3 to 4 weeks aer starting the
radiation treatment. In the control group, the mucositis
developed sooner and with more severe pain, indicated
by the greater number of patients using a narcotic drug
in that group than in the study group.
As to the secondary outcomes, only one-third of the
patients in each group required enteral feeding. NG-tube
feeding was mostly selected because it is less invasive than
PEG. Approximately the same proportion of patients in
each group had experienced a signicant weight loss by the
third week of treatment. is correlated with the onset of
mucositis and dermatitis, which were the common acute
complications. Unfortunately, nutritional assessments
using the PG-SGA classication were only conducted
for the patients in the study group. e descriptive data
showed that most of the patients were well-nourished at
their initial diagnosis. However, during their treatment,
they became moderately to severely malnourished. We
assume that was because of the natural course of the
disease and/or the acute complications of the treatment,
such as mucositis, which might have interfered with the
patients’ oral food intake. However, most of the patients
recovered to a normal, or mildly malnourished, status
within 3 months of treatment cessation.
CONCLUSION
The active nutrition improvement program,
a noninvasive procedure conducted by a physician,
was clinically beneficial. Compared to the patients
receiving only routine dietary counseling, the nutrition
improvement program produced a higher scheduled
treatment completion rate (i.e., without interruption or
postponement), improved chemotherapy tolerance, and
a lower and delayed incidence of mucositis. Given that
the active counseling program is not too complicated
and could be easily conducted by any paramedic, we
recommend that it be provided as part of treatment
protocols.
ACKNOWLEDGMENTS
We express our sincere appreciation to the physicians,
nurses, technicians, and support sta at the Division of
Radiation Oncology, Department of Radiology, Faculty
of Medicine Siriraj Hospital, for assisting with the data
collection and facilitating the study. We also thank all
patients who enrolled in this study.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
58
REFERENCES
1. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C,
Rebelo M, et al. Cancer incidence and mortality worldwide:
Sources, methods and major patterns in GLOBOCAN 2012.
Int J Cancer 2015;136:E359-86.
2. Chinswangwatanakul V, ed. Siriraj cancer registry 2016. Siriraj
Cancer Center, Faculty of Medicine Siriraj Hospital, Mahidol
University, ailand. ISSN 00857-3840.
3. Unsal D, Mentes B, Akmansu M, Uner A, Oguz M, Pak Y.
Evaluation of nutritional status in cancer patients receiving
radiotherapy: a prospective study. Am J Clin Oncol 2006;29:
183-8.
4. Langius JA, Bakker S, Rietveld DH, Kruizenga HM, Langendijk
JA, Weijs PJ, et al. Critical weight loss is a major prognostic
indicator for disease-specic survival in patients with head and
neck cancer receiving radiotherapy. Br J Cancer 2013;109:
1093-9.
5. Isenring EA, Capra S, Bauer JD. Nutrition intervention is
benecial in oncology outpatients receiving radiotherapy to the
gastrointestinal or head and neck area. Br J Cancer 2004;91:
447-52.
6. White JV, Guenter P, Jensen G, Malone A, Schoeld M. Consensus
statement: Academy of nutrition and dietetics and American
society for parenteral and enteral nutrition: Characteristics
recommended for the identication and documentation of adult
malnutrition (undernutrition). J Parenter Enteral Nutr 2012;
36:275-83.
7. O’Neill JP, Shaha AR. Nutrition management of patients with
malignancies of the head and neck. Surg Clin North Am 2011;
91:631-9.
8. Beaver MES, Matheny KE, Roberts DB, Myers JN. Predictors
of weight loss during radiation therapy. Otolaryngology-Head
and Neck Surgery 2001;125:645-8.
9. Kang WX, Li W, Huang SG, Dang Y, Gao H. Eects of nutritional
intervention in head and neck patients undergoing radiotherapy:
a prospective randomized clinical trial. Molecular and Clinical
Oncology 2016;5:179-82.
10. Ravasco P, Monteiro-Grillo I, Marques Vidal P, Camilo ME.
Impact of nutrition on outcome: a prospective randomized
controlled trial in patients with head and neck cancer undergoing
radiotherapy. Head Neck 2005;27:659-68.
11. Bahl M, Siu LL, Pond GR, Kim J, Tannock IF, Bayley A, et al.
Tolerability of the Intergroup 0099 (INT 0099) regimen in
locally advanced nasopharyngeal cancer with a focus on patients’
nutritional status. Int J Radiat Oncol Biol Phys 2004;60:1127-
36.
12. Pramyothin P, Manyanont S, Trakarnsanga A, Petsuksiri J,
Ithimakin S. A prospective study comparing prophylactic
gastrostomy to nutritional counselling with a therapeutic
feeding tube if required in head and neck cancer patients
undergoing chemoradiotherapy in ai real-world practice.
J Hum Nutr Diet 2016;29:768-76.
13. Brown TE, Banks MD, Hughes BGM, Lin CY, Kenny LM,
Bauer JD. Randomised controlled trial of early prophylactic
feeding vs standard care in patients with head and neck cancer.
BJC 2017;117:15-24.
14. Langius JA, Zandbergen MC, Eerenstein SE, van Tulder MW,
Leemans CR, Kramer MH, et al. Eect of nutritional interventions
on nutritional status, quality of life and mortality in patients with
head and neck cancer receiving (chemo)radiotherapy: a systematic
review. Clinical Nutrition 2013;32:671-8.
15. van der Linden NC, Kok A, Leermakers-Vermeer MJ, de Roos
NM, de Bree R, van Cruijsen H, et al. Indications for enteral
nutrition use and prophylactic percutaneous endoscopic
gastrostomy placement in patients with head and neck cancer
undergoing chemoradiotherapy. Nutr Clin Pract 2017;32:225-
32.
16. Cady J. Nutritional support during radiotherapy for head and
neck cancer: the role of prophylactic feeding tube placement.
Clinical Journal of Oncology Nurse 2007;11:875-80.
Veerasarn et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
59
Original Article
SMJ
Cholavech Chavasiri, M.D.*, Nopchanok Sukprasert, M.A.*, Srinual Chavasiri, M.D.**
*Department of Orthopaedic Surgery, **Department of Rehabilitation Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok,
ailand.
Depression and Quality of Life in Spinal Cord Injury
Patients Living in the Community After Hospital
Discharge
ABSTRACT
Objective: To investigate quality of life (QoL) and the prevalence of depression, and to identify factors signicantly
associated with QoL and depression in spinal cord injury (SCI) patients living in the community aer hospital
discharge.
Methods: is prospective study included SCI patients that have a follow-up evaluation and care at the Siriraj Spinal
Unit of the Department of Orthopaedic Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok,
ailand during April 2015 to February 2018. Presence and level of depression and QoL were assessed using Patient
Health Questionnaire-9 (PHQ-9) and World Health Organization Quality of Life Brief-ai (WHOQOL-BREF-
THAI), respectively. Age, gender, education level, income, marital status, level of impairment, injury severity, cause
of injury, and time since injury were collected and recorded.
Results: One hundred and twelve spinal injury patients (64.3% male) with a mean age of 44.3±15.3 years were
enrolled. e prevalence of depression was 39.3%, and the mean overall QoL was a moderate 90.3±14.7. Multivariate
analysis revealed marital status to be the only independent predictor of depression aer hospital discharge (odds
ratio [OR]: 2.99, 95% condence interval [CI]: 1.19-7.51; p=0.020). Regarding QoL, multivariate analysis revealed
educational level (OR: 16.18, 95% CI: 3.01-87.03; p=0.001), level of impairment (OR: 9.20, 95% CI: 1.84-46.13;
p=0.007), and depression (OR: 50.39, 95% CI: 7.94-319.83; p<0.001) to be independent predictors of quality of life.
Conclusion: Depression was observed in 39.3% of SCI, and most study patients had moderate QoL. Marital status
predicts depression; and, educational level, level of impairment, and presence of depression predict QoL.
Keywords: Quality of life; depression; spinal cord injury; community; postdischarge (Siriraj Med J 2020; 72: 59-66)
Corresponding author: Srinual Chavasiri
E-mail: csiccv@gmail.com
Received 9 April 2019 Revised 18 July 2019 Accepted 8 August 2019
ORCID ID: http://orcid.org/0000-0002-4439-4975
http://dx.doi.org/10.33192/Smj.2020.08
INTRODUCTION
Depression is a psychological complication that
is regularly observed among spinal cord injury (SCI)
patients.
1
Early onset depression may be found aer SCI
and is considered a normal part of the patient adaptation
process
2
, Prevalence of depressive symptoms in SCI in
in-patient ward (clinic) was report about 30%.
3
Moreover,
almost 20% of patients were found to have a continuation
of depressive symptoms aer hospital discharge
4
, and
that percentage increased to 25-30% aer these patients
returned to live in a community setting.
5
e prevalence
of depression was reported to be three times higher
among SCI patients living in the community than among
general population.
6
One of the negative consequences of depression
in SCI patients is reduced quality of life.
7
Craig, et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
60
reported a decrease in quality of life and an increase in
anxiety and/or depression in patients who sustained a
spinal cord injury.
3
Several studies found depression to be
associated with reduced quality of life in SCI patients.
8,9,10,11
Holicky and Charlifue studied 225 English SCI patiens
who had lived with SCI for 26 or more years. ey found
a decrease in the level of depression among patients
who got married. ey also found a higher level of life
satisfaction, overall mental health, and quality of life in
married SCI compared to unmarried SCI.
12
Depression in SCI patient dose not spontaneous
resolved, therefore continuation of treatment is necessary
for SCI patients aer hospital discharge.
13
Accordingly,
follow-up to monitor for depressive symptoms among SCI
patiens living in the community is essential. Moreover,
SCI patients are now living longer
14
, which highlights
the need for long-term support to improves in patient
quality of life. Improved understanding about depression
and quality of life in this patient population will improve
the mental health evaluation and treatment processes
among SCI patiens living in a community setting. e
aim of this study was to investigate quality of life and
the prevalence of depression, and to identify factors
signicantly associated with quality of life and depression
in SCI patients living in the community aer hospital
discharge.
MATERIALS AND METHODS
Participants
e study population consisted of SCI patients
who received follow-up evaluation and care at the
Siriraj Spinal Unit of the Department of Orthopaedic
Surgery, Faculty of Medicine Siriraj Hospital, Mahidol
University, Bangkok, ailand during the April 2015
to February 2018. Patients that met all of the following
criteria were included: (1) age greater than 18 years; (2)
having good consciousness; and, (3) having good ability
and a willingness to answer the PHQ-9 questionnaire
and the WHOQOL-BREF-THAI. Patients that have a
history or diagnosis of depression before SCI occurred
and active psychosis patients were excluded.
Assessment instrument
e PHQ-9 questionnaire is a self-report depression
screening tool that consists of 9 items that are derived
from DSM-IV depression diagnostic criteria. ere are
four levels of scoring for each item that include: “Not
at all” (score = 0); “Some days, not oen” (score = 1);
“Quite regularly” (score = 2); and, “Nearly every day”
(score = 3). e overall score can range from 0-27. Based
on studies conducted in ailand, a participant with a
score of 9 or more was diagnosed as having depression
(sensitivity = 0.84, specicity = 0.77).
15
In the present study,
a participant with a score higher than 7 was considered
to have depressive symptoms.
World Health Organization Quality of Life Brief-
ai (WHOQOL-BREF-THAI) is a 5-point rating scale
instrument. It is comprised of 26 items, of which 23
questions are positive, and 3 questions are negative. e
positive questions are scored, as follows: “Not at all”
(score = 1), “Slightly” (score = 2), “Moderate” (score =
3), “Very” (score = 4), and “Extremely” (score = 5). e
3 negative questions are scored, as follows: “Not at all”
(score = 5), “Slightly” (score = 4), “Moderate” (score =
3), “Very” (score = 2), and “Extremely” (score = 1). e
instrument measures four domains of quality of life,
including: physical health, psychological well-being,
social relationships, and environment. Quality of life
is classied by score into one of the three following
categories, including: poor quality of life (score 26-60),
moderate quality of life (score 61-95), or good quality
of life (score 96-130). Its reliability (Cronbach’s alpha
coecient) was 0.84, and the validity was 0.65 when
tested against the ai version of the WHOQOL-100.
is questionnaire is ocially approved by WHO.
16
Moreover, the WHOQOL-BREF-THAI is widely used
in the study of SCI patients. us, it is considered a
suitable tool for assessing the QoL of patients who have
sustained a spinal cord injury.
17
Aer receiving study approval from the Siriraj
Institutional Review Board (SIRB), Faculty of Medicine
Siriraj Hospital, Mahidol University, Bangkok, ailand
(Si 233/2015), written informed consent was obtained
from all enrolled study participants. The following
data were collected from patient medical records: age,
gender, education level, income, marital status, level of
impairment, injury severity, cause of injury, and time
since injury. Patient level of depression and quality of
life was assessed using the PHQ-9 questionnaire and
WHOQOL-BREF-THAI, respectively.
Statistical analysis
e prevalence of depression among spinal injury
patients was reported as number and percentage, and
the depression and quality of life scores were reported
as mean ± standard deviation (SD). Student’s t-test was
used to compare continuous data, and chi-square test
or Fisher’s exact test was used to compare categorical
data. Factors with a p-value less than 0.20 in univariate
analysis were included in multivariate analysis. e results
of multivariate analysis are presented as adjusted odds
ratio (OR) and 95% condence interval (CI). All data
Chavasiri et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
61
Original Article
SMJ
analyses were performed using SPSS Statistics version 18
(SPSS, Inc., Chicago, IL, USA), and a p-value<0.05 was
regarded as being an indicator of statistical signicance.
RESULTS
One hundred and twelve spinal injury patients
with a mean age of 44.3±15.3 years (range: 18-84)
were enrolled. Of those, 64.3% were male and 35.7%
were female. One-quarter (25.9%) of patients had an
elementary-level education, 40.2% were married, and
the average income was 11,810.9±18,197.7 ai baht per
month (range: 0-100,000). Regarding the cause of injury
and outcome, 57.1% of patients were injured in trac
accidents and 37.1% became paraplegic. e majority of
patients (41.1%) had incomplete injury, and the mean
duration of disease aer injury was 55.8±64.4 months
(range: 1 month to 28 years), as shown in Table 1.
e prevalence of depression was 39.3%, and the
mean depression score was 5.8±4.4. Regarding quality of
life, the majority of patients (59.8%) had a moderate level
of quality of life, followed by good quality of life (38.4%)
and poor quality of life (1.8%). e mean quality of life
score was 90.3±14.7. e mean quality of life score in
each domain was, as follows: physical domain: 23.4±4.9,
psychological domain: 22.2±3.8, social relationship domain:
9.8±2.2, and environmental domain: 28.2±4.9. e quality
of life in each domain was observed to be moderate
(Table 2).
Univariate analysis that included age, gender,
educational level, marital status, income, cause of injury,
time since injury, severity of impairment, and severity of
injury as factors potentially associated with depression
revealed only marital status to be signicantly associated
with depression (p=0.026). We then included all factors
from univariate analysis with a p-value less than 0.20 in
multivariate analysis (i.e., gender, marital status, and
time since injury). at analysis revealed marital status
to be the only independent predictor of depression aer
hospital discharge. More specically, patients who were
married had a 2.99 times (95% CI: 1.19-7.51, p=0.020)
greater probablity of developing depression than patients
who were single (Table 3).
Univariate analysis revealed educational level
(p=0.001), income level (p=0.019), level of impairment
(p=0.004), severity of injury (p=0.005), and depression
(p<0.001) to be signicantly associated with quality of
life. Multivariate analysis that included factors from
univariate analysis with a p-value less than 0.20 (i.e.,
educational level, income level, time since injury, depression,
severity of injury, and level of impairment) revealed
educational level, level of impairment, and depression
to be independent predictors of quality of life. Patients
with an undergraduate degree or higher had a 16.18
times (95% CI: 3.01-87.03, p=0.001) greater probability
of having good quality of life when compared to patients
with a lower level of education. Patients with non-severe
disability had a 9.20 times (95% CI: 1.84-46.13, p=0.007)
greater probability of having good quality of life when
compared to patient with tetraplegia. Lastly, patients with
an absence of depressive symptoms had a 50.39 times
(95% CI: 7.94-319.83, p<0.001) greater probability of
having good quality of life when compared with patients
with depression (Table 4).
DISCUSSION
e present study found the prevalence of depression
among SCI patients aer discharge from Siriraj Hospital
to be 39.3%. e prevalence of depression reported in this
study is far higher than the 18%
4
, 20%
18
, and 22% rates
reported from other countries.
19
However and importanly,
a score of 10 or greater reected the presence of depressive
symptoms in those studies, while a score greater than
7 reected the presence of depressive symptoms in our
study. Moreover, we used diagnostic criteria established
and recommended by the ailand Department of Mental
Health. Concerning the measurement of quality of life
among SCI patients aer discharge from the hospital, we
selected the WHOQOL-BREF Questionnaire as a tool
to demonstrated level of QOL, because it is widely used
in the study of SCI patients and it has proven acceptable
reliability.
20
e results of our study showed the overall
QoL and the QoL for each domain to be at moderate
level, which is similar to the results from previous studies
conducted in ai SCI population.
17,21,22,23,26
Considering factors related to depression among
SCI patients, the current study found marital status to be
signicantly associated with depression. More specically,
we found that married patients were more likely to
develop depression than unmarried SCI. Similarly, Tzanos,
et al.
25
investigated depressive mood among SCI patients
in Greece, and they found that having a spouse did not
play an important role in the prevention of depression.
On the contrary, In some previous study found marriage
to be signicantly associated with a low level of emotional
distress.
26
In some instances, SCI patients experience an
adverse change in marital status – a divorce. A previous
study found that divorced SCI patients were more likely
to have depressive mood when compared with other
groups of SCI patients.
27
Other previous ndings revealed
family relationship and family support are represent key
factors that inuence a reduction in depression among
SCI patients.
28
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
62
TABLE 1. Demographic and clinical characteristics of included spinal injury patients.
