SMJ

E-ISSN 2228-8082

Volume 75, Number 10, October 2023

MONTHLY

Indexed by

THAILAND SECTION 1954

https://he02.tci-thaijo.org/index.php/sirirajmedj/index

E-mail: sijournal92@gmail.com

ORIGINAL ARTICLE LETTER TO THE EDITOR

Siriraj Medical Journal

The world-leading biomedical science of Thailand

ORIGINAL ARTICLE‌

699 Prevalence of Urinary Tract Infection in Pregnancies with Premature Uterine Contractions

Pattarawalai Talungchit, Pornpimol Ruangvutilert,

SMJ

Siriraj Medical Journal

The world-leading biomedical science of Thailand

Volume 75 Number 10

October 2023

Tachjaree Panchalee boonbowornpong, Buraya Phattanachindakun

707 Efficacy of Music Therapy and Zingiber officinale Roscoe Aromatherapy for Reducing Pain during the First Stage of Labor:

A Randomized Controlled Trial

Pontip Paoin, Pichita Prasongvej, Athita Chanthasenanont, Titchayakorn Niumpradit, Densak Pongrojpaw, Komsun Suwannarurk

713 Heart Rate Variability and Baroreflex Sensitivity: Factors and Reference Ranges

Paveena Somboon, Surat Kulapatana, Wattana B. Watanapa, Tanchanok Chattaris, Phatadon Sirivongrangson,

Panomporn Livikorn, Thongchai Nonthadid, Poungpetch Hussarin

725 Cost-Effectiveness Analysis of Long-acting Injectable Once-monthly of Aripiprazole Compared with Long-acting Injectable Oncemonthly Paliperidone Palmitate for the Treatment of Stable

Schizophrenia Patients in Thailand

Piyanut Ueapanjasin, Wiwat Thavornwattanayong, Jadesada Lertsirimunkong, Kamolpat Chaiyakittisopon

736 Burnout and Associated Factors among Thai Anesthesiology Residents

Neranchala Soonthornkes, Orattha Sakornwattananon, Jitpanu Jirapitakkul, Winitra Nuallaong

744 Associated Factors of Quality of Life in Adult Female Acne Coexisting with Hyperandrogenism and Polycystic Ovarian Syndrome using the Dermatology Life Quality Index

Pattriya Jirawattanadon, Leena Chularojanamontri, Yanisorn Nanchaipruek, Panicha Chantrapanichkul, Papapit Tuchinda, Kanokvalai Kulthanan

752 Evaluation of the Efficacy and Safety of the ITM 68Ge/68Ga Generator After its Recommended Shelf-life

759

Efficacy of Oral Iron Supplementation in Treating Patients with Female Pattern Hair Loss and Low Serum Ferritin: A Pilot Study

Daranporn Triwongwaranat, Kanchalit Thanomkitti, Supisara Wongdama, Nuttagarn Jantanapornchai,

Chanika Subchookul, Chayanee Likitwattananurak, Noldtawat Viriyaskultorn, Supenya Varothai, Rattapon Thuangtong

LETTER TO THE EDITOR

Tossaporn Sriprapa, Thanete Doungta, Nopparath Sakulsamart, Nilmanee Taweewatthanasopon, Lanyawat Madputeh, Pitima Ragchana, Napamon Sritongkul, Malulee Tantawiroon, Somlak Kongmuang, Benjapa Khiewvan, Shuichi Shiratori

SMJ

SIRIRAJ MEDICAL JOURNAL

https://he02.tci-thaijo.org/index.php/sirirajmedj/index

Executive Editor: Apichat Asavamongkolkul Editorial Director: Aasis Unnanuntana

Editor-in-Chief: Thawatchai Akaraviputh, Mahidol University, Thailand

Associate Editors

Adisorn Ratanayotha, Mahidol University, Thailand Chenchit Chayachinda, Mahidol University, Thailand Pornprom Muangman, Mahidol University, Thailand Phunchai Charatcharoenwitthaya, Mahidol University, Thailand

Varut Lohsiriwat, Mahidol University, Thailand

Andrew S.C. Rice, Imperial College London, UK

International Editorial Board

Morris Solomon Odell, Monash University, Australia

Anusak Yiengpruksawan, The Valley Robotic Institute, USA Barbara Knowles, The Jackson Laboratory, USA Christopher Khor, Singapore General Hospital, Singapore Ciro Isidoro, University of Novara, Italy

David S. Sheps, University of Florida, USA

David Wayne Ussery, University of Arkansas for Medical Sciences, USA Davor Solter, The Jackson Laboratory, USA