Variables Mean ± SD or n (%)
Age (years) 44.3±15.3
Male gender 72 (64.3%)
Education
No education 3 (2.7%)
Primary school 29 (25.9%)
Secondary school 27 (24.1%)
Vocationalcerticate 20(17.9%)
Undergraduate 27 (24.1%)
Postgraduate 6 (5.4%)
Marital status
Single 41 (36.6%)
Married 51 (45.5%)
Divorced/Separated/Widowed 20 (17.9%)
Income (Thai baht/month) 11,810.9±18,197.7
Cause of injury
Trafcaccident 64(57.1%)
Fall 36 (32.1%)
Violence 7 (6.3%)
Other 5 (4.5%)
Level of impairment
No 37 (33.0%)
Paraplegia 43 (38.4%)
Tetraplegia 32 (28.6%)
Severity of injury (n=75)
Complete 29 (25.9%)
Incomplete 46 (41.1%)
Time since injury (months) 55.8±64.4
Abbreviation: SD = standard deviation
TABLE 2. Quality of life scores of included spinal injury patients.
Domains Level of QOL QOL score
Poor (%) Moderate (%) Good (%) Mean±SD Range Level
Physical health 31 (27.7%) 73 (65.2%) 8 (7.1%) 23.4±4.9 11-34 Moderate
Psychological 50 (44.6%) 60 (53.6%) 2 (1.8%) 22.2±3.8 12-30 Moderate
Social relationship 25 (22.3%) 67 (59.8%) 20 (17.9%) 9.8±2.2 4-15 Moderate
Environmental 48 (42.9%) 63 (56.3%) 1 (0.9%) 28.2±4.9 18-40 Moderate
Overall QOL 43 (38.4%) 67 (59.8%) 2 (1.8%) 90.0±14.7 52-126 Moderate
Abbreviations: QOL= quality of life; SD = standard deviation
Chavasiri et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
63
Original Article
SMJ
TABLE 3. Univariate and multivariate analysis for factors signicantly associated with depression.
Factors
No depression Depression
P-value
Adjusted odds
P-value
(n=68) (n=44) ratio (95% CI)
Gender 0.160 0.301
Male 40 (55.6%) 32 (44.4%) 1.59 (0.66-3.81)
Female 28 (70.0%) 12 (30.0%) 1.00
Education 0.832
Lower than bachelor’s degree 47 (59.5%) 32 (40.5%)
Bachelor’s degree or higher 21 (63.6%) 12 (36.4%)
Marital status 0.026
Single 30 (73.2%) 11 (26.8%) 1.00
Married 24 (47.1%) 27 (52.9%) 2.99 (1.19-7.51) 0.020
Divorced/separated/widow 14 (70.0%) 6 (30.0%) 1.00 (0.29-3.40) 0.997
Income level (Thai baht/month) 0.377
No income 21 (55.3%) 17 (44.7%)
<15,000 21 (56.8%) 16 (43.2%)
≥15,000 26(70.3%) 11(29.7%)
Cause of injury 0.209
Fall 40 (62.5%) 24 (37.5%)
Trafcaccident 24(66.7%) 12(33.3%)
Violence 2 (28.6%) 5 (71.4%)
Other 2 (40.0%) 3 (60.0%)
Level of impairment 0.233
Paraplegia 26 (60.5%) 17 (39.5%)
Tetraplegia 16 (50.0%) 16 (50.0%)
Normal 26 (70.3%) 11 (29.7%)
Severity of injury 0.235
Complete 18 (62.1%) 11 (37.9%)
Incomplete 24 (52.2%) 22 (47.8%)
Normal 26 (70.3%) 11 (29.7%)
Time since injury (months) 0.137
<60 48 (57.8%) 35 (42.2%) 1.00
60-120 6 (50.0%) 6 (50.0%) 2.81 (0.71-11.12) 0.141
˃120 14(82.4%) 3(17.6%) 4.89(0.81-29.41) 0.083
Factors with a p-value <0.2 in univariate analysis were included in multivariate analysis
A p-value<0.05 in multivariate analysis indicates statistical signicance
Abbreviation: CI = condence interval
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
64
TABLE 4. Univariate and multivariate analysis for factors signicantly associated with quality of life.
Factors
Good Poor/moderate
P-value
Adjusted odds
P-value
(n=43) (n=69) ratio (95% CI)
Gender 0.547
Male 26 (36.1%) 46 (63.9%)
Female 17 (42.5%) 23 (57.5%)
Education 0.001
Lower than bachelor’s degree 22 (27.8%) 57 (72.2%) 1.00
Bachelor’s degree or higher 21 (63.6%) 12 (35.4%) 16.18 (3.01-87.03) 0.001
Marital status 0.560
Single 18 (43.9%) 23 (56.1%)
Married 19 (37.3%) 32 (62.7%)
Divorced/separated/widow 6 (30.0%) 14 (70.0%)
Income level (Thai baht/month) 0.019
No income 11 (28.9%) 27 (71.1%) 0.79 (0.18-3.54) 0.756
<15,000 11 (29.7%) 26 (70.3%) 0.84 (0.19-3.67) 0.817
≥15,000 21(56.8%) 16(43.2%) 1.00
Cause of injury 0.379
Fall 28 (43.8%) 36 (56.3%)
Trafcaccident 13(36.1%) 23(63.9%)
Violence 1 (14.3%) 6 (85.7%)
Other 1 (20.0%) 4 (80.0%)
Level of impairment 0.004
Tetraplegia 7 (21.9%) 25 (78.1%) 1.00
Paraplegia 14 (32.6%) 29 (67.4%) 2.97 (0.81-10.94) 0.102
Normal 22 (59.5%) 15 (40.5%) 9.20 (1.84-46.13) 0.007
Severity of injury 0.005
Complete 9 (31.0%) 20 (69.0%)
Incomplete 12 (26.1%) 34 (73.9%)
Normal 22 (59.5%) 15 (40.5%)
Time since injury (months) 0.142
<60 32 (38.6%) 51 (61.4%) 0.83 (0.18-3.89) 0.813
60-120 2 (16.7%) 10 (83.3%) 0.21 (0.02-2.06) 0.181
˃120 9(52.9%) 8(47.1%) 1.00
PHQ-9 <0.001
Depression 3 (6.8%) 41 (93.2%) 1.00
No depression 40 (58.8%) 28 (41.2%) 50.39 (7.94-319.83) <0.001
Factors with a p-value <0.2 in univariate analysis were included in multivariate analysis
A p-value<0.05 in multivariate analysis indicates statistical signicance
Abbreviations: CI = condence interval; PHQ-9 = Patient Health Questionnaire-9
Chavasiri et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
65
Original Article
SMJ
Concerning factors related to the QoL of SCI patients,
the present study found having a bachelor degree or higher,
having non-severe disability, or having no depressive
symptoms to be signicantly associated with patients
having a good QoL score. Similarly, previous studies
conducted in other countries and in ailand found the
factors that inuence QoL are the educational level
24
,
severity of disability
21,30,31
, and presence of depression.
Tzanos, et al.
25,29
found that high educational level enabled
SCI patients to acquire skills necessary for learning new
knowledge, and to better understand the steps required
for using assisting technology. Furthermore, higher
education level promoted greater access to technology,
and influenced a tendency toward higher income.
Employment and sucient income were factors that helped
to facilitate a better QoL. In contrast, low educational
level may eectuate or worsen social barriers among SCI
patients, and this may further limit all varieties of potential
opportunities that could improve their state of mind
and overall well-being. Regarding severity of disability,
people who become disabled due to an injury are more
likely to have a lower level of quality of life than people
in the general population.
21
Aer sustaining an injury,
the physical performance of patients decreases, and this
adversely aects their coping skills, which consequently
decreases their quality of life their quality of life.
23
An
absence of depressive symptoms was reported to be a
factor signicantly associated with good quality of life.
26
Tzanos, et al.
25
found the PHQ-9 score to be negatively
related with the QoL score in every domain (WHOQOL-
BREF). Similarly, Tate, et al.
32
also found good quality
of life to be signicantly related to reduced prevalence
of depression.
e limitation of this study is its cross-sectional
design, which means that the data was collected at one
point in time. However, the study participants selected the
answers that they felt best described their situation and
state of mind, so further elicitation to extract additional
information was not required.
CONCLUSION
e prevalence of depression among study participants
was 39.3% according to PHQ-9. Primary care, based on the
depression diagnostic criteria of the ailand Department
of Mental Health, includes managing emotions and
consulting a psychiatrist, psychologist or some other type
of training mental health professional. Moreover, self-
assessment must be elicited when a patient is receiving
follow-up services. Having an undergraduate degree or
higher, presence of non-severe disability, and no presence
of depressive symptoms were found to be signicantly
related with good quality of life among SCI. Moreover,
the educational level and severity of disability variables
were personal and illness-related factors that cannot
be easily prevented or treated among these patients.
However, depression, which is a factor that was found to
be associated with good quality of life, can be screened for
and prevented prior to its development. If depression is
identied, it can be treated, and this will lead to improved
QoL in patients with SCI that return to living in the
community aer their discharge from the hospital.
ACKNOWLEDGMENTS
e authors gratefully acknowledge the patients
that generously agreed to participate in this study,
and Ms. Julaporn Pooliam of the Division of Clinical
Epidemiology, Research Department, Faculty of Medicine
Siriraj Hospital, Mahidol University for assistance with
statistical analysis.
Conict of interest declaration: e authors hereby
declare no personal or professional conicts of interest
regarding any aspect of this study.
Funding disclosure: is was an unfunded study.
REFERENCES
1. Wong AWK, Heinemann AW, Wilson CS, Neumann H, Fann
JR, Tate DG, et al. Predictors of participation enfranchisement
aer spinal cord injury: the mediating role of depression and
moderating role of demographic and injury characteristics.
Arch Phys Med Rehabil 2014;95:1106-13.
2. Fullerton DT, Harvey RF, Klein MH, Howell T. Psychiatric
disorders in patients with spinal cord injuries. Arch Gen
Psychiatry 1981;38:1369-71.
3. Craig A, Tran Y, Middleton J. Psychological morbidity and
spinal cord injury: a systematic review. Spinal Cord 2009;47:
108-14.
4. Homan JM, Bombardier CH, Graves DE, Kalpakjian CZ,
Krause JS. A longitudinal study of depression from 1 to 5 years
aer spinal cord injury. Arch Phys Med Rehabil 2011;92:411-8.
5. Krause JS, Kemp B, Coker J. Depression aer spinal cord injury:
relation to gender, ethnicity, aging, and socioeconomic indicators.
Arch Phys Med Rehabil 2000;81(8):1099-109.
6. Bombardier CH, Fann JR, Tate DG, Richards JS, Wilson CS,
Warren AM, et al. An exploration of modiable risk factors for
depression aer spinal cord injury: which factors should we
target? Arch Phys Med Rehabil 2012;93:775-81.
7. National Spinal Cord Injury Model Systems. Spinal cord injury
facts and gures at a glance. [Internet]. 23 Jan 2015. Available
from: www.nscisc.uab.edu/PublicDocuments/fact_gures_docs/
Facts%202013.
8. Elliott TR, Frank RG. Depression following spinal cord injury.
Arch Phys Med Rehabil 1996;77:816-23.
9. Kennedy P, Rogers BA. Anxiety and depression aer spinal cord
injury: a longitudinal analysis. Arch Phys Med Rehabil 2000;81:
932-7.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
66
10. Dijkers MP. Quality of life of individuals with spinal cord
injury: a review of conceptualization, measurement, and research
ndings. J Rehabil Res Dev 2005;42:87-110.
11. Post M, Noreau L. Quality of life aer spinal cord injury. J
Neurol Phys er. 2005;29:139-46.
12. Holicky R, Charlifue S. Ageing with spinal cord injury: the
impact of spousal support. Disabil Rehabil 1999;21:250-7.
13. Dryden DM, Saunders LD, Rowe BH, May LA, Yiannakoulias
N, Svenson LW, et al. Depression following traumatic spinal
cord injury. Neuroepidemiology 2005;25:55-61.
14. Tate DG, Kalpakjian CZ, Forchheimer MB. Quality of life
issues in individuals with spinal cord injury. Arch Phys Med
Rehabil 2002;83:S18-25.
15. Lotrakul M, Sumrithe S, Saipanish R. Reliability and validity
of the ai version of the PHQ-9. BMC Psychiatry 2008;8:46.
16. Mahatnirankul S, Tantipiwatanaskul W, Pumpaisalchai W,
Wongsuwan K, Promanajirankul R. Comparison of the
WHOQOL-100 and the WHOQOL-BREF (26 items). J Ment
Health ai 1997;5:4-15.
17. Pongboriboon P, Tongprasert S, Kovindha A. Quality of life in
persons with spinal cord injury: a comparative study between
those with indwelling catheterization and intermittent
catheterization. J ai Rehabil Med 2011;21:13-20.
18. Krause JS, Saunders LL, Reed KS, Coker J, Zhai Y, Johnson E.
Comparison of the Patient Health Questionnaire and the Older
Adult Health and Mood Questionnaire for self-reported depressive
symptoms aer spinal cord injury. Rehabil Psychol 2009;54:
440-8.
19. Bombardier CH, Richards JS, Krause JS, Tulsky D, Tate DG.
Symptoms of major depression in people with spinal cord injury:
implications for screening. Arch Phys Med Rehabil 2004;85:
1749-56.
20. Hill MR, Noonan VK, Sakakibara BM, Miller WC, SCIRE Research
Team. Quality of life instruments and denitions in individuals
with spinal cord injury: a systematic review. Spinal Cord 2010;
48:438-50.
21. Choochart N, Vichiansiri R, Arayawichanon P, Manimmanakorn
N. Quality of life of traumatic spinal cord injured patients in
North-East region of ailand. J ai Rehabil Med 2015;25:15-21.
22. Wannapakhe J, Amatachaya S, Siritaratiwat W, Arrayawichanon
P, Wattanapan P. Physical abilities, quality of life, incidences
of complications and falls in patients with chronic spinal cord
injury aer discharge. J Med Tech Phy er 2010;22:188-9.
23. Trevittaya P, Wattanavittawat P. Quality of life and perceived
health status among people with spinal cord injury. Bull Chiang
Mai Assoc Med Sci 2016;49:185-96.
24. Scivoletto G, Petrelli A, Di Lucente L, Castellano V. Psychological
investigation of spinal cord injury patients. Spinal Cord 1997;
35:516-20.
25. Tzanos IA, Mavrogenis A, Gioti K, Papagelopoulos P,
Panagiotopoulos E. Depressive mood in individuals with spinal
cord injury (SCI) living in Greece. Spinal Cord 2018;56:883-9.
26. Craig AR, Hancock KM, Dickson HG. Spinal cord injury:
a search for determinants of depression two years aer the
event. Br J Clin Psychol 1994;33:221-30.
27. Kalpakjian CZ, Albright KJ. An examination of depression
through the lens of spinal cord injury. Comparative prevalence
rates and severity in women and men. Womens Health Issues
2006;16:380-8.
28. Orenczuk S, Mehta S, Slivinski J, Teasell RW. Depression
following spinal cord injury. In: Eng JJ, Teasell RW, Miller WC,
Wolfe DL, Townson AF, Hsieh JTC, Connolly SJ, Noonan
VK, Loh E, McIntyre A, editors. Spinal Cord Injury Rehabilitation
Evidence. Version 5.0. 2014. p. 1-33.
29. Dajpratham P, Kongkasuwan R. Quality of life among the
traumatic spinal cord injured patients. J Med Assoc ai
2011;94:1252-9.
30. Salilpramote B. Quality of life of patients with spinal cord
lesion. 12
th
Reg Med J 2016;27:112-8.
31. Jain NB, Sullivan M, Kazis LE, Tun CG, Garshick E. Factors
associated with health-related quality of life in chronic spinal
cord injury. Am J Phys Med Rehabil 2007;86:387-96.
32. Tate D, Forchheimer M, Maynard F, Dijkers M. Predicting
depression and psychological distress in persons with spinal
cord injury based on indicators of handicap. Am J Phys Med
Rehabil 1994;73:175-83.
Chavasiri et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
67
Original Article
SMJ
Karnkawin Patharateeranart, M.D.*, Jitladda Wasinrat, M.D.*, Jirawadee Yodying, M.D.*, Jitraporn Wongwiwatchai,
M.D.**
*Department of Radiology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, **Department of Radiology, Faculty of Medicine, Khon
Kaen University, Khon Kaen, ailand
Comparison of Computed Tomography Angiography
(CTA) Findings in Post-Endovascular Aortic
Aneurysm Repair CTA between Persistent and
TransientTypeIIEndoleak
ABSTRACT
Objective: To compare rst post-endovascular aortic aneurysm repair (EVAR) computed tomography angiography
(CTA) imaging characteristics between transient and persistent type II endoleaks.
Methods: is retrospective study enrolled patients who underwent EVAR and were diagnosed with type II endoleak
from rst post-operative CTA during January 2005 to October 2017 at the Department of Radiology, Faculty of
Medicine Siriraj Hospital, Mahidol University, Bangkok, ailand. Aneurysmal sac size, aneurysmal sac growth,
and endoleak were recorded among patients whose endoleak disappeared within 6 months (transient group), and
among patients whose endoleak persisted for more than 6 months (persistent group).
Results: Eighty-eight patients with a mean age of 75.3±7.3 years were included. Of those, 12 and 76 patients were
in the transient group and persistent group, respectively. ere were 71 males and 17 females. Univariate analysis
showed number of feeding arteries (odds ratio [OR]: 9.9, p=0.012) and presence of inferior mesenteric artery (IMA)
as an endoleak source (OR: 4.3, p=0.026) to be found more frequently in the persistent group than in the transient
group; however, neither factor survived multivariate analysis. No signicant dierence between two groups was
seen for endoleak diameter, endoleak complexity, or aneurysmal sac enlargement.
Conclusion: e number of feeder arteries and presence of IMA as an endoleak source on rst postoperative CTA
to be more likely found in patients with persistent type II endoleak. Further prospective study in a larger study
population is necessary to identify any existing statistically signicant dierences and/or associations.
Keywords: Comparison; computed tomography angiography; CTA; post-endovascular aortic aneurysm repair
CTA; persistent and transient type II endoleak (Siriraj Med J 2020; 72: 67-73)
Corresponding author: Jitladda Wasinrat
E-mail: jitladda.siriraj@gmail.com
Received 16 August 2019 Revised 30 September 2019 Accepted 30 October 2019
ORCID ID: http://orcid.org/0000-0001-8835-4599
http://dx.doi.org/10.33192/Smj.2020.09
INTRODUCTION
Endovascular aortic aneurysm repair (EVAR) is now
generally accepted worldwide as a treatment procedure
for abdominal aortic aneurysm because it has a lower rate
of perioperative mortality than open surgical repair.
1,2
However, despite EVAR being shown to have superior
perioperative survival advantage, its most well-known
complication – endoleak, which is dened as continuous
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
68
perfusion within the aneurysm sac – can cause aneurysmal
sac expansion that can lead to compromised long-term
survival outcome.
Type II endoleaks originate from retrograde ow
from collateral arteries to the aneurysmal sac. Typical
aortic side branches that cause type II endoleaks are the
inferior mesenteric artery (IMA), the lumbar arteries
(LAs), the median sacral artery, and the accessory renal
arteries. Early-onset type II endoleak is dened as an
endoleak found within 90 days aer EVAR, while late-
onset type II endoleak is detected aer 90 days. Early-onset
type II endoleak can be classied as transient (one that
spontaneously resolves within 6 months) or persistent
(endoleak that lasts longer than 6 months aer EVAR)
type II endoleak.
Persistent type II endoleaks have been reported as
a cause of late adverse outcome including aneurysmal
sac diameter enlargement, the need for switching to
open repair/ reintervention and rupture.
3
Researchers
have investigated for factors that predict persistent type
II endoleaks. In addition to demographic and patient
characteristics that have been discussed and reported
in previous studies as predictors of persistent type II
endoleak4, several other studies examined preprocedural
imaging ndings to identify factors that predict this
particular variety of endoleak.
5-9
e objective of this study was to identify the relation
of computed tomography angiography (CTA) ndings in
rst post-EVAR CTA between transient and persistent
type II endoleaks.
MATERIALS AND METHODS
Patients
is study was approved by the Siriraj Institutional
Review Board (SIRB) of the Faculty of Medicine Siriraj
Hospital, Mahidol University (Si 051/2018). e study
retrospectively enrolled patients who were diagnosed with
type II endoleak from rst post-operative CTA during
January 2005 to October 2017 at the Department of
Radiology, Faculty of Medicine Siriraj Hospital, Mahidol
University. The first post-operative CTA should be
performed within 90 days aer EVAR to evaluate for
early-onset type II endoleak, and the next post-operative
CTA should be performed at least 6 months aer EVAR
according to the criteria for persistent type II endoleak.
A search of our center’s database and radiology reporting
system revealed 88 patients (mean age 75.3±7.3 years,
71 males and 17 females) that met the inclusion criteria.
Patients with other types of endoleak or who were lost
to follow-up were excluded.
Imaging study
Two sessions of CTA were interpreted. e rst
postoperative CTA within the rst 90 days aer EVAR
was used for measurement of CTA ndings, and the
follow-up CTA was used to assess the outcome. CT scans
were obtained using helical CT scanners (1. Somatom
Denition dual source CT; Siemens, Forchheim, Germany
or 2. Discovery CT 750HD; GE Healthcare, Milwaukee,
USA or 3. Revolution CT, GE Healthcare, Milwaukee,
USA). e CTA protocol consists of unenhanced CT
scan, contrast-enhanced CTA phase, and delayed phase.
Contrast-enhanced CTA phase was used by bolus-tracking
technique with a threshold of 150 Hounseld units (HUs)
at abdominal aorta proximal to the endovascular stent
gra. Delayed phase was performed 2 minutes aer
bolus trigger.
Imaging interpretation
Evaluation and measurement of CTA findings
were performed by two radiologists who separately
interpreted the image ndings on our center’s Picture
Archive Communication System (PACS). e following
CTA ndings were analyzed from the rst postoperative
CTA: size of the aneurysmal sac, name of feeder artery and
total number of arteries feeding the endoleak, endoleak
complexity, diameter of endoleak, attenuation of the
endoleak, and attenuation of aorta.
Aneurysmal sac diameter was measured using
electronic caliper to the nearest millimeter. Measurement
of the orthogonal sac diameter using double oblique
short axis was performed (Fig 1). e number of feeder
arteries included all visible arterial feeders. A common
vessel of lumbar arteries, and a common vessel of the
fourth lumbar artery and the median sacral artery were
counted as one feeding artery. Endoleak complexity was
classied into simple type and complex type. A simple
type endoleak was dened as an endoleak that is fed by a
single artery, while a complex type endoleak was dened
as being fed by two or more arteries. Measurement of
endoleak diameter was performed in axial plane of delayed
phase CT images. e maximal diameter was measured
and recorded (Fig 2). Attenuation of endoleak and aorta
was measured using the greatest possible circular region
of interest (ROI) on both CTA and delayed phases in
Hounseld units. Relative attenuation of endoleak was
calculated using the following formula: (endoleak cavity
attenuation on CTA and delayed phases – endoleak cavity
attenuation on unenhanced images) / (attenuation of stent
gra lumen on CTA and delayed phases – attenuation of
stent gra lumen on unenhanced images). e outcome
Patharateeranart et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
69
Original Article
SMJ
Fig 1. Post-EVAR CTA images show the method used to measure
the aneurysmal sac. (A) Double oblique axial view used to measure
the orthogonal diameter of the aneurysm. (B) and (C) Coronal and
sagittal reformatted images, respectively.
Statistical analysis
All data analyses were performed using SPSS
Statistics software (SPSS, Inc., Chicago, IL, USA).
Values are shown as mean and/or standard deviation
or number and percentage, as appropriate. Univariate
analysis was performed using chi-square test, t-test
or Mann-Whitney U test, and multivariate analysis
was done using logistic regression analysis to identify
factors independently associated with persistent type II
endoleak. Interobserver agreement of aneurysmal sac
diameter measurement, number of feeder arteries, endoleak
diameter measurement, and attenuation measurement
was calculated using intraclass correlation coecient
(ICC, r). e r values were classied, as follows: 1.0,
perfect agreement; 0.81-0.99, almost perfect agreement;
0.61-0.80, substantial agreement; 0.41-0.60, moderate
agreement; 0.21-0.40, fair agreement; and, less than 0.2,
slight agreement. Interobserver agreement of endoleak
complexity, presence of each feeder, and aneurysmal sac
enlargement were calculated using kappa statistic. e
kappa values (k) were interpreted, as follows: 0.81-1.00,
very good agreement; 0.61-0.80, good agreement; 0.41-
0.60, moderate agreement; 0.21-0.40, fair agreement;
and, less than 0.20, poor agreement. A p-value < 0.05
was considered statistically signicant.
RESULTS
Patients
Eighty-eight patients were included. ere was no
signicant dierence in age between the transient and
persistent groups (73.92±7.15 years vs. 75.57±7.30 years,
respectively). Of the 88 early-onset type II endoleak
patients, 12 patients had spontaneous resolution (transient
group), and 76 patients still had type II endoleak on the
6-month follow-up CTA (persistent group). ere was
no signicant dierence in gender distribution between
groups (p=0.236). Patient demographic data are shown
in Table 1.
Endoleak characteristics
Initial mean aneurysmal sac diameter in the transient
group and persistent group was 57.0±11.57 mm and
58.92±14.34 mm, respectively (p=0.66). In terms of
endoleak complexity, simple endoleak was found in 7
transient type II endoleak patients, and in 24 persistent type
II endoleak patients. Complex endoleak was observed in
5 transient type II endoleak patients, and in 52 persistent
type II endoleak patients (p=0.103).
e number of arterial feeders of endoleak was
separated into two groups: 2 or less arterial feeders,
and more than 2 feeders. e transient type II endoleak
Fig 2. Post-EVAR CTA, delayed phase, axial plane image demonstrates
measurement of endoleak diameter.
(transient or persistent type II endoleak and aneurysmal
sac diameter enlargement) was recorded based on the
ndings from CTA study performed 6 months aer
EVAR.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
70
group had a signicantly greater proportion of 2 or less
arterial feeders than the persistent group (91.7% vs.
52.6%, respectively; odds ratio [OR]: 9.9, 95% condence
interval [CI]: 1.217-80.526; p=0.012).
Regarding endoleak sources (dierent arterial feeders),
presence of inferior mesenteric artery (IMA) was found
more frequently in the persistent group (68.4%) than in
the transient group (33.3%) (OR: 4.3, 95% CI: 1.18-15.8;
p=0.026). However, there was no signicant dierence
between two groups for presence of lumbar artery (LA),
median sacral artery, or accessory renal artery as an
arterial feeder of type II endoleak.
Endoleak diameter showed no signicant dierence
between groups. Median diameter of type II endoleak
in the transient group and persistent group was 15 mm
and 13 mm, respectively (p=0.87). e mean attenuation
of the endoleak cavity was not signicantly dierent
between the transient and persistent groups in unenhanced
(43.0 vs. 43.8 HUs), arterial (281.6 vs. 273.7 HUs), and
delayed (140.1 vs. 128.9) phases. e relative attenuation
of endoleak in both the arterial and delayed phases was
also not signicantly dierent between two groups. e
data are shown in Table 2. Regarding aneurysmal sac
enlargement, the persistent type II endoleak group had
6 patients (7.9%) with enlarged aneurysmal sac, while
the transient type II endoleak group had no patients
(0%) with enlarged aneurysmal sac (p=0.591).
Multivariate analysis
Univariate analysis revealed statistically signicant
dierences in three variables, including endoleak complexity,
number of arterial feeders, and presence of inferior
mesenteric mesentery feeder. However, when entered
into multivariate analysis, none of these factors was
found to be an independent predictor of persistent type
II endoleak. e odd ratio for endoleak complexity,
number of arterial feeders, and presence of inferior
mesenteric feeder was 1.302 (95% CI: 0.318-5.333), 5.532
(95% CI: 0.487-62.797), and 2.285 (95% CI: 0.557-9.365),
respectively (Table 3).
Interobserver agreement
Intraclass correlation coecient (ICC) revealed
almost perfect agreement for aneurysmal sac diameter
measurement (r=0.961) and endoleak diameter measurement
(r=0.922). ere was substantial agreement in the counting
of the number of arterial feeders (r=0.635). Attenuation
measurement was moderate to substantial agreement
(r=0.413-0.781). Cohen’s kappa was used to calculate
the kappa value. Presence of type II endoleak on follow-
up study, endoleak complexity, and identication of
arterial feeders showed good agreement (k=0.779-0.855).
Aneurysmal sac enlargement evaluation showed moderate
agreement (k=0.482).
DISCUSSION
A few studies
10,11
investigated type II endoleak and
found no signicant association relative to age, gender,
or initial postoperative aneurysmal sac diameter in both
transient and persistent type II endoleak – all of which
is similar to our study.
Type II endoleaks can be classied into two subtypes,
simple and complex subtypes based on number of inow/
outow vessels. e authors compared simple and complex
type II endoleak and found complex type II endoleak
to be observed more frequently in the persistent group
(68.4%) than in the transient group (41.7%). However,
the dierence between two groups was not statistically
signicant. is result is not directly comparable to that
reported by Muller-Wille R, et al.
8
because they used a
dierent method of outcome measurement. Muller-Wille
R, et al.
8
classied simple and complex type II endoleak
into four subtypes, and they found complex IMA-LA
subtype to be signicantly associated with aneurysmal
sac enlargement (OR: 20.8, 95% CI: 4.9-88.9; p<0.001).
TABLE 1. Patient characteristics.
TypeIIendoleakduringfollow-up
P-value
Transient (n=12) Persistent (n=76)
Mean age (years) 73.92±7.15 75.57±7.30 0.468
Gender
Male 8 (66.7%) 63 (82.9%) 0.236
Female 4 (33.3%) 13 (17.1%)
A p-value<0.05 indicates statistical signicance
Patharateeranart et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
71
Original Article
SMJ
TABLE 2. Endoleak CTA characteristics.
TypeIIendoleakduringfollow-up Oddsratio
Transient (n=12) Persistent (n=76) (95% CI)
P-value
Aneurysmal sac diameter 57.0±11.57 58.92±14.34 0.66
(mm), mean±SD
Endoleakcomplexity,n(%)
Simple 7 (58.3%) 24 (31.6%) 3.03(0.87-10.5) 0.10
Complex 5 (41.7%) 52 (68.4%)
Number of feeders, n (%) 9.9 (1.217-80.526) 0.012
≤2 11(91.7%) 40(52.6%)
>2 1(8.3%) 36 (47.4%)
Endoleaksource,n(%)
IMA 4 (33.3%) 52 (68.4%) 4.3 (1.18-15.8) 0.026
LAs 10 (83.3%) 60 (78.9%) 0.75 (0.15-3.77) 1
Median sacral artery 0 (0.0%) 2 (2.6%) 0.86 (0.79-0.94) 1
Accessory renal artery 0 (0.0%) 2 (2.6%) 0.86 (0.79-0.94) 1
Endoleakdiameter(mm),median(range) 15 (6, 23) 13 (3, 62) 0.87
Attenuation (HU±SD)
Unenhanced 43±9.4 43.86±13.77 0.835
Arterial phase 281.58±100 273.7±104.2 0.807
Delayed phase 140.1±27.14 128.9±27.47 0.257
Relative attenuation CTA 0.638±0.278 0.693±0.394 0.64
Relative attenuation delay 1.058±0.601 0.914±0.307 0.248
Sac enlargement, n (%) 0 (0.0%) 6 (7.9%) 0.854 (0.78-0.934) 0.591
A p-value<0.05 indicates statistical signicance
Abbreviations: CTA, computed tomography angiography; CI, condence interval; SD, standard deviation; IMA, inferior mesenteric artery;
LAs, lumbar arteries; HU, Hounseld unit
TABLE 3. Multivariate analysis.
Odd ratios (95%CI) P-value
Complexity 1.302 (0.318-5.333) 0.714
Number of feeders 5.532 (0.487-62.797) 0.168
IMA feeder 2.285 (0.557-9.365) 0.251
A p-value >0.05 indicates statistical signicance
Abbreviations: CI, condence interval; IMA, inferior mesenteric artery
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
72
Number of feeder arteries showed a signicantly
greater proportion of 2 or less arterial feeders in the
transient group (91.7%) than in the persistent group
(52.6%); p = 0.012. e authors found this result to be
similar to that from a study by Maeda, et al.
10
Maeda,
et al.
10
found multiple vessels responsible for type II
endoleak to be a signicant factor for prediction of
persistent type II endoleak. In the present study, that
relationship did not survive multivariate analysis. is
dierence in results may be due to the small size of our
transient group that inuenced a widely ranging 95%
condence interval.
Regarding sources of type II endoleak, we found a
signicantly greater proportion of inferior mesenteric
artery feeder in the persistent group (68.4%) than in the
transient group (33.3%). According to a study by Muller-
Wille R, et al.
8
(despite this result being not directly
comparable due to dierences in outcome measurement
between studies), presence of inferior mesenteric artery
as a feeder artery was shown to be more frequent in the
aneurysm enlargement group than in the non-enlargement
group.
In contrast to Timaran, et al.
11
, Keedy, et al.
12
, and
Dudeck, et al.
13
, the endoleak diameter measured on axial
CT in our study was not signicantly dierent between
the groups. Timaran, et al.
11
found maximum endoleak
cavity diameter more than 15 mm to be associated with
increased risk of increase aneurysmal sac diameter (relative
risk [RR]: 11.1, 95% CI: 1.4-85.8; p=0.02). Keedy, et al.
12
reported transverse diameter of the endoleak cavity in
the intervention group to be signicantly greater than
in the nonintervention group (1.85±1.01 vs. 1.13±0.83
cm, respectively; p=0.007). Dudeck, et al.
13
also identied
diameter and area of endoleak nidus in the reintervention
group to be signicant greater than in the surveillance
group. However, our endoleak diameter study result was
similar to reports from Mursalin, et al.
7
and Muller-Wille
R, et al.
8
, both of which reported that simple evaluation of
diameter and area of endoleak cavity was not of predictive
value.
Attenuation measurement of endoleak from our study
showed no signicant dierence between that transient
and persistent groups for both absolute measurement and
relative attenuation. e results of absolute attenuation
measurement corresponded to a report from Muller-
Wille R, et al.
8
that reported no signicant dierence
between the no aneurysm enlargement group and the
aneurysm enlargement group.
Presence of aneurysmal sac enlargement in the
present study showed no signicant dierence between
the transient and persistent groups, and this corresponds
with the findings of Kaley Pippin, et al.
14
However,
both our study and that study had a small sized study
population.
Limitations
is study has three major limitations. First, the
retrospective design of this study made it dicult to
control the CT protocol. Second, the small number of
patients in the transient type II endoleak group may have
adversely aected the ability of our statistical analysis to
detect signicant dierences between groups. ird, our
results cannot be applied in late-onset type II endoleak
because we included only patients with early-onset type
II endoleak.
TABLE 4. Interobserver agreement.
Intraclasscorrelationcoefcient(ICC,r)
Sac diameter 0.961
Number of feeders 0.635
Endoleak diameter 0.922
Attenuation measurement 0.781
Kappavalue(k)
Presence of type II endoleak 0.782
Endoleak complexity 0.779
Presence of each feeder 0.855
Aneurysmal sac enlargement 0.482
Patharateeranart et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
73
Original Article
SMJ
CONCLUSION
e results of this study revealed endoleak complexity,
the number of feeder arteries and presence of IMA as an
endoleak source on rst postoperative CTA to be more
likely found in patients with persistent type II endoleak.
Further prospective study in a larger study population
is needed to identify any existing statistically signicant
dierences and/or associations.