Dennis J. Janisse, Medical College of Wisconsin, USA

Dong-Wan Seo, University of Ulsan College of Medicine, Republic of Korea Folker Meyer, Argonne National Laboratory, USA

Frans Laurens Moll, University Medical Center Ultrecht, Netherlands

G. Allen Finley, Delhousie University, Canada

George S. Baillie, University of Glasgow, United Kingdom

Gregory Bancroft, London School of Hygiene of Tropical Medicine, United Kingdom Gustavo Saposnik, St. Michael’s Hospital, Canada

Harland Winter, Harvard Medical School, USA

Hidemi Goto, Nagoya University Graduate School of Medicine, Japan Ichizo Nishino, National Institute of Neuroscience NCNP, Japan Intawat Nookaew, University of Arkansas for Medical Sciences, USA James P. Doland, Oregon Health & Science University, USA

John Damian Smith, Texas A&M University-San Antonio, USA John Hunter, Oregon Health & Science University, USA

Juri Gelovani, Wayne State University, USA

Karl Thomas Moritz, Swedish University of Agricultural Sciences, Sweden Kazuo Hara, Aichi Cancer Center Hospital, Japan

Keiichi Akita, Tokyo Medical and Dental University Hospital, Japan Kym Francis Faull, David Geffen School of Medicine, USA

Kyoichi Takaori, Kyoto University Hospital, Japan Marcela Hermoso Ramello, University of Chile, Chile Marianne Hokland, University of Aarhus, Denmark

Matthew S. Dunne, Institute of Food, Nutrition, and Health, Switzerland Mitsuhiro Kida, Kitasato University & Hospital, Japan

Moses Rodriguez, Mayo Clinic, USA

Nam H. CHO, Ajou University School of Medicine and Hospital, Republic of Korea Nima Rezaei, Tehran University of Medical Sciences, Iran

Noritaka Isogai, Kinki University, Japan

Paul James Brindley, George Washington University, USA

Pauline Mary Rudd, National Institute for Bioprocessing Research and Training Fosters Avenue Mount Merrion Blackrock Co., Dublin, Ireland

Peter Hokland, Aarhus University Hospital, Denmark

Philip A. Brunell, State University of New York At Buffalo, USA Philip Board, Australian National University, Australia

Richard J. Deckelbaum, Columbia University, USA Richard W. Titball, University of Exeter, USA Robert W. Mann, University of Hawaii, USA

Robin CN Williamson, Royal Postgraduate Medical School, United Kingdom Sara Schwanke Khilji, Oregon Health & Science University, USA

Seigo Kitano, Oita University, Japan

Shomei Ryozawa, Saitama Medical University, Japan Shuji Shimizu, Kyushu University Hospital, Japan

Stanlay James Rogers, University of California, San Francisco, USA Stephen Dalton, University of Georgia, USA

Sue Fletcher, Murdoch University, Australia

Tai-Soon Yong, Yonsei University, Republic of Korea Tomohisa Uchida, Oita University, Japan

Victor Manuel Charoenrook de la Fuente, Centro de Oftalmologia Barraquer, Spain Vincent W.S. Chan, University of Toronto, Canada

Wen-Shiang Chen, National Taiwan University College of Medicine, Taiwan Wikrom Karnsakul, Johns Hopkins Children’s Center, USA

Yasushi Sano, Director of Gastrointestinal Center, Japan Yik Ying Teo, National University of Singapore, Singapore Yoshiki Hirooka, Nagoya University Hospital, Japan

Yozo Miyake, Aichi Medical University, Japan Yuji Murata, Aizenbashi Hospital, Japan

Ampaiwan Chuansumrit, Mahidol University, Thailand Anuwat Pongkunakorn, Lampang Hospital, Thailand Jarupim Soongswang, Mahidol University, Thailand Nopphol Pausawasdi, Mahidol University, Thailand Nopporn Sittisombut, Chiang Mai University, Thailand Pa-thai Yenchitsomanus, Mahidol University, Thailand Pornchai O-Charoenrat, Mahidol University, Thailand Prapon Wilairat, Mahidol University, Thailand Puttinun Patpituck, Mahidol University, Thailand Rungroj Krittayaphong, Mahidol University, Thailand Saranatra Waikakul, Mahidol University, Thailand

Editorial Board

Sayomporn Sirinavin, Mahidol University, Thailand Suneerat Kongsayreepong, Mahidol University, Thailand Supakorn Rojananin, Mahidol University, Thailand Surapol Issaragrisil, Mahidol University, Thailand