ACKNOWLEDGMENTS
e authors gratefully acknowledge Miss Julaporn
Pooliam of the Division of Clinical Epidemiology,
Department of Research, Faculty of Medicine Siriraj
Hospital, Mahidol University, Bangkok, ailand for
assistance with statistical analysis.
Conict of interest declaration: Both authors declare
no personal or professional conicts of interest, and no
nancial support from the companies that produce and/
or distribute the drugs, devices, or materials described
in this report.
Funding disclosure: is was an unfunded study.
REFERENCES
1. Investigators TUKET. Endovascular versus open repair of
abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71.
2. Lederle FA, Freischlag JA, Kyriakides TC, Matsumura JS, Padberg
FTJ, Kohler TR, et al. Long-term comparison of endovascular
and open repair of abdominal aortic aneurysm. N Engl J Med
2012;367:1988-97.
3. Jones JE, Atkins MD, Brewster DC, Chung TK, Kwolek CJ,
LaMuraglia GM, et al. Persistent type 2 endoleak aer endovascular
repair of abdominal aortic aneurysm is associated with adverse
late outcomes. J Vasc Surg 2007;46:1-8.
4. Lo RC, Buck DB, Herrmann J, Hamdan AD, Wyers M, Patel VI,
et al. Risk factors and consequences of persistent type II
endoleaks. J Vasc Surg 2016;63:895-901.
5. Christopher J. Abularrage, Robert S. Crawford, Mark F. Conrad,
Hang Lee, Christopher J. Kwolek, David C. Brewster, et al.
Preoperative Variables Predict Persistent Type 2 Endoleak
Aer Endovascular Aneurysm Repair. J Vasc Surg 2010;52:
19-24.
6. Couchet G, Pereira B, Carrieres C, Maumias T, Ribal J-P, Ben
Ahmed S, et al. Predictive factors for type II endoleaks aer
treatment of abdominal aortic aneurysm by conventional
endovascular aneurysm repair. Ann Vasc Surg 2015;29:1673-9.
7. Mursalin R, Sakamoto I, Nagayama H, Sueyoshi E, Tanigawa
K, Miura T, et al. Imaging-based predictors of persistent type
II endoleak aer endovascular abdominal aortic aneurysm
repair. AJR Am J Roentgenol 2016;206:1335-40.
8. Müller-Wille R, Schötz S, Zeman F, Uller W, Güntner O, Pster
K, et al. CT features of early type II endoleaks aer endovascular
repair of abdominal aortic aneurysms help predict aneurysm
sac enlargement. Radiology 2015;274:906-16.
9. Ward TJ, Cohen S, Patel RS, Kim E, Fischman AM, Nowakowski
FS, et al. Anatomic risk factors for type-2 endoleak following
EVAR: A retrospective review of preoperative CT angiography
in 326 patients. Cardiovasc Intervent Radiol 2014;37:324-8.
10. Maeda T, Ito T, Kurimoto Y, Watanabe T, Kuroda Y, Kawaharada
N, et al. Risk factors for a persistent type 2 endoleak aer
endovascular aneurysm repair. Surg Today 2015;45:1373-7.
11. Timaran CH, Ohki T, Rhee SJ, Veith FJ, Gargiulo NJ, Toriumi
H, et al. Predicting aneurysm enlargement in patients with
persistent type II endoleaks. J Vasc Surg 2004;39:1157-62.
12. Keedy AW, Yeh BM, Kohr JR, Hiramoto JS, Schneider DB,
Breiman RS. Evaluation of potential outcome predictors in
type II endoleak: A retrospective study with CT angiography
feature analysis. AJR Am J Roentgenol 2011;197:234-40.
13. Dudeck O, Schnapau D, Herzog L, Löwenthal D, Bulla K,
Bulla B, et al. Can early computed tomography angiography
aer endovascular aortic aneurysm repair predict the need
for reintervention in patients with type II endoleak? Cardiovasc
Intervent Radiol 2015;38:45-52.
14. Pippin K, Hill J, He J, Johnson P. Outcomes of type II endoleaks
aer endovascular abdominal aortic aneurysm (AAA) repair:
a single-center, retrospective study. Clin Imaging 2016;40:
875-9.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
74
Ekkarat Wongsawat, B.Sc.*, Panisadee Avirutnan, M.D., Ph.D.**, Yoon-Won Kim, Ph.D.***, Yupin Suputtamongkol,
M.D.*
*Department of Medicine, **Division of Dengue Hemorrhagic Fever Research, Department of Research and Development, Faculty of Medicine Siriraj
Hospital, Mahidol University, Bangkok, ailand, ***Department of Microbiology, College of Medicine, Hallym University, Chuncheon, South Korea.
Performance of Two Commercial Dengue NS1
Rapid Tests for the Diagnosis of Adult Patients
with Dengue Infection
ABSTRACT
Objective: to determine the diagnostic performance of two commercially available dengue NS1 antigen rapid
detection tests (RDTs); namely the SD BIOLINE Dengue NS1 RDT and the ImmuneMed Dengue NS1 Rapid, for
the diagnosis of adult patients with dengue infection.
Methods: e study was performed by using archived samples from 237 patients with the laboratory-conrmed
dengue infection. Archived, well-characterized samples from an additional 208 febrile individuals from ailand
were used as the control group. Reference testing to conrm the diagnosis in dengue patients included RT-PCR
and in-house NS1 ELISA, in-house IgM and IgG capture ELISAs.
Results: e sensitivity of the SD BIOLINE Dengue NS1 RDT was 100%, and the specicity was 99% (95%CI 96.6
to 99.7%). False positive was found in 2 samples from patient with scrub typhus. e sensitivity and specicity of
the ImmuneMed Dengue NS1 Ag Rapid were 97.4% (95% CI, 95.5 to 99.5%) and 96.6% (95%CI 93.5 to 98.4%)
respectively. False positive results were found in 7 patients with murine typhus, scrub typhus, and inuenza.
Conclusion: Both RDTs showed comparable sensitivity and specicity in this study population. Data from this study
could be used to facilitate data-driven laboratory test choices for managing patient care during dengue outbreaks.
Keywords: Dengue infection; nonstructural protein -1; adult patients; rapid detection test (Siriraj Med J 2020; 72:
74-78)
Corresponding author: Yupin Suputtamongkol
E-mail: ysuputtamongkol@gmail.com
Received 24 October 2019 Revised 4 December 2019 Accepted 19 December 2019
ORCID ID: http://orcid.org/0000-0001-7324-1698
http://dx.doi.org/10.33192/Smj.2020.10
INTRODUCTION
Dengue fever is an arboviral disease that is a public
health priority in most tropical countries including
ailand.
1
Dengue fever is caused by the dengue virus
(DENV), a avivirus that can be classied into four
predominant serotypes (DENV-1, -2, -3, and -4).
2
DENV
comprises three structural proteins (capsid, membrane,
and envelope) and seven nonstructural (NS) proteins
(NS1, NS2a, NS2b, NS3, NS4a, NS4b, and NS5NV.
DENV is transmitted by mosquitoes, principally Aedes
aegypti and Aedes albopictus. Clinically, dengue fever
is characterized by fever, headache, myalgia, arthralgia,
rash, leukopenia, and sometimes thrombocytopenia.
2,3
e severity of the disease varies from asymptomatic or
mild to severe with high fever, hemorrhage, and shock.
2,3
ere is no antiviral drug to cure dengue fever.
2
e only
available treatment options are supportive therapies,
including bed rest, uids, and symptomatic relief using
Wongsawat et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
75
Original Article
SMJ
analgesics. As many patients present with nonspecic fever
requiring dierential diagnosis, laboratory conrmation
using a rapid, accurate, and relatively low-cost diagnostic
test is especially important.
3
Laboratory diagnosis for
DENV infection includes detection of the virus, genome,
NS-1 antigen or IgM/IgG antibodies, or a combination
of these tests.
2-4
Rapid diagnostic tests (RDTs), using
immunochromatographic assay (ICT) to detect antigen
and/ or IgM, IgG, are commonly used for DENV diagnosis
because of their simplicity and rapidity.
4
Several RDTs
are now widely available from dierent manufacturers.
e sensitivities of these NS1 RDTs vary among dierent
DENV serotype and duration of illnesses.
5
e SD BIOLINE
Dengue NS1 Ag RDT (Standard Diagnostic Incorporation,
Gyeonggi-do, Republic of Korea) and the ImmuneMed
dengue NS1 Ag rapid (ImmuneMed, Inc., South Korea)
are among the widely use RDTs in ailand. We reported
here the comparative performance of these two RDTs
in adult patient with dengue infection.
MATERIALS AND METHODS
Patient population
Plasma samples were obtained from 237 adult
patients with laboratory-conrmed dengue infection,
as part of the clinical study of dengue infection, carried
out at Siriraj Hospital, Bangkok and Loei Hospital, Loei
Province, ailand. e diagnosis of dengue infection
was conrmed by RT-PCR in all patients. e dengue
serotype was determined by both reverse transcriptase
quantitative PCR (RT-qPCR) protocol and NS1 serotype
specic ELISA.
6-9
For control group, plasma samples
were collected from patients (male: female = 2:1), aged
15- 84 years (mean age 45 years) with acute febrile
illnesses from Siriraj Hospital, and 4 more hospitals in
ailand (Maharaj Nakhon Ratchasima Hospital, Nakhon
Ratchasima Province, Loei Hospital, Loei Province, and
Banmai Chaiyapod Hospital, Burirum Province). Blood
samples were collected as a part of studies investigating
the causes of fever.
10-12
All of these samples were tested
by indirect immunofluorescent assay (IFA) for the
diagnosis of leptospirosis, scrub typhus, and murine
typhus. Details of the test and interpretation of IFA is
as described previously.
10-12
In addition, 2 aerobic blood
cultures were done in patients with suspected septicemia
or bacteremia. All patients in the control group had
negative NS1 and IgM and IgG test for dengue. ese
clinical studies were approved by the Ethical Review
Subcommittee of the Public Health Ministry of ailand
and Siriraj Institutional Review Board Faculty of Medicine
Siriraj Hospital, Mahidol University (Si 148/2019). All
patients provided the informed written consent before
sample collection. Plasma from patients in both groups
were divided for immediate use and the leover samples
were stored at –70°C. until further used. e present
study protocol was approved Siriraj Institutional Review
Board, Faculty of Medicine Siriraj Hospital, Mahidol
University, ailand.
e IFA assay for the laboratory conrmation of
leptospirosis, scrub typhus, and murine typhus was
performed as described previously.
6,11,12
e Leptospira
interrogans, serovar autumnalis; pooled O. tsutsugamushi
from Karp, Kato and Gilliam strains; and Rickettsia
typhi (Wilmington strain) were used as the antigens
for the detection of IgM and IgG antibodies for the
diagnosis of leptospirosis, scrub typhus, and murine
typhus respectively.
The SD BIOLINE dengue NS1 RDT and the ImmuneMed
Dengue NS1 Ag Rapid were performed according to the
company’s instruction. In brief, approximately 80 μL
of plasma was applied to the ICT sample well of each
RDT, and then approximately 2 drops (80 μL) of sample
diluent was added into the sample well immediately. e
results were interpreted visually at 15 minutes. e test
result was considered negative if only the control band
was stained. If the test and control bands were stained,
the test result was considered positive. Weakly positive
of both RDTs was considered as positive in this study.
Statistical methods
e analysis was performed on admission samples of
all patients. e standard diagnostic accuracy indices of
sensitivity, specicity, with the 95% condence intervals
(CIs) were calculated, using the SPSS18.0 software
(SPSS Inc., Chicago, IL, USA). We did not calculate
the positive and the negative predictive value for both
rapid ICT because we used the stored samples collected
from various hospitals, at dierent period of times. As
a result, the proportion of samples from patients with
dengue infection did not represent the true prevalence
of dengue infection among patients with acute fever in
ailand.
RESULTS
In dengue group, sample from 237 patients (M: F=
135: 102, median age 24, range 15-72 years) were tested.
e median duration of fever was 2 days (range 1-5 day).
Dengue -4 was the most common (n=129 or 54.4%, all
secondary infection) serotype found during the collection
period, followed by DENV-3 (n=52 or 21.9%, with 11.5%
primary infection), DENV-2 (n=34 or 14.3%, with 8.8%
primary infection), and DENV-1 (n= 22 or 9.3%, with
13.6% primary infection). e distribution of cases by
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
76
day of onset of illness and DENV serotype is shown in
Fig 1. For control group, there were 208 patients with
conrmed other infections including; scrub typhus, n=
65, 31.3%; murine typhus, n= 43, 207%; leptospirosis,
n=41, 19.7%; inuenza infection, n=24, 11.5%; zika virus
infection, n= 18 (8.7%), bacteremia from various bacteria
such as Escherichia coli, n=10, 4.8%, and malaria, n=7,
3.4%.
e sensitivity of SD BIOLINE Dengue NS1 Ag
RDT was 100% and the specicity was 99.0% (95% CI
96.6%- 99.7%). ere were 2 samples from patients
with scrub typhus that had false positive SD BIOLINE
Dengue NS1 Ag RDT. e sensitivity and specicity
of the ImmuneMed dengue NS1 Ag Rapid were 97.4%
(95.5- 99.5%) and 96.6% (95% CI 93.2-98.4) respectively.
Six dengue patients with secondary infection (5 patients
with DENV-4, and 1 patient with DENV-3) had false
negative test. False positive test results occurred among
7 patients; murine typhus (3), zika virus infection (1),
leptospirosis (1), scrub typhus (1). Details of results of
the analysis is shown in Table 1.
Fig 1. e distribution of samples (%) of patients with dengue infection by duration of illness in day and DENV- serotypes.
TABLE 1. Sensitivities (95%CI) and specicities (95%) compared between two RDTs for the rapid diagnosis of
dengue infection in adults.
Group Dengue infection Other Infections Total
RDT (n=237) (n= 208)
SD BIOLINE NS1 RDT
Positive, n 237 2 239
Negative, n 0 206 206
Total, n (%) 237 (100%) 208 (99.0%) 445
ImmuneMed NS1 RDT
Positive, n 231 7 238
Negative, n 6 201 207
Total, n (%) 237 (97.5%) 208 (96.6%) 445
Wongsawat et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
77
Original Article
SMJ
DISCUSSION
Dengue is a common cause of fever in adult patients
in tropical countries. e incidence of dengue among
non-malaria fever varied from 5 to 9% in these countries.
In addition, recent report from ailand indicated that
a number of adult patients who died of dengue virus are
misdiagnosed as severe sepsis and septic shock.
13
us
diagnosis of dengue based on clinical features alone is
dicult. Rapid NS1 antigen detection, using ICT method,
is the most widely used method for the diagnosis of
dengue in endemic areas.
4,5
However, the sensitivity
and specicity of NS1 antigen vary, due to the dierent
study population (children or adult), serotype circulation,
immune status (primary or secondary infection) and
onset of illness.
5,14,15
ese factors vary from place to
place and time to time. us we evaluated the sensitivity
and specicity of two NS1 Ag RDT that are currently
used in ailand. e gold standard diagnosis of dengue
infection in this study included both RT-PCR and NS1
ELISA. In this population where most adult patients
had DENV-4 infection and secondary infection. Both
RDT showed a comparable sensitivities and specicities.
False positive of either RDTs should not explained by
the coinfection of dengue with scrub typhus or murine
typhus or inuenza because none of them had positive
of both RDTs. In addition we performed NS1 antigen
detection using in-house ELISA, and none of them had
detectable NS1. e SD BIOLINE NS1 RDT showed a
consistent performance with the previous reports.
4,5,16
On the other hand, this is the rst study for the evaluation
of the ImmuneMed Dengue NS1 Ag Rapid. Although
the comparable performance with the more widely used,
SD BIOLINE NS1 RDT was shown, more studies with
broader population ranges, to conrm result of the
performance of ImmuneMed Dengue NS1 Ag Rapid,
are needed.
ere was a selection bias of sample of dengue patient
in this study because all sample tested were collected from
patient who had positive SD BIOLINE NS1 RDT as the
inclusion criteria of those clinical studied. erefore, it
was not possible to calculate the true sensitivity and 95%
CI of this RDT in this study. Another limitation of this
study is that we only enrolled patient within 5 days of
onset of illness. Although patient with suspected dengue
infection were commonly presented to the hospital within
this period of illness, sensitivity and specicity of both
RDT beyond this duration were not determined in this
study.
In conclusion, we found that both commercially
point-of care NS1 antigen detection RDTs had consistent
performance for the initial diagnosis of adult patient
with dengue infection in the early phase of illness. is
information could be used to facilitate data-driven laboratory
test choices for managing patient care during dengue
outbreaks.
ACKNOWLEDGMENTS
e authors thank Standard Diagnostic Incorporation,
Gyeonggi-do, Republic of Korea and ImmuneMed
Incorporation, Gangwon-do, Republic of Korea for
donation of the RDT used in this study. ey do not
have involvement in the study design or results analysis.
Conicts of interest and source of funding: Y.S. and
P.A. have been supported by a Siriraj Chalermprakiat
Grant and P.A. has been supported by a Research Lecturer
Grant, Faculty of Medicine Siriraj Hospital, Mahidol
University. K.Y. advised the scientic contribution when
ImmuneMed Dengue NS1 Ag Rapid has been developed.
REFERENCES
1. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW,
Moyes CL, et al. e global distribution and burden of dengue.
Nature 2013;496:504-7.
2. World Health Organization. Dengue guidelines for diagnosis,
treatment, prevention and control: New Edition. WHO/HTM/
NTD/DEN/2009.1 (World Health Organization, 2009).
3. Centers for Disease Control and Prevention. 2012. Laboratory
guidance and diagnostic testing. Centers for Disease Control
and Prevention, Atlanta, GA: http://www.cdc.gov/ dengue/
clinicalLab/laboratory.html.
4. Andries AC, Duong V, Ngan C, Ong S, Huy R, Sroin KK, et al.
Field evaluation and impact on clinical management of a rapid
diagnostic kit that detects dengue NS1, IgM and IgG. PLoS
Negl Trop Dis 2012;6:e1993.
5. Lee H, Ryu J H, Park H-S, Park K H, Bae H, Yun S, et al.
Comparison of six commercial diagnostic tests for the detection
of dengue virus non-structural-1 antigen and IgM/IgG antibodies.
Ann Lab Med 2019;39:566-71.
6. Yenchitsomanus PT, Sricharoen P, Jaruthasana I, Pattanakitsakul
SN, Nitayaphan S, Mongkolsapaya J, et al. Rapid detection
and identication of dengue viruses by polymerase chain reaction
(PCR). Southeast Asian J Trop Med Pub Health 1996;27:228- 36.
7. Puttikhunt C, Prommool T, U-thainual N, Ong-ajchaowlerd P,
Yoosook K, Tawilert C, et al. e development of a novel
serotyping-NS1-ELISA to identify serotypes of dengue virus.
J Clin Virol 2011;50:314-9.
8. Innis BL, Nisalak A, Nimmannitya S, Kusalerdchariya S,
Chongswasdi V,Suntayakorn S,et al. An enzyme-linked
immunosorbent assay to characterize dengue infections where
dengue and Japanese encephalitis co-circulate. Am J Trop
Med Hyg 1989;40:418- 27.
9. Shu PY, Chang SF, Kuo YC, Yueh YY, Chien LJ, Sue CL, et al.
Development of group- and serotype-specic one-step SYBR
green I-based real-time reverse transcription-PCR assay for
dengue virus. J Clin Microbiol 2003;41:2408-16.
10. Suttinont C, Losuwanaluk K, Niwatayakul K, Hoontrakul S,
Intaranongpai W, Silpasakorn S, et al. Causes of acute
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
78
undierentiated febrile illness in rural ailand: a prospective
observational study. Ann Trop Med Hyg 2006;100:363-70.
11. Suputtamongkol Y, Niwattayakul K, Suttinont C, Losuwanaluk
K, Limpaiboon R, Chierakul W, et al. An open, randomized,
controlled trial of penicillin, doxycycline, and cefotaxime for
patients with severe leptospirosis. Clin Infect Dis 2004;39:1417-
24.
12. ipmontree W, Suputtamongkol Y, Tantibhedhyangkul W,
Suttinont C, Wongsawat E, Silpasakorn S. Human leptospirosis
trends: northeast ailand, 2001-2012. Int J Environ Res Public
Health 2014;11:8542-51.
13. Teparrukkul P, Hantrakun V, Day NPJ, West TE. Management
and outcomes of severe dengue patients presenting with sepsis
in a tropical country. PloS One 2017;12:e0176233.
14. Tricou V, Vu HT, Quynh NV, Nguyen CV, Tran HT, Farrar J,
et al. Comparison of two dengue NS1 rapid tests for sensitivity,
specicity and relationship to viraemia and antibody responses.
BMC Infect Dis 2010;10:142.
15. Hsieh CJ, Chen MJ. e commercial dengue NS1 antigen-
capture ELISA may be superior to IgM detection, virus isolation
and RT-PCR for rapid laboratory diagnosis of acute dengue
infection based on a single serum sample. J Clin Virol 2009;44:
102.
16. Pal S, Dauner AL, Mitra I, Forshey BM, Garcia P, Morrison
AC, et al. Evaluation of dengue NS1 antigen rapid tests and
ELISA kits using clinical samples. PLoS One 2014;9:e113411.
Wongsawat et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
79
Original Article
SMJ
Wipawee Booranapong, M.D.*, Pinnita Prabhasawat, M.D.*, Suksri Chotikavanich, M.D.*, Sasima Tongsai,
Ph.D.**, Patsathorn Naranunn, M.D.*, Wilaipun aweerattanasilp, M.D.*, Panida Kosrirukvongs, M.D.*
*Department of Ophthalmology, **Clinical Epidemiology Unit, Oce for Research and Development, Faculty of Medicine Siriraj Hospital, Mahidol
University, Bangkok, ailand.
Comparison of an Automated Thermodynamic
Treatment System (LipiFlow) and Warm
Compresses for the Treatment of Moderate
Severity of Meibomian Gland Dysfunction
ABSTRACT
Objective: To compare the ecacies of a single thermodynamic treatment system (LipiFlow®) and warm compresses
used for 3 months in patients with a moderate severity of meibomian gland dysfunction (MGD).
Methods: is prospective, randomized, controlled clinical study enrolled 28 patients (22 females, 6 males; mean
age, 53.9 ± 14.8 years) diagnosed as having moderate MGD by plugging at the meibomian gland orices between
one-third and two-thirds of lid margins and at least one of the following: a Standard Patient Evaluation of Eye
Dryness (SPEED) questionnaire score of 6-12; a lipid layer thickness (LLT) score of 40-70 Interferometric Color
Units; upper eyelid meiboscore of 1-2; and 3-6 expressible meibomian glands (EMGs) in the lower eyelid. Both
eyes of each patient were randomized into study and control eyes. Study eyes were treated with a single, 12-minute
LipiFlow® system, while control eyes received 5-minute warm compresses twice daily for 3 months. e dry eye
symptoms, the number of EMGs, and LLTs were evaluated.
Results: ere were no signicant dierences in the dry eye symptoms, number of EMGs, and LLTs for both groups
at baseline and at each follow-up. However, the total SPEED scores for the LipiFlow® group reduced signicantly
from baseline at each follow-up until 6 months. As to the warm compress group, the total SPEED scores reduced
signicantly from baseline at each follow-up until 3 months.
Conclusion: e single LipiFlow® treatment and twice-daily warm compresses relieved the dry eye symptoms of
patients with a moderate severity of MGD compared with their baseline symptoms, despite no statistical dierences
in the dry eye symptoms, number of EMGs, and LLTs of both treatments.
Keywords: Meibomian gland dysfunction; warm compresses; automated thermodynamic treatment system;
LipiFlow® (Siriraj Med J 2020; 72: 79-86)
Corresponding author: Wipawee Booranapong
E-mail: wipawee.boo@mahidol.edu
Received 25 March 2019 Revised 22 July 2019 Accepted 7 August 2019
ORCID ID: http://orcid.org/0000-0002-8812-0319
http://dx.doi.org/10.33192/Smj.2020.11
INTRODUCTION
Meibomian gland dysfunction (MGD) is a chronic,
diuse, terminal duct obstruction of the Meibomian glands
(MGs) involving qualitative or quantitative changes in
glandular secretions.
1
It results in an alteration of the
tear lm and causes the majority of evaporative dry eyes,
ranging from 38% to 68% of the population.
2
Patients
with MGD experience eye irritation, a burning sensation,
diculty opening the eyelids, and blurred vision. e
management of MGD includes the replacement of the MG
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
80
secretion with topical oil emulsion formulations, the use
of ocular ointments, the release of the MG obstruction,
and the restoration of normal MG secretion. Current
methods to relieve MG obstruction include eyelid margin
cleansing, eyelid warming and eyelid massaging twice
daily
3,4
and practitioner-administered manual expression.
5
However, it is dicult to maintain warm compresses
at a constant temperature between 32ºC and 40ºC (the
reported melting temperature range of normal MG
secretions) for 5 minutes.
6
Moreover, patients nd that
the twice-daily administration of an eective warm
compress is both inconvenient and time-consuming,
resulting in poor compliance. Although practitioner-
administered manual expressions of MGs are more
eective at releasing MG obstructions, their application
is limited due to the associated pain. A new treatment
modality, the LipiFlow® Thermal Pulsation System
(TearScience Inc., Morrisville, NC, USA), may help to
solve the limited success rate of warm compresses. is
technology involves a single, 12-minute treatment that
allows heat to be applied to the palpebral surfaces of the
upper and lower eyelids directly over the MGs while
simultaneously applying graded, pulsatile pressure to the
cutaneous eyelid surfaces, thereby expressing the MGs
during heating.
7-9
Many studies have shown that a single
LipiFlow® treatment results in sustained improvement
in both the signs and symptoms of MGD at 1 month,
10-12
3 months,
9,13-15
6 months,
16
9 months
17
and 1 year post-
treatment.
18-19
However, Finis and colleagues found that
this single LipiFlow® treatment poorly beneted patients
with severe MGD or MG atrophy.
16
Because of the high
cost of this single LipiFlow® treatment, it is essential to
identify potential non-responders or poor responders
before treatment. To our knowledge, there has been no
prospective, randomized, controlled clinical study of the
ecacy of this treatment in patients with a moderate
severity of MGD. us, this study evaluated the ecacy
of this new treatment and compared it with the twice-
daily application of warm compresses for patients with
moderate MGD.
MATERIALS AND METHODS
is prospective, observer-masked, randomized,
controlled clinical trial was conducted at Siriraj Hospital,
Faculty of Medicine Siriraj Hospital, Mahidol University,
ailand, between January 2015 and June 2016. e
study was approved by the Institutional Review Board
and followed the tenets of the Declaration of Helsinki
for research involving human subjects (Si 725/2014; ai
clinical trial registration number: TCTR 20170905001).
Written, informed consent was obtained from all patients
before enrollment. e inclusion criteria were patients
at least 18 years of age who had a moderate severity of
MG obstruction, dened in this study as follows: the
presence of plugging at the MG orices between one-
third and two-thirds of total lid margin area and at least
one of the following: (1) a Standard Patient Evaluation
of Eye Dryness (SPEED) questionnaire score of 6-12
15,20-21
;
(2) a lipid layer thickness (LLT) score of 40-70 ICU
(Interferometric Color Units), measured by using a
LipiView® interferometer (TearScience Inc., Morrisville,
NC, USA)
22
; (3) an upper eyelid meiboscore of 1-2,
obtained by using a Meibography® (Oculus, Wetzlar,
Germany), using the grading scales developed by Arita
et al.
23
; and (4) 3-6 expressible MGs (EMGs) in the
lower eyelid.
15,21
e exclusion criteria were patients
with systemic diseases resulting in dry eyes (such as
systemic lupus erythematosus, rheumatoid arthritis, and
allergic diseases); patients who had eyelid abnormalities,
previous ocular surgery, or trauma; and patients who
used systemic medications aecting dry eyes (for instance,
antihistamines, tetracycline derivatives, isotretinoin,
and topical cyclosporine A) or steroids in the preceding
month.
All patients completed the SPEED questionnaires by
themselves. en, the MGs of all patients were evaluated
using a Meibography® and LLTs were measured using
a LipiView® interferometer. Both the meiboscores and
LLTs were measured by one examiner (P. Chonpimai).
Complete ocular examinations of the anterior and posterior
segments, including an assessment of the expression of the
MGs of the lower eyelids of both eyes using a Meibomian
Gland Evaluator (TearScience Inc., Morrisville, NC,
USA), were performed by cornea specialists (W.B., P.P.,
S.C., and P.K. and two cornea fellows (P.N. and W.T.).
Dry eye symptoms were recorded as total SPEED
scores, derived from the sum of the frequency and severity
scores for all symptoms over a range from 0 to 28. A
higher score represented more frequent and/or more
severe symptoms.
20,24
e number of EMGs was quantied, as described
by Korb and Blackie.
21
e MGs of the lower eyelids
were expressed by the Meibomian Gland Evaluator,
being pressed with approximately 1.2 g/mm
2
of pressure
along the distal end of the MGs at the nasal, central, and
temporal parts of the lower eyelids. Each part consisted
of 5 consecutive MG orices, which meant that a total
of 15 MGs were expressed. e number of EMGs was
recorded.
e LLTs were ascertained by using a LipiView®
interferometer, as described by Blackie et al.
22
e patients’
tear lm interference patterns were recorded and analyzed
Booranapong et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
81
Original Article
SMJ
as a value of interferometric color units (ICU; 1 ICU
refers to approximately a 1-nm thickness of lipid layer).
e right and le eyes of each patient were randomized
into a study eye and a contralateral control eye. e
study eye was treated with a single, 12-minute, LipiFlow®
ermal Pulsation System, whereas the control eye was
treated with a 5-minute warm compress twice daily
for 3 months. e LipiFlow® treatment was utilized as
described in detail by Lane et al.
11
Briey, two drops
of a topical anesthetic (0.5% tetracaine hydrochloride)
were applied to the study eye prior to treatment. e
LipiFlow® device applied heat (42.5ºC) to the upper and
lower inner eyelid surfaces directly over the MGs, while
a pulsating pressure was simultaneously applied to the
outer eyelids using an inatable air bladder.
11
is device
was capable of melting the MG contents without thermal
injury, and of simultaneously evacuating the MGs of the
upper and lower eyelids. e control eye was treated
with warm compresses. e patients were instructed
to soak a hand towel or washcloth in hot water, wring
out the excess water, and then place the warm towel or
cloth over the skin of the closed eyelids of the control
eye, applying gentle pressure for 5 minutes.
6
e study parameters were the dry eye symptoms
determined by using the total SPEED scores and an
assessment of the MG function, measured by the number
of EMGs and the LLT scores. ese three parameters
were evaluated at baseline and at the follow-up times of
1 day, 1 week, 1month, 6 weeks, 3 months and 6 months.
Articial tear eye drops (0.18% sodium hyaluronate)
were prescribed for application into both eyes every 2
hours for all patients during the study.
Sample size
To calculate the sample size, a power analysis and the
eect size using the data of Lane et al.,
11
which showed the
improvements in the number of EMGs, were performed.
On this basis, the minimum sample size was 20 patients
per group. However, to ensure adequate reliability and
to compensate for patient loss, 30 patients per group
were enrolled.
Statistical analysis
Descriptive statistics were used to demonstrate the
patients’ baseline characteristics. e categorical data
were described as numbers with percentages, while the
continuous data were presented as mean with standard
deviation (SD) or mean dierences with the corresponding
95% condence interval. e unpaired t-test was utilized
to compare the continuous data of the LipiFlow® and
control groups, whereas the paired t-test was used to
compare the continuous data within the groups. All
analyses were performed using PASW Statistics for
Windows, version 18 (SPSS Inc., Chicago, IL, USA). All
tests of signicance were two tailed, and a p-value<0.05
was considered statistically signicant.
RESULTS
A total of 33 patients were enrolled. However,
only 28 completed the study because 3 patients were
lost to follow-up aer the baseline examination, and
2 patients did not return for the 6-month follow-up.
Twenty-two patients were female (78.6%), and six were
male (21.4%). e mean age was 53.9±14.8 years. ere
were no statistical dierences in the meiboscores of
the LipiFlow® and warm compress groups at baseline
(p=0.745).
As to the total SPEED scores, there were no statistically
signicant dierences between the LipiFlow® and the
warm compress groups at baseline and at each follow-up
(Table 1). However, in the case of the LipiFlow® group,
there was a statistically signicant, greater reduction in
the total SPEED scores at each follow-up until 6 months,
compared to the scores at baseline (Table 1). Similarly, the
warm compress group showed a statistically signicant,
greater reduction in the total SPEED scores at the 1-week,
1-month, 6-week and 3-month follow-ups, compared
to those at baseline (Table 1). However, aer stopping
the warm compresses at the 3-month follow-up, the
total SPEED scores for the warm compress group were
not changed signicantly at the 6-month follow-up,
compared to those at baseline (Table 1). In other words,
the total SPEED scores for the warm compress group
returned to near-baseline level at the 6-month follow-up
aer the application of the warm compresses ceased at
the 3-month time point.
In a comparison of the number of EMGs of the two
groups, the LipiFlow® group showed a greater increase
in the number of EMGs than the warm compress group
at every follow-up. Nevertheless, those greater increase
were statistically signicant on only two occasions: at
the 1-day and the 3-month follow-ups (p=0.048 and
p=0.049, respectively; Table 2).
Regarding the LLTs, there were no statistically
signicant dierences in the LLTs of both groups at
baseline and at each follow-up (Table 3). However, the
warm compress group had a statistically signicant greater
reduction in the LLT scores at the 1-day and 1-week
follow-ups, compared to those of baseline (p=0.004 and
p=0.031, respectively; Table 3).
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
82
TABLE 1. Comparison of the dry eye symptoms (SPEED scores) for both groups and between baseline and at
each follow-up for each group.
SPEED scores (mean ± SD)
Warm compresses LipiFlow
®
P-value
a
Baseline 9.2 ± 4.4 9.5 ± 4.5 0.781
Follow-up time
1 day 7.4 ± 5.2 7.8 ± 4.2 0.531
∆
1 day -1.6 ± 4.4 -1.8 ± 2.9
p-value
b
0.062 0.003
1 week 7.4 ± 4.7 7.1 ± 4.5 0.924
∆
1 week -2.1 ± 4.0 -2.7 ± 3.8
p-value
b
0.007 0.001
1 month 6.3 ± 4.6 5.8 ± 4.1 0.731
∆
4 weeks -2.5 ± 4.1 -3.3 ± 3.6
p-value
b
0.006 <0.001
6 weeks 6.7 ± 4.7 6.5 ± 4.2 0.777
∆
6 weeks -2.3 ± 5.2 -3.0 ± 4.5
p-value
b
0.035 0.003
3 months 6.7 ± 4.4 6.7 ± 4.4 1.000
∆
3 months -2.7 ± 4.2 -2.7 ± 3.5
p-value
b
0.004 0.001
6 months 7.9 ± 5.8 7.8 ± 5.3 0.954
∆
6 months -1.3 ± 5.1 -1.9 ± 4.7
p-value
b
0.238 0.034
a
comparison of SPEED scores for both groups,
b
comparison of SPEED scores between baseline and at each follow-up for each group,
∆
= mean change in SPEED scores from baseline to each follow-up for each group
Adverse events
ere were no serious, device-related, adverse events
in the LipiFlow® group, and no adverse events in the
warm compress group. e only device-related, adverse
event was eye discomfort or pain, which occurred among
patients with small eyes, narrow palpebral ssure, or
deep-set eyes.