Suttipong Wacharasindhu, Chulalongkorn University, Thailand Vasant Sumethkul, Mahidol University, Thailand

Vitoon Chinswangwatanakul, Mahidol University, Thailand Watchara Kasinrerk, Chiang Mai University, Thailand Wiroon Laupattrakasem, Khon Kaen University, Thailand Yuen Tanniradorn, Chulalongkorn University, Thailand

Journal Manager: Nuchpraweepawn Saleeon, Mahidol University, Thailand

Medical Illustrator: Nuchpraweepawn Saleeon, Mahidol University, Thailand

Proofreaders: Noochpraweeporn Saleeon, Mahidol University, Thailand, Amornrat Sangkaew, Mahidol University, Thailand

Office: His Majesty the King’s 80th Birthday Anniversary 5th December 2007 Building (SIMR), 2nd Fl., Room No.207 Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand Tel: 02-419-2888 Fax: 02-411-0593 E-mail: sijournal92@gmail.com

Prevalence of Urinary Tract Infection in Pregnancies with Premature Uterine Contractions

Pattarawalai Talungchit, M.D., Ph.D., Pornpimol Ruangvutilert, M.D., Ph.D., Tachjaree Panchalee boonbowornpong, M.D., Buraya Phattanachindakun, M.D.

Department of Obstetrics and Gynaecology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand.

ABSTRACT

Objective: This study aimed to determine the prevalence of urinary tract infection (UTI) among pregnant women with premature uterine contractions. Roles of the current practice of routine simultaneous urinalysis and urine culture in these cases were also evaluated.

Materials and Methods: Medical records of pregnant women admitted with premature uterine contractions at Siriraj Hospital between January 2013 and December 2017 were reviewed. Prevalence of UTI in these women was determined. Women were divided into 2 groups based on diagnosis at admission; group 1 included preterm labor/preterm prelabor rupture of membranes (PTL/PPROM), and group 2 was threatened preterm labor (TPL). Evaluations of urinalysis and/or urine culture were performed in these two groups to establish a strategy to reduce unnecessary urine culture.

Results: The prevalence of UTI among 2,286 women with premature uterine contractions was 4.9%. Prevalence of UTI were not different between the two groups of women. A positive urine culture was found in 2.4%. The most common organism identified in both groups of women was Escherichia coli. Urinalysis with abnormal cell counts and/or significant presence of bacteria could be used to screen for women in TPL group who should have urine culture with a positive screening rate of 6.5% and 90.9% sensitivity. With this strategy, unnecessary urine culture could be reduced.

Conclusion: The prevalence of UTI among pregnant women admitted with premature uterine contractions was 4.9% and was not different between PTL/PPROM group and TPL group. Urinalysis with abnormal cell counts or significant bacteriuria could be a screening tool to reduce number of urine culture in women with TPL.

Keywords: Asymptomatic bacteriuria; urinalysis; urine culture; premature uterine contractions; pregnancy outcomes (Siriraj Med J 2023; 75: 699-706)

INTRODUCTION

Asymptomatic bacteriuria (ASB) is a significant bacterial infection of the urine without signs or symptoms of urinary infection. It is defined by the presence of one or more species of bacteria identified in the urine culture at a quantitative count of > 105 colony-forming units (CFU/ml).1 Reported prevalence of ASB in pregnancy ranges from 2% to 7%2 and may be higher in some conditions, such as diabetic pregnancies.3 If pregnant

women with ASB are not treated, up to 30% will develop acute pyelonephritis, which has been associated with low birth weight and preterm birth.4 Therefore, once detected, active management with antibiotics is recommended.5 Although early routine screening for and treatment of ASB in pregnancy are recommended in antenatal care guidelines,6 the benefits and harm of screening and treatment are still questionable in low-risk singleton pregnancies.7-9 Like most hospitals in Thailand, Siriraj

Corresponding author: Buraya Phattanachindakun E-mail: buraya.pha@mahidol.ac.th

Received 24 July 2023 Revised 16 September 2023 Accepted 19 September 2023 ORCID ID:http://orcid.org/0000-0002-5232-7959 https://doi.org/10.33192/smj.v75i10.264256

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

Hospital does not perform screening for ASB during the antenatal period due to the low prevalence (2.3%) without a significant difference in prevalence by trimester.9

Preterm birth is an important cause of neonatal morbidity and mortality due to the immaturity/prematurity of the newborn various organ systems. Preterm birth rate at Siriraj Hospital was reported to be 9-14%.10 Preterm birth usually is preceded by regular uterine contractions11 or membrane rupture. Infection is one of the proposed mechanisms involved in this condition. Urinary tract infections (UTIs), genital tract infections, and periodontal diseases have been reported to be associated with spontaneous preterm labor.12