DISCUSSION
e present study showed that both the single,
12-minute, thermodynamic treatment system (LipiFlow®)
and the twice-daily warm compresses were able to relieve
the dry eye symptoms of patients with a moderate MGD
compared to their baseline symptoms, despite there
being no statistical dierences in the dry eye symptoms
for both treatments. In fact, the dry eye symptoms of the
LipiFlow® group decreased faster at the rst day post-
treatment, and they continued to decrease to 6 months
post-treatment. In contrast, the dry eye symptoms of
the warm compress group began to reduce later, at the
one-week follow-up, and sustained the reduction only
while the participants applied the warm compresses
twice daily for 3 months. When they stopped the warm
compresses, their symptoms reappeared. Many studies
9- 19
have reported that a single LipiFlow® treatment had
a statistically signicant greater reduction in dry eye
symptoms than warm compresses at 1 month until 1
year post-treatment. In contrast to those ndings, the
Booranapong et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
83
Original Article
SMJ
The number of EMGs (mean ± SD)
Warm compresses LipiFlow
®
P-value
a
Baseline
Follow-up time 2.5 ± 2.3 2.6 ± 2.8 0.885
1 day 2.0 ± 2.0 3.0 ± 2.0 0.048
∆
mean (95%CI) -0.7 (-1.6, 0.2) 0.3 (-1.2, 1.7)
p-value
b
0.156 0.536
1 week 2.5 ± 2.1 3.7 ± 3.3 0.153
∆
mean (95%CI) -0.1 (-1.1, 1.0) 1.3 (-0.5, 3.0)
p-value
b
0.951 0.152
1 month 2.6 ± 2.2 3.5 ± 3.1 0.471
∆
mean (95%CI) 0.1 (-1.1, 1.3) 0.7 (-0.8, 2.2)
p-value
b
0.847 0.352
6 weeks 2.3 ± 2.4 2.4 ± 2.3 0.827
∆
mean (95%CI) -0.4 (-1.7, 0.9) -0.6 (-2.1, 1.0)
p-value
b
0.556 0.660
3 months 2.3 ± 2.0 3.6 ± 2.2 0.049
∆
mean (95%CI) -0.3 (-1.4, 0.8) 0.8 (-0.5, 2.0)
p-value
b
0.494 0.066
6 months 2.7 ± 2.2 3.1 ± 2.1 0.369
∆
mean (95%CI) 0 (-1.2, 1.2) 0.3 (-1.1, 1.6)
p-value
b
0.862 0.557
TABLE 2. Comparison of the number of expressible meibomian glands (EMGs) for both groups and between
baseline and at each follow-up for each group.
a
comparison of the number of EMGs for both groups,
b
comparison of the number of EMGs between baseline and at each follow-up for each
group, ∆ mean (95%CI) = mean change in the number of EMGs from baseline to each follow-up for each group (95% condence interval)
results of the present study were the same as those of
other reports
16,25
that did not demonstrate a statistically
signicant dierence in the dry eye symptoms between
the single LipiFlow® treatment and the twice-daily warm
compresses, nor in the number of EMGs and LLTs, despite
a tendency to improve those three parameters in the
patients receiving the LipiFlow® treatment. e current
study could not demonstrate a statistical dierence in
the dry eye symptoms of both treatments, and the mean
SPEED scores found in this study were less than those
reported by other researches.
8-12,17-18
is may be due to
the small sample size and the enrollment of only patients
with a moderate MGD.
Concerning the MG function, the number of EMGs
and LLT scores were evaluated. e number of EMGs
was derived from the number of MGs in which the
obstructed meibum had been melted and was able to
be squeezed through the opening of the MG ducts.
is implied that the MG obstruction was relieved by
melting the stagnated lipid of MGs, which subsequently
owed through the ducts and passed out their openings.
Consequently, LLTs should be increased as per a report
26
that showed a signicant correlation between EMGs and
LLTs. Although the present study found that the LipiFlow®
treatment resulted in a greater number of EMGs than
the warm compresses at each follow-up, those greater
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
84
TABLE 3. Comparison of lipid layer thicknesses (LLTs) for both groups and between baseline and at each follow-
up for each group.
LLT scores (mean ± SD)
Warm compresses LipiFlow
®
P-value
a
Baseline
Follow-up time 74.9 ± 18.8 70.3 ± 19.9 0.368
1 day 65.4 ± 19.6 65.5 ± 19.0 0.984
∆
mean (95%CI) -9.4 (-15.6, -3.2) -6.4 (-13.1, 0.3)
p-value
b
0.004 0.062
1 week 69.0 ± 18.8 65.5 ± 15.9 0.491
∆
mean (95%CI) -8.5 (-16.1, -0.9) -5.5 (-15.0, 3.9)
p-value
b
0.031 0.238
1 month 65.0 ± 18.4 69.7 ± 19.9 0.520
∆
mean (95%CI) -7.6 (-15.7, 0.4) -2.4 (-12.7, 8.0)
p-value
b
0.062 0.630
6 weeks 73.8 ± 18.0 75.7 ± 16.7 0.798
∆
mean (95%CI) -1.3 (-15.0, 12.3) -1.4 (-7.4, 4.6)
p-value
b
0.834 0.613
3 months 80.6 ± 17.8 76.4 ± 13.3 0.491
∆
mean (95%CI) 1.9 (-6.8, 10.7) 5.6 (-5.3, 16.4)
p-value
b
0.642 0.286
6 months 67.1 ± 23.0 66.9 ± 21.2 0.979
∆
mean (95%CI) -8.0 (-16.9, 0.8) -2.2 (-12.3, 8.0)
p-value
b
0.073 0.662
a
comparison of LLT scores for both groups,
b
comparison of LLT scores between baseline and at each follow-up for each group, ∆ mean
(95%CI) = mean change in LLT scores from baseline to each follow-up for each group (95% condence interval)
numbers were only statistically signicant at the 1-day
and 3-month follow-ups. In other words, our study
demonstrated that the automated LipiFlow® began to be
eective at an earlier time than the warm compresses. As
the result of releasing the MG obstruction, LLTs should
be evaluated. e current study did not show a statistically
signicant dierence in the LLTs of both groups at
baseline and at each follow-up, which is dierent from
the ndings of previous studies.
16,27
Although the LLT
scores post-LipiFlow® treatment increased insignicantly
from baseline, the LLT scores for the warm compress
group surprisingly decreased signicantly at the 1-day
and 1-week follow-ups. is may be due to the small
sample size, diculties in controlling the temperature of
the warm compresses, or the enrollment of only patients
with a moderate MGD. With that MGD severity, the
obstructed meibum needs a higher temperature than
the heat provided in a recent program of LipiFlow® and
the warm compresses in order to achieve the melting
point of the obstructed meibum for severely-aected
patients, as reported by Bron et al.
6
and Finis et al.
16
Moreover, the application of the warm compresses needs
a great deal of attention to the procedures described in
previous studies
3,4
to achieve constant and sucient
heat, and the warm compresses hardly provide a high
enough degree of heat to melt the obstructed meibum
in patients with moderate to severe MGD. is may
explain the surprising result of the signicant decrease
Booranapong et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
85
Original Article
SMJ
in the LLT scores at the 1-day and 1-week follow-ups
for the warm compress group. In addition to inadequate
temperatures, the formulation of the lipid or meibum
in cases of moderate MGD is not normal lipid.
6
us, it
cannot be easily expressed through the MG duct to its
opening to increase the LLT. Even though the results of
our study could not show that the eectiveness of the
single LipiFlow® treatment was signicantly superior to
that of the warm compresses, the dry eye symptoms and
the number of EMGs of the patients with moderate MGD
tended to improve better when they were treated with the
LipiFlow® system. e advantages of a single, 12-minute,
automated thermodynamic LipiFlow® treatment are that
it is less painful than manually expressing the MGs, less
time-consuming than the warm compresses, and convenient
due to the absence of any self-administered treatment,
Moreover, a single treatment has been reported to have a
prolonged eect by easing MGD for at least the following
6 months and for up to 3 years.
16-19,28
Nevertheless, the
LipiFlow® treatment has several disadvantages: its high
cost; the LipiFlow® devices’ use of too large an eye cup
and lid warmer parts for Asian eyes; and its potentially
lower eectiveness for moderate and severe degrees of
MGD or MG atrophy, as reported by earlier studies.
16,25,29
e limitations of this study were its small sample
size; a high female to male ratio; the enrollment of only
patients with a moderate MGD; no assessment of the tear
break-up time; no ocular surface staining; no Schirmer
test and an incapacity to monitor the temperature of the
hot water and the warm compress procedures performed
by the patients at home. ere were also disadvantages
in having the study eye and the contralateral control eye
in the same patient, such as the possibility of an inter-
eye interaction and an inability to mask the patients
to the specic treatment given to each eye. is may
have had some impact on the subjective results, such as
the SPEED scores. Since the patients with a moderate
severity of MGD had obvious symptoms of dryness, a
long follow-up duration of 6 months without any tear
supplements might have caused corneal complications.
However, the application of the articial tear eye drops
every two hours by the patients in both groups could have
improved some of the symptoms of dry eyes, resulting
in better SPEED scores being reported by both groups.
On the other hand, the advantages of the present
study were that it was a prospective, observer-masked,
randomized, controlled clinical trial comparing both
treatments in the same patient, which enabled control
of the internal factors of both trial groups and provided
a long follow-up time of 6 months.
CONCLUSION
e single LipiFlow® treatment and the twice-daily
warm compresses relieved the dry eye symptoms of the
patients with moderate MGD compared with their baseline
symptoms, despite there being no statistical dierences
in the dry eye symptoms, the number of EMGs, or the
LLTs of both treatments.
Further studies should be conducted with a larger
sample size and investigated by classifying the patients
as having mild, moderate and severe degrees of MGD.
In the future, the LipiFlow® system should develop a new
program for the treatment of moderate and severe degrees
of MGD by slightly increasing the heat or increasing
the duration of the treatment or its repetition during a
specic time period. Moreover, the eye cup and lid warmer
parts of the LipiFlow® system should be available in an
optional, smaller size for Asian eyes in order to apply
an adequate amount of heat and eectively massage the
eyelids.
ACKNOWLEDGMENTS
e authors thank Miss Pratuangsri Chonpimai
for her assistance in measuring the meiboscores using
the Meibography® and the lipid layer thicknesses using
the LipiView® interferometer. e authors are also very
grateful to the Faculty of Medicine Siriraj Hospital,
Mahidol University, for supporting this study.
REFERENCES
1. Nichols KK, Foulks GN, Bron AJ, Glasgow BJ, Dogru M, Tsubota
K, et al. e international workshop on meibomian gland
dysfunction: executive summary. Invest Ophthalmol Vis Sci
2011;52:1922-9.
2. Stapleton F, Alves M, Bunya VY, Jalbert I, Lekhanont K, Malet
F, et al. TFOS DEWS II epidemiology report. Ocul Surf 2017;
15:334-65.
3. Goto E, Endo K, Suzuki A, Fujikura Y, Tsubota K. Improvement
of tear stability following warm compression in patients with
meibomian gland dysfunction. Adv Exp Med Biol 2002;506:
1149-52.
4. Olson MC, Korb DR, Greiner JV. Increase in tear lm lipid
layer thickness following treatment with warm compresses
in patients with meibomian gland dysfunction. Eye Contact
Lens 2003;29:96-9.
5. Korb DR, Greiner JV. Increase in tear lm lipid layer thickness
following treatment of meibomian gland dysfunction.Adv
Exp Med Biol 1994;350:293-8.
6. Bron AJ, Tiany JM, Gouveia SM, Yokoi N, Voon LW. Functional
aspects of the tear lm lipid layer. Exp Eye Res 2004;78:347-60.
7. Korb DR, Blackie CA. Restoration of meibomian gland
functionality with novel thermodynamic treatment device: a
case report. Cornea 2010;29:930-3.
8. Majmudar PA; LipiFlow Study Group. A Novel Thermal
pulsation treatment for obstructive meibomian gland dysfunction:
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
86
applying heat to the inner eyelid surfaces. Invest Ophthalmol
& Vis Sci 2010;51:6281.
9. Friedland BR, Fleming CP, Blackie CA, Korb DR. A novel
thermodynamic treatment for meibomian gland dysfunction.
Curr Eye Res 2011;36:79-87.
10. Greiner JV; LipiFlow Study Group. Treatment of meibomian
gland dysfunction (MGD) with the novel LipiFlow® thermal
pulsation system restores meibomian gland function. Invest
Ophthalmol & Vis Sci 2010;51:6282.
11. Lane SS, DuBiner HB, Epstein RJ, Ernest PH, Greiner JV,
Hardten DR, et al. A new system, the LipiFlow, for the treatment
of meibomian gland dysfunction. Cornea 2012;31:396-404.
12. Korb DR, Blackie CA. Case report: a successful LipiFlow
treatment of a single case of meibomian gland dysfunction
and dropout. Eye Contact Lens 2013;39:e1-e3.
13. Finis D, Hayajneh J, König C, Borrelli M, Schrader S, Geerling
G. Evaluation of an automated thermodynamic treatment
(LipiFlow®) system for meibomian gland dysfunction: a prospective,
randomized, observer-masked trial. Ocul Surf 2014;12:146-54.
14. Blackie CA, Coleman CA, Nichols KK, Jones L, Chen PQ,
Melton R, et al. A single vectored thermal pulsation treatment
for meibomian gland dysfunction increases mean comfortable
contact lens wearing time by approximately 4 hours per day.
Clin Ophthalmol 2018;12:169-83.
15. Hagen KB, Bedi R, Blackie CA, Christenson-Akagi KJ. Comparison
of a single-dose vectored thermal pulsation procedure with
a 3-month course of daily oral doxycycline for moderate-to-
severe meibomian gland dysfunction. Clin Ophthalmol
2018;12:161-8.
16. Finis D, König C, Hayajneh J, Borrelli M, Schrader S, Geerling
G. Six-month eects of a thermodynamic treatment for MGD
and implications of meibomian gland atrophy. Cornea 2014;
33:1265-70.
17. Greiner JV. A single LipiFlow® thermal pulsation system
treatment improves meibomian gland function and reduces
dry eye symptoms for 9 months. Curr Eye Res 2012;37:272-8.
18. Greiner JV. Long-term (12-month) improvement in meibomian
gland function and reduced dry eye symptoms with a single
thermal pulsation treatment. Clin Experiment Ophthalmol
2013;41:524-30.
19. Blackie CA, Coleman CA, Holland EJ. e sustained eect
(12 months) of a single-dose vectored thermal pulsation
procedure for meibomian gland dysfunction and evaporative
dry eye. Clin Ophthalmol 2016;10:1385-96.
20. Ngo W, Situ P, Keir N, Korb D, Blackie C, Simpson T, et al.
Psychometric properties and validation of the standard patient
evaluation of eye dryness questionnaire. Cornea 2013;32:1204-
10.
21. Korb DR, Blackie CA. Meibomian gland diagnostic expressibility:
correlation with dry eye symptoms and gland location. Cornea
2008;27:1142-7.
22. Blackie CA, Solomon JD, Scadi RC, Greiner JV, Lemp MA,
Korb DR. e relationship between dry eye symptoms and
lipid layer thickness. Cornea 2009;28:789-94.
23. Arita R, Itoh K, Inoue K, Amano S. Noncontact infrared
meibography to document age-related changes of the meibomian
glands in a normal population. Ophthalmology 2008; 115:
911-5.
24. Korb DR, Herman JP, Greiner JV, Scadi RC, Finnemore VM,
Exford JM, et al. Lid wiper epitheliopathy and dry eye symptoms.
Eye Contact Lens 2005;31:2-8.
25. Zhao Y, Veerappan A, Yeo S, Rooney DM, Acharya RU, Tan
JH, et al. Clinical trial of thermal pulsation (LipiFlow) in
meibomian gland dysfunction with preteatment meibography.
Eye Contact Lens 2016;42:339-46.
26. Finis D, Pischel N, Schrader S, Geerling G. Evaluation of lipid
layer thickness measurement of the tear lm as a diagnostic
tool for meibomian gland dysfunction. Cornea 2013;32:1549-
53.
27. Michee S, Rabut G, Baudouin C, Labbe A. Treatment of
meibomian gland disease with the Lipiow® system: A prospective
study. Acta Ophthalmol 2013; 91:0.
28. Greiner JV. Long-term (3 year) eects of a single LipiFlow
thermal pulsation system treatment on meibomian gland
function and dry eye symptoms. Eye Contact Lens 2016; 42:
99-107.
29. Satjawatcharaphong P, Ge S, Lin MC. Clinical outcomes
associated with thermal pulsation system treatment. Optom
Vis Sci 2015;92:e334-41.
Booranapong et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
87
Review Article
SMJ
Lee Kian Khoo, D.D.S.*,**, Sutichai Norranitchaiyakul, M.D., D.D.S.**, Suphachai Suphangul, D.D.S**.,
Natthamet Wongsirichat, B.Sc., D.D.S.**,***
*Faculty of Dentistry, Vinayaka Missions University, Ariyanur, Salem, Tamil, Nadu, India, **Department of Oral & Maxillofacial Surgery, Faculty
of Dentistry, Mahidol University, Bangkok, ***Consultant of Walailak University International College of Dentistry Phahonyothin Road, Phaya ai
District, Bangkok 10400, ailand.
Overview of Surgical Therapies for Obstructive
Sleep Apnea: a Concise Review Literature
ABSTRACT
e surgical therapy of Obstructive Sleep Apnea (OSA) requires addressing anatomical obstructions or collapse
of the pharyngeal airway by skeletal and so tissue surgeries. Numerous surgical options have been documented for
OSA therapy with varying success. OSA surgery is vital when patients refuse continuous positive airway pressure
(CPAP). e aim of this article is to review surgeries for OSA and the eectiveness of each surgery in terms of
Epworth sleepiness scale (ESS), Apnea Hypopnea Index (AHI) or Respiratory Disturbance Index (RDI) reduction.
Keywords: Obstructive sleep apnea; continuous positive airway pressure; epworth sleepiness scale; apnea hypopnea
index; respiratory disturbance index (Siriraj Med J 2020; 72: 87-94)
Corresponding author: Suphachai Suphangul
E-mail: kungomfs@gmail.com
Received 10 April 2019 Revised 17 July 2019 Accepted 7 August 2019
ORCID ID: http://orcid.org/0000-0002-1908-2398
http://dx.doi.org/10.33192/Smj.2020.12
INTRODUCTION
Obstructive sleep apnea (OSA) is a medical condition
characterized by episodes of partial (hypopnea) or complete
(apnea) constriction of the upper airway during sleep.
Patients with untreated OSA are usually aicted by
excessive daytime sleepiness (EDS), which is a symptom
that has been associated with an increased risk of motor-
vehicle
1
and workplace accidents.
2
Moreover, several
reports have linked OSA with various cardiovascular
events such as stroke,
2,3
myocardial infarction,
4
and
hypertension.
5
Nowadays, the usual rst-line therapy
for OSA is continuous positive airway pressure (CPAP).
Although CPAP is still regarded as the gold-standard
therapy for OSA the non-adherence rate of patients to
CPAP has been documented to be as high as 34.1% (based
on studies over a twenty-year time frame).
6
Consequently,
CPAP non adherent patients are then managed by surgery
either to increase CPAP compliance
7
or to eectively
reduce the Apnea Hypopnea Index (AHI) or Respiratory
Disturbance Index (RDI) to a level that alleviates the
need for CPAP.
8
However, each surgical modality comes
with distinct pros and cons, and it is up to the clinician
to weigh them for the patient’s best interest. is article
serves to review the surgical treatments for OSA.
Diagnosis of OSA
Polysomnograhy (PSG) is the most accurate
method for diagnosing the presence of OSA, and the
severity of OSA is usually based on the Apnea Hypopnea
Index(AHI): Mild OSA (5-14.9), Moderate OSA (15-
29.9) and Severe OSA(≥30).
9
The PSG is incapable
of determining the exact anatomical location of the
obstruction or collapse. Anatomical obstructions could
be well demarcated deformities of the so tissues that
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
88
TABLE 1. e abbreviations were used in this concise review literature.
Full name word Abbreviation
Obstructive sleep apnea OSA
Excessive daytime sleepiness EDS
Continuous positive airway pressure CPAP
Apnea hypopnea index AHI
Respiratory disturbance index RDI
Polysomnograhy PSG
Drug induced sleep endoscopy DISE
Velum, oropharynx lateral wall, tongue base and epiglottis VOTE
Uvulopalatopharyngoplasty UPPP
Genioglossus advancement GGA
Laser assisted uvulopalatoplasty LAUP
Radiofrequencyablation RFA
Radiofrequency RF
Maxillomandibular advancement MMA
Bilateral sagittal split ramus osteotomy BSSRO
Hypoglossal nerve stimulation HGNS
Transoral oral robotic surgery TORS
can be visualized through physical examinations, ber
optic pharyngoscopy and lateral cephalograms. ese
obstructions generally remain constant, regardless of
whether the patient is awake or sleeping.
10
Some forms of
obstructions or collapse occur as a result of a narrowing
of certain anatomical sites such as the lateral pharyngeal
wall or tongue during sleep.
10
Complete concentric
collapse of the velum during sleep is one of the main
determinants of failure of OSA surgery.
11
erefore,
drug induced sleep endoscopy (DISE) has become vital
for selecting the right treatment plan in OSA cases.
12
e main anatomical structures that are visualized and
evaluated via DISE are the Velum, Oropharynx lateral
wall, Tongue base and Epiglottis (VOTE).
12
Besides aiding
in diagnosis, DISE post-operation gives an objective and
dynamic visualization of the airway, and provides a key
marker for success which is the stability of the lateral
pharyngeal wall.
13
Tracheostomy
e tracheostomy as a procedure for resolving
hypoventilation was rst proposed in 1965 by Valero and
Alroy, and the case report featured a 55 year old male
with chronic traumatic micrognathia who complained of
excessive sleepiness which improved aer tracheostomy.
14
Multiple preceding reports up until the early of 1980s
rearmed the tracheostomy as the surgery of choice
for OSA, particularly in morbidly obese patients.
15-17
However, the complications of tracheostomy include
minor hemorrhage, cu leakage, and in severe cases
tube obstruction which could lead to death.
18
ere is
an inevitable reduction in the quality of life of patients
following tracheostomy due to speech problems, body
image issues, and daily physical limitations.
19
erefore,
the American Academy of Sleep Medicine recommended
that tracheostomy should only be used when all other
treatment options have been exhausted and failed or
when clinically urgent.
20
Khoo et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
89
Review Article
SMJ
Nasal surgeries
e consensus on nasal surgery as a stand-alone
procedure for OSA is that it improves daytime sleepiness
and snoring, but does not reduce AHI scores.
21
Nevertheless,
OSA patients with increased nasal resistance from
hypertrophic inferior turbinate and deviated septum are
oen non-adherent to CPAP.
22
Nasal surgeries increase
the nasal airow and reduce the pressure requirements
on CPAP therefore improving CPAP adherence.
21,22
Nakata et al.,
22
found significant reduction of nasal
resistance measured via Rhinomanometry aer inferior
turbinectomy and submucus resection of nasal septum
was done on CPAP non-compliant patients. ESS scores
post-op improved from 11.7±4.1 to 3.3 ±1.3.
Uvulopalatopharyngoplasty
e Uvulopalatopharyngoplasty (UPPP) is a surgery
that was rst developed by Ikematsu in the 1950’s for the
reduction of snoring, his ndings were later published
in 1964.
23
In 1981 Fujita et al.,
24
introduced the UPPP
technique (Fig 1) to English-speaking surgeons.
visible when tongue is in a neutral position) cases that
had UPPP.
28
Genioglossus advancement with/without hyoid
myotomy
Riley et al., rst documented success with genioglossus
advancement (GGA) and hyoid myotomy as part of
a multi-step surgical protocol. GGA was done via a
genial tubercle rectangular window which was advanced
and rotated before xation whereas hyoid myotomy
involved the suspension of the hyoid bone over the thyroid
cartilage.
29
Neruntarat replicated the same technique under
local anesthesia on 31 OSA patients with hypopharynx
obstruction as diagnosed by polysomnography and
awake nasopharyngoscopy, and the RDI improved from
48.2±10.8 to 14.5±5.8.
30
Both studies utilized additional
UPPP for the correction of retropalatal obstructions.
In 2017 Vargo et al.,
31
documented recently GGA as a
stand-alone surgery for OSA with success in terms of
mean AHI reduction (36/h to 4.3/h). Vargo et al., also
advocated a sliding genioplasty (Fig 2) as opposed to genial
tubercle window in isolated GA cases for better cosmetic
outcomes and the possibility of larger advancements
with glossoplexy sutures.
31
Fig 1. Fujita Uvulopalatopharyngoplasty.
ereaer, the UPPP became a common procedure
despite the lack of evidence supporting its ecacy especially
as a stand-alone surgery in the treatment of moderate to
severe OSA.
20
Sher et al., concluded in a meta-analysis
that only 40.79% of patients who had undergone UPPP
exhibited success in terms of AHI reduction of 50 %
or AHI less than 20.
25
e UPPP entails the widening
of the oropharynx space by means of shortening the
uvula thus, long-term side eects include dysphagia,
nasal regurgitation, dysphonia, and throat pain.
26
When
a staging system that is based on the anatomy of the
palate is used, the success of the UPPP could be more
predictable.
27
Friedman et al., had success in 80.6%
of Friedman stage 1 (entire uvula, tonsils or pillar are
Fig 2. Sliding Genioplasty. e genial osteotomy should involve the
anterior inferior border of the mandible to advance the genioglossus
and suprahyoid muscles.
Laser assisted uvulopalatoplasty
Laser assisted uvulopalatoplasty (LAUP) is an
outpatient procedure for OSA therapy that could be
performed under local anesthesia without the risk of
hemorrhage. Kamami
32
was the rst proponent of the
LAUP as a treatment modality for OSA. In a 1994
publication’ Kamami
32
documented 43.4% resolution of
OSA, with only 13% non-responders aer multi session
LAUP. In a similar vein, Peng et al., reported a 79.17%
rate of eectiveness based on polysomnography results
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
90
from 96 patients who had received LAUP.
33
Contrarily,
Göktas Ö et al., found worsening AHI scores (>5/h) post
LAUP in 12 out of 25 patients.
34
A recent meta-analysis
and systematic review by Camacho et al., recommended
clinicians either to abandon LAUP as a therapy for OSA
or to perform it with caution because individual data
revealed 44% of cases with worsening AHI following
LAUP.
35
Camacho et al., attributed the deterioration
of OSA aer LAUP to the destruction of the so palate
surface which leads to the reduction of the reexogenic
dilation of the upper airway.
35
Radiofrequency ablation
Radiofrequency ablation (RFA) is considered a
conservative procedure that involves the insertion of a
Radiofrequency (RF) electrode probe into the submucosa
of the so palate or the tongue to decrease the size of these
structures. RFA is usually delivered at low temperatures
to prevent post-operative pain and discomfort. Powell
et al., pioneered the use of RFA in mild sleep disordered
breathing in 1998 by inserting a custom RF electrode to
the so palate under local anesthesia, and aer 8 to 12
weeks, radiographic results showed a mean shrinkage
of 5.5±3.7 mm which objectively improved ESS scores
(8.5±4.4 to 5.2±3.3) and the mean 95
th
percentile inspiratory
nadir (-19.71±5.29 cm to -12.78±6.28) cm H
2
O.
36
Powell
et al., followed up in 1999 with the RFA of the dorsal
surface of the tongue demonstrating positive results in
tongue size reduction (-17%), mean RDI (39.6 to 17.9)
and mean SaO
2
nadir (81.9% to 88.3%).
37
Additionally,
Riley et al., found promising results (mean AHI 35.1 to
15.1) aer multiple RFA applications to the ventral and
dorsal surface of the tongue. In a study on ai patients,
Sonsuwan et al., documented 16 out of 51 patients that
had AHI scores of < 5 following RFA of the so palate.
ese 16 patients had lower baseline AHI scores pre-op
which was inferred as a key factor of success in RFA for
OSA.
38
Maxillomandibular advancement
Maxillomandibular advancement (MMA) routinely
involves a Lefort 1 osteotomy (Fig 3) and Bilateral Sagittal
Split Ramus Osteotomy (BSSRO) (Fig 4) with or without
genioplasty advancement. Although MMA is usually
performed on moderate to severe OSA patients with
retrognathic maxilla and mandibles, some OSA cases
with class 1 skeletal relationships could be candidates for
MMA as well.
39
Morbidly obese patients who are non-
adherrent to CPAP have been documented to respond
favorably aer MMA.
40,41
Despite morbid obesity not being
a denite contraindication for MMA, the eectiveness
of MMA in this group of patients remains inconclusive
due to the lack of data and insucient studies.
42
Clear
indications to perform MMA as a rst line surgical
therapy includes retrognathic jaws, severe OSA and
cases with complete concentric closure of the velum
and lateral pharyngeal wall diagnosed with DISE.
38
Fig 3. Lateral view (1 side) of a Lefort 1 osteotomy. e thick dotted
lines show the osteotomies at the pterygomaxillary junction. e
thin dashed lines show the horizontal osteotomies that form a wedge
shape which facilitates anti-clockwise movement of the maxilla.
Fig 4. Bilateral sagittal split osteotomy of the mandible. Necessary
precautions are necessary to prevent inferior alveolar nerve injury.
e Stanford MMA technique developed by Powell
and Riley has allowed larger advancements, up to 10
mm in the maxilla and 11 mm in the mandible which
eectively expands the upper airway space without the
need for hyoid suspension.
43
An updated version of this
technique involves Lefort 1 and Bilateral Sagittal split
osteotomies with the maxillomandibular complex rotated
in an anti-clockwise manner to maximize esthetic results
and tension of the lateral pharyngeal wall.
39
e success rate of MMA for OSA when performed
in either phase 1 or phase 2 of the Stanford protocol
has been reported to be more than 90% in patients with
BMI<40 and 81% in patients with BMI>40.
43
Studies
that conducted 3-dimensional Computer Tomography
analysis of the upper airway in a static model post MMA
Khoo et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
91
Review Article
SMJ
documented improvement in retropalatal and retroglossal
parameters.
44,45
With the aid of DISE, dynamic measurements
were recorded post MMA which revealed the lateral
pharyngeal wall tension during sleep as a signicant
marker for success.
13
Kastoer et al., documented promising
results with MMA for treating OSA patients with complete
concentric collapse of the velum by eliminating the
complete closure thereby reducing the AHI values by
9.9 ± 7.2 events per hour from a baseline of 40.2 ± 25.6
events per hour.
46
Despite the high success rates and predictability of
MMA as a treatment for OSA, the advancement of the
maxillomandibular complex will undoubtedly alter the
facial aesthetics of an individual. Li et al., conducted a
questionnaire survey on patients 6 to 12 months aer
MMA for OSA, and 55% (24/44) of the patients reported
feeling more attractive or youthful whereas 4 patients
felt unfavorable changes to their facial appearance.
47
Large Lefort 1 advancements causes nasal aring. Liu
et al., reported the need for corrective nasal surgery
in 18.7% of patients aer MMA.
48
ese unfavorable
cosmetic results aer MMA could be more pronounced
in Asians, because Asians tend to have lips that are more
outwardly projected
49
and atter nasal bridges.
50
With
this in mind, Liao et al., proposed a modied MMA
technique for Asians that added anterior segmental
osteotomies from premolar extraction sites, and they found
a marked reduction in mean AHI scores (41.6±19.2/h to
5.3±4.0/h) with improved facial aesthetics post operation.
50
Nonetheless, according to Liu et al., MMA in Asians
improves their facial appearance therefore surgeons
should not be deterred from using MMA as a modality
to resolve OSA in Asians as long as proper planning is
done prior to surgery.
51
Recently, a prospective multicenter study which
evaluated the success of MMA for OSA reported signicant
improvements in mean ESS (13.3 to 4.9) and AHI (39.6
to 7.9) values.
52
ese results are consistent with the
2016 meta-analysis by Zaghi et al., which documented
98.8% improvement in AHI and RDI values aer MMA,
based on 45 studies with a total of 518 unique cases.
53
Every surgery carries risks and potential complications,
likewise MMA is an invasive surgery that could cause
numbness, pain and swelling around the maxillofacial
region.
53
Adverse events such as plate exposure and
wound infection especially in the mandible may arise
following MMA surgery.
52
Hypoglossal nerve stimulation
In 1993 the inuence of hypoglossal nerve stimulation
(HGNS) on the upper airway was rst studied in cats by
Schwartz et al.
54
In 2011 Eastwood et al., conducted the rst
human trials with HGNS, with 21 CPAP non-compliant
patients with moderate to severe OSA who were included
in that trial, and 19/21 had polysomnography results at
baseline and 6 months aer HGNS device installation. AHI
scores improved from 43.15±17.5/h to 19.5±16,7/h, ESS
scores reduced from 12.1±4.7 to 8.1±4.4.
55
Subsequently,
a multicenter prospective single-group clinical trial (STAR
trial) with 126 participants was performed to assess the
ecacy of HGNS for moderate to severe OSA, and the
mean AHI demonstrated improvement aer 12 months of
HGNS (32.0±11.8 to 15.3±16.1).
56
Additionally, 116 of the
126 participants from the STAR trial were reviewed again
aer 36 months, and 98 agreed to a polysomnography
test at 36 months aer HGNS implantation. e results
demonstrated stable AHI and ODI in these participants,
thus, proving the long-term eectiveness of HGNS for
moderate to severe OSA.
57
HGNS oers CPAP non-
compliant OSA patients who are averse to facial skeletal
changes from MMA, an eective second-line treatment
option. However, one of the contraindications in HGNS
is the complete concentric closure of the velum during
DISE examination, so these cases are best treated with
MMA.
39
Palatal implants
Palatal implants or pillar implants are plastic implants
that are inserted into the so palate to stien the so
palate in OSA or snoring (Fig 5). Pillar implants have
been largely eective in mild to moderate OSA.
58,59
In
2008 A randomized double-blinded placebo clinical trial
was conducted by Friedman et al., with encouraging
results aer palatal stiening with polyester implants
(AHI reduction was -7.9±7.7).
58
In 2013, a meta-analysis
published by Choi et al., found improved ESS and AHI
scores post pillar implantation in mild and moderate OSA.
Pillar implants were also found to have an overall low
complication rate which was mostly trivial in nature.
59
Fig 5. Pillar/palatal implants. e implants help stien the palate to
prevent collapse.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
92
Transoral robotic surgery
Transoral oral robotic surgery (TORS) was introduced
by Vicini et al., and they documented tongue base reduction,
supraglottoplasty, nasal surgeries and UPPP with the Da
Vinci Robot.
60
Vicini et al., reported AHI reduction in 8
out of 10 patients aer TORS, although 2 patients had
worsening AHI.
60
In 2019, Vicini et al described TORS
supraglottoplasty and tongue base reduction in detail,
and they advocated TORS due to the precise dissection
and acceptable surgical time (30minutes for tongue
base reduction, and 15 minutes for Supraglottoplasty).
61
Despite accurate and minimally invasive surgeries, TORS
is still not cost eective enough to warrant widespread
use. Because of the price of the Da Vinci Surgical System,
only selected OSA cases with tongue base hypertrophy
and epiglottis collapse could potentially benet from
TORS.
DISCUSSION
CPAP is recommended as the rst-line therapy in
adult OSA cases but non-compliance remains a tough
conundrum for patients and clinicians. McEvoy et al.,
published an alarming study in the New England Journal
of Medicine which documented the ineectiveness of
CPAP for preventing cardiovascular events in moderate
to severe OSA cases.
62
ese ndings coupled with the
CPAP non-compliance rate could surmise a need for a
shi in ideology wherein surgery is placed as a rst-line
therapy which xes the etiology of OSA, particularly in
cases with notable anatomical obstructions. Rotenberg
et al., concluded in a systematic review that CPAP should
no longer be regarded as gold-standard therapy, because
of the high non-compliance and the long-term success
of various surgical options. A glaring deciency in the
majority of CPAP studies is the short-term follow-up
compared to the longer follow-up period in many surgical
trials.