Premature uterine contraction is a common problem in pregnancy. As mentioned, the contractions can precede a preterm birth. On the other hand, the contractions can be a mere false alarm. Generally, regular uterine contractions with a cervical dilatation of ≥ 2 cm and

≥ 80% effacement have a high risk of labor progression, hence the condition is termed preterm labor (PTL). Condition of regular premature uterine contractions with less cervical changes is traditionally termed “threatened preterm labor” (TPL). However, with a subtle or borderline cervical change, it is difficult at the initial assessment to differentiate between early PTL and TPL. Further observation is needed to evaluate cervical progression. Another condition associated with preterm birth is preterm prelabor rupture of membranes (PPROM). It defines spontaneous membrane rupture before 37 completed weeks and before labor onset or cervical changes. With PPROM, risk of infection is increased along with time. Management of pregnancies with either of the three conditions (PTL, PPROM, TPL) aims to delay delivery usually by tocolytics to avoid or reduce the risk of preterm birth unless/until there is a contraindication to do so. The women are admitted for evaluation and managed accordingly. Pregnancies with advanced PTL or PPROM are more likely than TPL to deliver in the

same admission.

Guidelines suggested by the Royal Thai College of Obstetricians and Gynaecologists (RTCOG) for managing preterm labor recommend performing septic work-ups to determine if an infection is the cause of preterm labor. Urine analysis and urine culture are often used to diagnose UTIs, despite the absence of signs and symptoms with midstream urine culture (MUC) being the gold standard.13 This is due to concern of ASB in these women.

At Siriraj Hospital, urinalysis is performed in women with premature uterine contractions on admission to detect UTI. Urine culture is also carried out simultaneously if feasible. This approach follows the RTCOG guidelines

with an idea that UTI is a possible cause of premature uterine contractions or membrane rupture, which in turn may lead to preterm birth. If found, treating UTI may improve the pregnancy outcomes. However, previous studies could not demonstrate clear association between positive midstream urine culture (MUCs) in pregnancies with preterm labor, preterm delivery or pregnancy outcomes.14 Furthermore, the rates of positive MUCs among women with preterm labor were reported to be 2.3% to 5% and the costs and benefits of performing MUCs were unclear.9,14-16

We aimed to determine the rate of UTI diagnosed by urinalysis and/or a positive urine culture in women admitted with premature uterine contractions. We carried out a study similar to the one previously performed in our institution.15 However, in that study, only urine culture results had been considered. We added urinalysis into the diagnostic process to investigate whether urine culture number could be reduced or not. Furthermore, currently, the urine specimen would be collected by a single urinary catheterization instead of clean voided urine catch to reduce the risk of contamination especially from the vaginal discharge/fluids which often presented in larger amount than normal in cases with premature uterine contractions and/or PPROM. The common organisms involved and the association between the urine culture results and pregnancy outcomes were also evaluated.

MATERIALS AND METHODS

This retrospective study was performed at Siriraj Hospital, a tertiary care and teaching hospital in Bangkok, Thailand. The Institutional Review Board of the Faculty of Medicine Siriraj Hospital approved this study (COA no. Si 320/2018). The medical records of pregnant women admitted between January 2013 and December 2017 with an ICD-10 diagnosis code related to preterm with delivery (O60.1), preterm without delivery or threatened preterm labor (O60.0), and preterm prelabor rupture of membranes (O42.91) were reviewed. They received medications to inhibit labor, and were worked-up to determine the causes of premature uterine contractions. Pregnant women with an underlying renal disease or with a known urinary tract disease or with signs or symptoms of UTI from history taking and/or physical examination were excluded, as were cases with a contraindication for labor inhibition.

An “abnormal urinary analysis suggestive of UTI” was defined as a white blood cell count > 5/high-power field17 or a red blood cell count > 5/high-power field without contamination and without crystal or casts.18 As

we excluded women with preexisting urinary tract diseases, we assumed that abnormal urinalysis was associated with UTI. For urine culture, “positive growth” was defined as a bacterial count > 105 CFU/ml, whereas lower counts or positive growth of multiple organisms were more likely to be associated with contamination.1 Since the women with symptoms of UTI had been excluded from the study, a positive growth in urine culture was also considered ASB.

Sample size was calculated based on the findings of Kiatsuda et al19 that UTI or ASB were presented in 15.5% of women admitted with preterm labor or threatened preterm labor. With alpha of 0.05 and d= 0.015, the sample size needed was 2,237 women.