63
MMA is currently deemed as the most eective
surgical modality aside from tracheostomy for moderate to
severe OSA. e advancement of the maxillo-mandibular
complex causes an increase in the width of the upper airway
length and tension of the lateral pharyngeal wall at the
expense of facial aesthetics for some patients. erefore,
proper treatment planning and careful consideration of
the baseline skeletal anatomy is needed for an acceptable
outcome. Although a modied MMA technique with
anterior segmental osteotomies to accommodate Asian
anatomy showed promising results,
50
further research
with long-term follow ups and larger samples are required
to generalize this eectiveness. UPPP is very eective in
selected mild to moderate OSA cases based on Friedman’s
staging system. CPAP non-compliant patients require
thorough examination of their nasal anatomy via nasal
endoscopy because nasal surgeries are proven to increase
CPAP compliance. Some OSA patients that are ideal
candidates for MMA may refuse MMA, in these cases,
OSA can be managed with a combination of UPPP, nasal
surgeries and HGNS.
64
OSA is a chronic disease with no denite cure, and
Liu et al., proposed an extension of the Stanford protocol
when they documented for the rst time a case of relapsed
OSA in a 65 years male previously treated with 2 phases
of surgeries: MMA when he was 49; UPPP, septoplasty
and RFA of tongue when he was 60.
65
e latest surgery
on the patient was the implantation of a HGNS device
which reduced ESS to 6 and eectively reduced AHI to
1 or less depending on sleep position.
65
e authors of
this review believe that customized treatment planning
and multiple surgeries are necessary for the long-term
stabilization of moderate to severe OSA. erefore, ENT
surgeons and oral and maxillofacial surgeons who deal
with OSA should be equipped with all the surgical skills
to oer patients the most appropriate treatment plan.
CONCLUSION
CPAP non-compliant patients with moderate to
severe OSA require surgical therapy to either resolve
OSA or to increase CPAP adherence. Further research
in each surgical modality for OSA should be conducted
for assessing the ecacy and long-term success rates.
Currently, MMA is the only surgical modality aside
from tracheostomy that could provide consistent AHI
and RDI reduction in moderate to severe OSA caused by
multilevel obstructions, although research documenting
MMA as an isolated rst line procedure is scarce. As of
today, we can conclude that moderate to severe OSA
in CPAP non-compliant patients is best treated with
surgeries that x multilevel obstructions. Additionally,
Hypoglossal nerve stimulation could play a more important
role in the future in cases with relapsed OSA who have
already undergone multiple surgeries.
ACKNOWLEDGMENTS
e authors would like to thank the sta and dental
assistants including colleagues and co-workers in the
Department of Oral and Maxillofacial Surgery, Faculty
of Dentistry, Mahidol University.
Funding: is research had no nancial support
Conict of interest: All authors declare no conict
interest in this research
Ethics approval: not required
Khoo et al.
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
93
Review Article
SMJ
REFERENCES
1. Ward KL, Hillman DR, James A, Bremner AP, Simpson L,
Cooper MN, et al. Excessive daytime sleepiness increases the
risk of motor vehicle crash in obstructive sleep apnea. Journal
of clinical sleep medicine.J Clin Sleep Med 2013;9:1013-21.
2. Arzt M, Young T, Finn L, Skatrud JB, Bradley TD. Association
of sleep-disordered breathing and the occurrence of stroke.
Am J Respir Crit Care Med 2005;172:1447-51.
3. Redline S, Yenokyan G, Gottlieb DJ, Shahar E, O’Connor GT,
Resnick HE, et al. Obstructive sleep apnea-hypopnea and
incident stroke: the sleep heart health study. Am J Respir Crit
Care Med 2010;182:269-77.
4. Hung J, Whitford EG, Parsons RW, Hillman DR. Association of
sleep apnoea with myocardial infarction in men. Lancet
1990;336:261-4.
5. Peppard PE, Young T, Palta M, Skatrud J. Prospective study of
the association between sleep-disordered breathing and
hypertension. N Engl J Med 2000;342:1378-84.
6. Rotenberg BW, Murariu D, Pang KP. Trends in CPAP adherence
over twenty years of data collection: a attened curve. J Otolaryngol
Head Neck Surg 2016;45:43.
7. Chandrashekariah R, Shaman Z, Auckley D. Impact of upper
airway surgery on cpap compliance in dicult-to-manage
obstructive sleep apnea. Arch Otolaryngol Head Neck Surg
2008;134:926-30.
8. Boyd SB, Walters AS, Waite P, Harding SM, Song Y. Long-
term eectiveness and safety of maxillomandibular advancement
for treatment of obstructive sleep apnea. J Clin Sleep Med 2015;
11:699-708.
9. Bjorvatn B, Rajakulendren N, Lehmann S, Pallesen S. Increased
severity of obstructive sleep apnea is associated with less anxiety
and depression. J Sleep Res 2018;27:e12647.
10. Li HY, Lee LA, Hsin LJ, Fang TJ, Lin WN, Chen HC, et al.
Intrapharyngeal surgery with integrated treatment for obstructive
sleep apnea. Biomed J 2019;42:84-92.
11. Koutsourelakis I, Saruddin F, Ravesloot M, Zakynthinos S,
de Vries N. Surgery for obstructive sleep apnea: Sleep endoscopy
determinants of outcome. Laryngoscope 2012;122:2587-91.
12. Kezirian EJ, Hohenhorst W, de Vries N. Drug-induced sleep
endoscopy: the VOTE classication. Eur Arch Otorhinolaryngol
2011;268:1233-6.
13. Liu SY, Huon LK, Powell NB, Riley R, Cho HG, Torre C,
et al. Lateral pharyngeal wall tension aer maxillomandibular
advancement for obstructive sleep apnea is a marker for
surgical success: observations from drug-induced sleep endoscopy.
J Oral Maxillofac Surg 2015;73:1575-82.
14. Valero A, Alroy G. Hypoventilation in acquired micrognathia.
Arch Intern Med 1965;115:307-10.
15. Guilleminault C, Simmons FB, Motta J, Cummiskey J, Rosekind M,
Schroeder JS, et al. Obstructive sleep apnea syndrome and
tracheostomy. Long-term follow-up experience. Arch Intern
Med 1981;141:985-8.
16. Kuhlo W, Doll E, Franck MC. Successful management of
Pickwickian syndrome using long-term tracheostomy. Dtsch
Med Wochenschr 1969;94:1286-90.
17. Weitzman ED, Kahn E, Pollak CP. Quantitative analysis of
sleep and sleep apnea before and aer tracheostomy in patients
with the hypersomnia-sleep apnea syndrome. Sleep 1980;3:
407-23.
18. El Solh AA, Jaafar W. A comparative study of the complications
of surgical tracheostomy in morbidly obese critically ill patients.
Crit Care 2007;11:R3.
19. Hashmi NK, Ransom E, Nardone H, Redding N, Mirza N.
Quality of life and self-image in patients undergoing tracheostomy.
Laryngoscope 2010;120 Suppl 4:S196.
20. Aurora RN, Casey KR, Kristo D, Auerbach S, Bista SR, Chowdhuri
S, et al. Practice parameters for the surgical modications
of the upper airway for obstructive sleep apnea in adults. Sleep
2010;33:1408-13.
21. Johnson DM, Soose RJ. Updated nasal surgery for obstructive
sleep apnea. Adv Otorhinolaryngol 2017;80:66-73.
22. Nakata S, Noda A, Yagi H, Yanagi E, Mimura T, Okada T, et al.
Nasal resistance for determinant factor of nasal surgery in CPAP
failure patients with obstructive sleep apnea syndrome. Rhinology
2005;43:296-9.
23. Ikematsu T. Study of snoring, 4th report: therapy. er J Jpn
Otorhinolaryngol. 1964;64:434-5.
24. Fujita S, Conway W, Zorick F, Roth T. Surgical correction
of anatomic abnormalities in obstructive sleep apnea syndrome:
uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg
1981;89:923-34.
25. Sher AE, Schechtman KB, Piccirillo JF. e ecacy of surgical
modications of the upper airway in adults with obstructive
sleep apnea syndrome. Sleep 1996;19:156-77.
26. Värendh M, Berg S, Andersson M. Long-term follow-up of
patients operated with uvulopalatopharyngoplasty from 1985
to 1991. Respir Med 2012;106:1788-93.
27. Rosvall BR, Chin CJ. Is uvulopalatopharyngoplasty eective
in obstructive sleep apnea? Laryngoscope 2017;127:2201-2.
28. Friedman M, Ibrahim H, Bass L. Clinical staging for sleep-
disordered breathing. Otolaryngol Head Neck Surg 2002;127:
13-21.
29. Riley RW, Powell NB, Guilleminault C. Obstructive Sleep
Apnea Syndrome: A review of 306 consecutively treated surgical
patients. Otolaryngol Head Neck Surg 1993;108:117-25.
30. Neruntarat C. Genioglossus advancement and hyoid myotomy
under local anesthesia. Otolaryngol Head Neck Surg 2003;129:
85-91.
31. Vargo JD, Ogan WS, Tanna N, Stevens D, Andrews BT. Modied
genioglossal advancement for isolated treatment of obstructive
sleep apnea. J Craniofac Surg 2017;28:1274-7.
32. Kamami YV. Outpatient treatment of sleep apnea syndrome
with CO
2
laser, LAUP: laser-assisted UPPP results on 46
patients. J Clin Laser Med Surg 1994;12:215-9.
33. Peng Y, Fan J, Wu J, Deng B, Sun A. Evaluate the eect of
CO
2
laser-assisted uvulopalatopharyngoplasty for obstructive
sleep apnea-hypopnea syndrome. Lin Chung Er Bi Yan Hou
Tou Jing Wai Ke Za Zhi 2009;23:302-4.
34. Göktas Ö, Solmaz M, Göktas G, Olze H. Long-term results
in obstructive sleep apnea syndrome (OSAS) aer laser-assisted
uvulopalatoplasty (LAUP). PLoS One. 2014;9:e100211.
35. Camacho M, Kushida CA, Lambert E, Nesbitt NB, Song SA,
Chang ET, et al. Laser-assisted uvulopalatoplasty for obstructive
sleep apnea: a systematic review and meta-analysis. Sleep
2017;40(3). doi: 10.1093/sleep/zsx004.
36. Powell NB, Riley RW, Troell RJ, Li K, Blumen MB, Guilleminault
C. Radiofrequency volumetric tissue reduction of the palate in
subjects with sleep-disordered breathing. Chest 1998;113:1163-
74.
37. Powell NB, Riley RW, Guilleminault C. Radiofrequency tongue
Volume 72, No.1: 2020 Siriraj Medical Journal
www.smj.si.mahidol.ac.th
94
base reduction in sleep-disordered breathing: a pilot study.
Otolaryngol Head Neck Surg 1999;120:656-64.
38. Sonsuwan N, Rujimethabhas K, Sawanyawisuth K. Factors
associated with successful treatment by radiofrequency treatment
of the so palate in obstructive sleep apnea as the rst-line
treatment. Sleep Disord 2015;2015:690425.
39. Liu SY, Awad M, Riley RW. Maxillomandibular Advancement:
contemporary approach at Stanford. Atlas Oral Maxillofac
Surg Clin North Am 2019;27:29-36.
40. Li KK, Powell NB, Riley RW, Zonato A, Gervacio L, Guilleminault
C. Morbidly obese patients with severe obstructive sleep apnea: is
airway reconstructive surgery a viable treatment option?
Laryngoscope 2000;110:982-7.
41. Do MH, Jansma J, Schepers RH, Hoekema A. Maxillomandibular
advancement surgery as alternative to continuous positive
airway pressure in morbidly severe obstructive sleep apnea: a
case report. Cranio 2013;31:246-51.
42. Camacho M, Teixeira J, Abdullatif J, Acevedo JL, Certal V,
Capasso R, et al. Maxillomandibular advancement and tracheostomy
for morbidly obese obstructive sleep apnea: a systematic review
and meta-analysis. Otolaryngol Head Neck Surg 2015;152:619-
30.
43. Camacho M, Liu SY, Certal V, Capasso R, Powell NB, Riley RW.
Large maxillomandibular advancements for obstructive sleep
apnea: an operative technique evolved over 30years. J Cranio-
maxillofac Surg 2015;43:1113-8.
44. Abramson Z, Susarla SM, Lawler M, Bouchard C, Troulis M,
Kaban LB. ree-dimensional computed tomographic airway
analysis of patients with obstructive sleep apnea treated by
maxillomandibular advancement. J Oral Maxillofac Surg
2011;69:677-86.
45. Abramson Z, Susarla S, August M, Troulis M, Kaban L. ree-
dimensional computed tomographic analysis of airway anatomy
in patients with obstructive sleep apnea. J Oral Maxillofac Surg
2010;68:354-62.
46. Kastoer C, Op de Beeck S, Dom M, Neirinckx T, Verbraecken J,
Braem MJ, et al. Drug-induced sleep endoscopy upper airway
collapse patterns and maxillomandibular advancement.
Laryngoscope 2019 Apr 29. doi: 10.1002/lary.28022. [Epub
ahead of print]
47. Li KK, Riley RW, Powell NB, Guilleminault C. Patient’s perception
of the facial appearance aer maxillomandibular advancement for
obstructive sleep apnea syndrome. J Oral Maxillofac Surg
2001;59:377-80.
48. Liu SY,Lee PJ,Awad M,Riley RW,Zaghi S.Corrective nasal
surgery aer maxillomandibular advancement for obstructive
sleep apnea: experience from 379 cases. Otolaryngol Head
Neck Surg 2017;157:156-9.
49. Gu Y, McNamara JA, Jr, Sigler LM, Baccetti T. Comparison of
craniofacial characteristics of typical Chinese and Caucasian
young adults. Eur J Orthod 2011;33:205-11.
50. Liao YF, Chiu YT, Lin CH, Chen YA, Chen NH, Chen YR.
Modied maxillomandibular advancement for obstructive sleep
apnoea: towards a better outcome for Asians. Int J Oral Maxillofac
Surg 2015;44:189-94.
51. Liu Sr, Yi Hl, Guan J, Chen B, Wu Hm, Yin Sk. Changes in
facial appearance aer maxillomandibular advancement for
severe obstructive sleep apnoea hypopnoea syndrome in
Chinese patients: a subjective and objective evaluation. Int J
Oral Maxillofac Surg 2012;41:1112-9.
52. Boyd SB, Chigurupati R, Cillo JE, Jr., Eskes G, Goodday R,
Meisami T, et al. Maxillomandibular advancement improves
multiple health-related and functional outcomes in patients
with obstructive sleep apnea: a multicenter study. J Oral
Maxillofac Surg 2019;77:352-70.
53. Zaghi S, Holty J-EC, Certal V, Abdullatif J, Guilleminault C,
Powell NB, et al. Maxillomandibular advancement for treatment
of obstructive sleep apnea: a meta-analysis. JAMA Otolaryngol
Head Neck Surg 2016;142:58-66.
54. Schwartz AR, ut DC, Russ B, Seelagy M, Yuan X, Brower
RG, et al. Eect of electrical stimulation of the hypoglossal
nerve on airow mechanics in the isolated upper airway. Am
Rev Respir Dis 1993;147:1144-50.
55. Eastwood PR, Barnes M, Walsh JH, Maddison KJ, Hee G,
Schwartz AR, et al. Treating obstructive sleep apnea with
hypoglossal nerve stimulation. Sleep 2011;34:1479-86.
56. Strollo PJ, Soose RJ, Maurer JT, de Vries N, Cornelius J,
Froymovich O, et al. Upper-Airway Stimulation for Obstructive
Sleep Apnea. N Engl J Med 2014;370:139-49.
57. Woodson T, Soose R, Gillespie M, Strohl K, Maurer J, de vries
N, et al. ree-Year Outcomes of Cranial Nerve Stimulation
for Obstructive Sleep Apnea: e STAR Trial2015. Otolaryngol
Head Neck Surg 2016;154:181-8.
58. Friedman M, Schalch P, Lin H-C, Kakodkar KA, Joseph NJ,
Mazloom N. Palatal implants for the treatment of snoring
and obstructive sleep apnea/hypopnea syndrome. Otolaryngol
Head Neck Surg 2008;138:209-16.
59. Choi JH, Kim S-N, Cho JH. Ecacy of the pillar implant in
the treatment of snoring and mild-to-moderate obstructive
sleep apnea: A meta-analysis. Laryngoscope 2013;123:269-76.
60. Vicini C, Dallan I, Canzi P, Frassineti S, La Pietra MG, Montevecchi
F. Transoral Robotic Tongue Base Resection in Obstructive
Sleep Apnoea-Hypopnoea Syndrome: A Preliminary Report.
ORL 2010;72:22-27.
61. Vicini C, Montevecchi F. Transoral Robotic Surgery for
Obstructive Sleep Apnea: Past, Present, and Future. Sleep Med
Clin 2019;14:67-72.
62. McEvoy RD, Antic NA, Heeley E, Luo Y, Ou Q, Zhang X, et al.
CPAP for Prevention of Cardiovascular Events in Obstructive
Sleep Apnea. N Engl J Med 2016;375:919-31.
63. Rotenberg BW, Vicini C, Pang EB, Pang KP. Reconsidering
rst-line treatment for obstructive sleep apnea: a systematic
review of the literature. J Otolaryngol Head Neck Surg 2016;45:23.
64. Liu SY, Wayne Riley R, Pogrel A, Guilleminault C. Sleep Surgery
in the Era of Precision Medicine. Atlas Oral Maxillofac Surg
Clin North Am 2019;27:1-5.
65. Liu SY, Riley RW. Continuing the Original Stanford Sleep
Surgery Protocol From Upper Airway Reconstruction to Upper
Airway Stimulation: Our First Successful Case. J Oral Maxillofac
Surg 2017;75:1514-8.
Khoo et al.