At the admission, an initial diagnosis was given according to the history taking and vaginal speculum and/or cervical examination as appropriate. Women were divided into two groups according to their diagnosis. As women with TPL were likely to be able to continue pregnancy and PTL/PPROM were likely to deliver in the

same admission, they were allocated into either PTL/ PPROM or TPL group. Patient characteristics, urinary results, and pregnancy outcomes were recorded on a case record form. Data analyses were performed using PASW Statistics for Windows, version 18 (SPSS Inc., Chicago, IL, USA). Descriptive data are presented as percentages, means, standard deviations (SDs), or medians with interquartile ranges (IQRs). Continuous variables were analyzed using Student’s t-test, while categorical variables were examined with Pearson’s Chi-square test or Fisher’s exact test. A statistically significant difference was defined as a value < 0.05.

RESULTS

A total of 2,286 medical records were reviewed. Of these, simultaneous urinalysis and urine culture were available in 1,805 cases. There were 479 cases who had only urinalysis carried out. The remaining 2 cases had only culture result due to a mistake in specimen handling. As shown in Table 1, the mean maternal age at admission

TABLE 1. Basic characteristics of the study cohort.

was 29 ± 7 years old and the mean gestational age was

31.3 ± 2.6 weeks. Approximately 60% of the women were primiparous. About 7% of the women had a history of a prior preterm birth. Table 2 shows that abnormal urinalysis suggestive of infection was presented in 94 cases (4.1% of those who had a urinalysis performed) and a positive urine culture was presented in 43 cases (2.4% of those who had urine culture performed). Among these two categories, there were 24 women who had both abnormal urinalysis and a positive urine culture. Therefore, altogether there were 113 women diagnosed with UTI by urinalysis and/or culture, comprising 4.9% of the total cohort.

As mentioned earlier, we divided women into two groups based on the initial diagnosis: PTL/PPROM group and TPL group. Approximately two-thirds of the women were diagnosed with TPL. The patient characteristics are summarized in Table 1. At the end, 20% of women with TPL delivered in the same admission, mostly from progression of labor. Ninety-five percent of women in PTL/PPROM group delivered in this admission. Women

in this group who had not delivered were either referred to another hospital or discharged against advice. Length of stay of the total cohort was 8.8 ± 10.8 days, 9.3 ± 9 days and 8.5 ± 11.6 days in the total cohort, PTL/PPROM and TPL groups respectively.

Table 1 also shows that lower gestational age at admission were more prevalent in TPL group. This might be due to uterine contractions at such gestational age was likely to be a false alarm. The lower proportions of dexamethasone administration or side effects from inhibiting labor in the PTL/PPROM group were explained by the discontinuation of medications once the women delivered.

The urinary results of the cohort are presented in detail in Table 2. There were no statistically significant differences between the proportions of abnormal urinalysis (4.3% vs. 4.0%) or positive urine culture (1.7% vs. 2.7%) between PTL/PPROM group and TPL group. The most common organism identified in urine culture in both groups was Escherichia coli, presented in 40% and 75% in PTL/PPROM group and TPL group respectively.

TABLE 2. Urinary results of the study cohort.

Others various gastrointestinal bacteria were presented randomly in both groups. Remarkably, Streptococcus agalactiae (Streptococcus Group B; GBS) was found in 2 cases who were diagnosed with PPROM. Membrane rupture increases risk of urinary ascending infection from a rectovaginal GBS colonization.

Pregnancy outcomes were analyzed according to the urinalysis results in Table 3. Delivery rates in the studied admission were similar between women with abnormal and normal urinalysis and between women with positive and negative culture. Pregnancies with abnormal urinalysis had the gestational age at delivery and newborn birth weight greater than the corresponding values of the normal urinalysis group. This finding was despite the delivery rate and mode of the 2 groups being approximately the same. Unfortunately, when breaking into more details in gestational age or birth weight, there was no clear or consistent pattern to draw conclusion from. Other neonatal outcomes, such as the rates of birth asphyxia and neonatal intensive care unit (NICU) admission, were not significantly different.

Table 4 reveals pregnancy outcomes according to the urine culture results. Similar to urinalysis, delivery rates in the studied admission were similar between the women with a positive and a negative culture. On the other hand, pregnancies with a positive urine culture resulted in newborns with higher rates of extremely low birth weight (< 1,000 g) and NICU admission than cases with a negative urine culture. Even though the mode of delivery of these 2 groups was statistically similar, the cesarean section rate of the positive urine culture group was distinctly higher than that of the MUC-negative group (77.8% vs 52.5%).

From Tables 3 and 4, urine culture appears more meaningful than urinalysis. However, universal urine culture in cases with premature uterine contractions to identify the 2.4% positive rate would be costly. Based on the obtained data, we looked for a strategy to use the urinalysis as a screening tool to reduce the number of urine culture. Considering TPL comprising two- thirds of cases with premature uterine contractions, we concentrated on TPL group. In this group, simultaneous

TABLE 3. Pregnancy outcomes according to urinalysis results.

Birth asphyxiaa : Apgar score at 5 min after birth < 7

TABLE 4. Pregnancy outcomes according to results of urine culture.

Birth asphyxiaa : Apgar score at 5 min after birth < 7

urinalysis and culture were performed in 1,206 cases, urinalysis only in 269 cases and urine culture only in 1 case. Of the 1,206 cases who had both modalities of urine testing, the urine culture was positive in 33 cases. Within these 33 cases, 19 had an abnormal urinalysis based on WBC or RBC counts in the urine sediment. Thus, “abnormal urinalysis” using these criteria would detect only 57.6% of positive urine culture cases. Another marker from urinalysis was needed to improve sensitivity and the presence of bacteria in the sediment at the time of urinalysis was added into consideration. Urinalysis with abnormal cell count and/or presence of bacteria of ≥ 2+ was considered a positive screening. Urinalysis without significant cell count and significant presence of bacteria was considered negative. With this urinary screening criteria, a positive urine culture could be identified with the performance as shown in Table 5. The urine culture would have been performed in 78 cases (6.5%) with 90.9% sensitivity.

DISCUSSION

The overall prevalence of positive urine culture and abnormal urinalysis among pregnant women with premature uterine contractions was 2.4% and 4.1%, respectively. Considering in combination, UTI was present in 4.9%. The prevalence of abnormal urinalysis and positive urine culture were not significantly different between the PTL/PPROM and TPL groups. The most common organism identified was Escherichia coli in both groups. Newborns of women with positive urine culture had higher rates of extremely low birth weight (birth weight < 1,000 g) and admission to the NICU than newborns of women with negative urine culture. Although ASB screening in early pregnancy is recommended as standard care, its benefits are still controversial. The reported prevalence of ASB among preterm labor or presumptive preterm labor groups has varied between studies.14-16,20 The current investigation revealed a lower prevalence of positive urine culture

TABLE 5. Urinalysis as a screening tool for urine culture in TPL cases.

Sensitivity: 90.9%; Specificity: 95.9%;

Negative predictive value: 99.7%; Positive predictive value: 38.5%

Positive urinalysis screening: Urinalysis with abnormal cell count and/or presence of bacteria of ≥ 2+

Negative urinalysis screening: Urinalysis without significant cell count and significant presence of bacteria (≤ 1+)

(2.4% vs 5%) among pregnant women admitted with premature uterine contractions than an older study at our institute.15 More recent research at our hospital in 2015 revealed a comparable prevalence among pregnancies in any trimester without premature uterine contraction.9 The prevalence might have decreased due to different urine collection techniques, which changed from voluntary urination to clean urinary catheterization.

The most common organism for positive urinary culture in the present investigation was Escherichia coli, which was similar to the findings of other investigations.9,15,20,21 The current work also found 2 women with Streptococcus agalactiae (Group B Streptococcus; GBS), both presented with PPROM which might have increased the risk of urinary infection from the rectovaginal colonization of GBS. Antenatal screening of GBS colonization and the use of an appropriate intrapartum antibiotic prophylaxis should be considered to prevent newborn infections.22 This study showed that abnormal urinalysis and positive urine culture were not associated with delivery rate in that particular admission. This evidence agrees with other work in which an MUC was not recommended as a routine laboratory test in pregnancy with preterm labor without clinical indications.14 However, newborns of pregnancies with positive MUCs were related to poor neonatal outcomes, such as risk of extremely low birth

weight and NICU admission.7,8

Our study revealed a rather low prevalence of abnormal urinalysis (4.1%) and positive urine culture (2.4%) in pregnancies with premature uterine contractions. We propose a strategy to identify those who need a urine culture in women diagnosed with TPL instead of universal culture. Urinalysis is first performed and the results are used for contemplating urine culture. The cases with abnormal WBC or RBC counts, or cases a

bacteria rating of ≥ 2+ should proceed to urine culture, while those with normal cell counts and lower rate of bacteria in the sediment should not. With this strategy, only 6.5% of cases with TPL would need urinary culture with 90.9% sensitivity. This would save cost, time and manpower considerably. The evidence from this study can be used to guide the routine ordering of appropriate laboratory tests.

This study assessed the benefit of simultaneous urinalysis and urine culture in women with premature uterine contractions with a large number of subjects. Moreover, a strategy to reduce number of unnecessary urine culture in TPL cases has been proposed. However, this retrospective study was unable to demonstrate a causal relationship. Besides, selection bias from incomplete screening might have affected the results as urine culture was performed in 79% of the total cohort.

In conclusion, routine urinalysis and urine culture in women admitted with premature uterine contractions did not predict whether the pregnancy could be continued or not. However, adverse neonatal outcomes were more common in pregnancies with positive urine culture. Urinalysis result could be used to identify women with TPL who should have a urine culture to reduce unnecessary culture.

ACKNOWLEDGMENTS

This study was supported by Siriraj Research Development Fund (Managed by Routine to Research: R2R), Faculty of Medicine Siriraj Hospital, Mahidol University.

Conflict of interest

None declared.

References

1. Schmiemann G, Kniehl E, Gebhardt K, Matejczyk MM, Hummers-Pradier E. The diagnosis of urinary tract infection: a systematic review. Dtsch Arztebl Int. 2010;107(21):361-7.

2. Nicolle LE, Gupta K, Bradley SF, Colgan R, DeMuri GP, Drekonja D, et al. Clinical Practice Guideline for the Management of Asymptomatic Bacteriuria: 2019 Update by the Infectious Diseases Society of America. Clin Infect Dis. 2019;68(10):e83-e110.

3. Alvarez JR, Fechner AJ, Williams SF, Ganesh VL, Apuzzio JJ. Asymptomatic bacteriuria in pregestational diabetic pregnancies and the role of group B streptococcus. Am J Perinatol. 2010; 27(3):231-4.

4. Moore A, Doull M, Grad R, Groulx S, Pottie K, Tonelli M, et al. Recommendations on screening for asymptomatic bacteriuria in pregnancy. CMAJ. 2018;190(27):E823-E30.

5. Smaill FM, Vazquez JC. Antibiotics for asymptomatic bacteriuria in pregnancy. Cochrane Database Syst Rev. 2019;2019(11):CD000490.

6. Force USPST, Owens DK, Davidson KW, Krist AH, Barry MJ, Cabana M, et al. Screening for Asymptomatic Bacteriuria in Adults: US Preventive Services Task Force Recommendation Statement. JAMA. 2019;322(12):1188-94.

7. Kazemier BM, Koningstein FN, Schneeberger C, Ott A, Bossuyt PM, de Miranda E, et al. Maternal and neonatal consequences of treated and untreated asymptomatic bacteriuria in pregnancy: a prospective cohort study with an embedded randomised controlled trial. Lancet Infect Dis. 2015;15(11):1324-33.

8. Angelescu K, Nussbaumer-Streit B, Sieben W, Scheibler F, Gartlehner G. Benefits and harms of screening for and treatment of asymptomatic bacteriuria in pregnancy: a systematic review. BMC Pregnancy Childbirth. 2016;16(1):336.

9. Srisompong J, Rahman S, Russameecharoen K, Tongsai S, Seenama C, Koomanachai P. Prevalence and Accuracy of Screening Test of Asymptomatic Bacteriuria During Pregnancy in Siriraj Hospital. Open Forum Infect Dis. 2017;4(Suppl 1): S348.

10. Chawanpaiboon S, Kanokpongsakdi S. Preterm Birth at Siriraj Hospital: A 9-Year Period Review (2002-2010). Siriraj Med J. 2011;63(5):143-6.

11. Bentley DL, Bentley JL, Watson DL, Welch RA, Martin RW, Gookin KS, et al. Relationship of uterine contractility to preterm labor. Obstet Gynecol. 1990;76(1 Suppl):36S-8S.

12. American College of Obstetricians and Gynecologists’ Committee on Practice B-O. Prediction and Prevention of Spontaneous Preterm Birth: ACOG Practice Bulletin, Number 234. Obstet Gynecol. 2021;138(2):e65-e90.

13. The Management of Preterm Labour and Preterm Premature Rupture of Membranes [Internet]. RTCOG Clinical Practice Guideline. 2017. Available from: http://www.rtcog.or.th/home/ ob-014-%e0%b8%81%e0%b8%b2%e0%b8%a3%e0%b8%94

    %e0%b8%b9%e0%b9%81%e0%b8%a5%e0%b8%a3%e0%b8

    %b1%e0%b8%81%e0%b8%a9%e0%b8%b2%e0%b8%a0%e0

    %b8%b2%e0%b8%a7%e0%b8%b0%e0%b9%80%e0%b8%88

    %e0%b9%87%e0%b8%9a%e0%b8%84/396/.

14. Bastek JA, Sammel MD, Srinivas SK, Elovitz MA. Is routine infectious and toxicologic screening in preterm labor effective in predicting preterm birth? Am J Obstet Gynecol. 2008;198(5): e38-42.

15. Chalermchockchareonkit A, Phoethong S, Ruangvutilert P, Thamkhantho M. Prevalence of positive culture of genitourinary tract microorganisms in pregnant women with presumptive preterm labor. J Med Assoc Thai. 2013;96(9):1111-8.

16. Hundley AF, Onderdonk AB, Greenberg JA. Value of routine urine culture in the assessment of preterm labor. J Reprod Med. 2003;48(11):853-7.

17. Komaroff AL. Acute dysuria in women. N Engl J Med. 1984;310(6):368-75.

18. Cohen RA, Brown RS. Clinical practice. Microscopic hematuria. N Engl J Med. 2003;348(23):2330-8.

19. Kiatsuda D, Thinkhamrop J, Prasertcharoensuk W. Success rate in preterm uterine contraction inhibition with tocolytic agents in a tertiary care center. Int J Womens Health. 2016;8: 663-7.

20. Kamel HAH, Hegab MHM, Al-sehrawey AA-s, Hassan HM. Prevalence of Asymptomatic Bacteriuria in Patients with Preterm Labor. The Egyptian Journal of Hospital Medicine. 2018;73(9): 7444-7.

21. Nicolle LE. Asymptomatic bacteriuria: when to screen and when to treat. Infect Dis Clin North Am. 2003;17(2):367-94.

22. Prevention of Group B Streptococcal Early-Onset Disease in Newborns: ACOG Committee Opinion, Number 797. Obstet Gynecol. 2020;135(2):e51-e72.

Efficacy of Music Therapy and Zingiber officinale Roscoe Aromatherapy for Reducing Pain during the First Stage of Labor: A Randomized Controlled Trial

Pontip Paoin, M.D.*, Pichita Prasongvej, M.D.*, Athita Chanthasenanont, M.D.*, Titchayakorn Niumpradit, M.D.**, Densak Pongrojpaw, M.D.*, Komsun Suwannarurk, M.D.*

*Department of Obstetrics and Gynecology, Thammasat University, Pathum Thani 12120, Thailand. , **Chulabhorn International College of Medicine, Thammasat University, Pathum Thani 12120, Thailand.

ABSTRACT

Objective: To study the pain-reducing effect of music therapy and aromatherapy with ginger essential oil (GEO; Zingiber officinale Roscoe) in the first stage of labor.

Materials and Methods: This randomized controlled trial was conducted from May 2022 to March 2023 in the delivery room of Thammasat University Hospital in Thailand. Participants were nulliparous pregnant women who came with actual labor pain and cervical dilation of 3 cm. They were divided into three groups: control, music, and aromatherapy. The control group received intrapartum care according to hospital guidelines, while participants in the music group listened to a “Musical Journey through Pregnancy” music playlist, and the aromatherapy group inhaled GEO. A visual analog scale was used for pain assessment every 2 hours. Interventions and pain assessment continued until the end of the first stage of labor or until cesarean section was indicated.

Results: Three hundred pregnant women were recruited. The mean age of the participants was 27.7 years old, and no difference in BMI among the three groups. Participants in the music and aromatherapy group had statistically significantly less labor pain than participants in the control group when cervical dilation was between 4 and 7 cm (6,1±0.7, 1.9±0.7, 2.6±0.4 when cervical dilation was 4-5 cm and 8.0±0.7, 2.8±0.4, 3.6±0.3 when cervical dilation was 6-7 cm in the control, music, and aromatherapy groups, respectively (p < 0.001). However, at cervical dilation of 8-10 cm, there was no difference among the three groups regarding labor pain. Both intervention groups had significantly shorter labor time and less estimated blood loss during vaginal delivery than the control group (p < 0.001).

Conclusion: Music and aromatherapy decreased labor pain, labor time, and estimated blood loss compared to the control group.

Keywords: Labor pain; music therapy; aromatherapy (Siriraj Med J 2023; 75: 707-712)

INTRODUCTION

Labor pain is one of the most painful experiences in a woman’s life. It results from contraction of the uterus, dilation of the cervix, and dilation of the adjacent pelvic organs. This pain has significant physiological effects

on the health of the mother and fetus.1 Therefore, its minimization in the intrapartum period is considered crucial. Conventional methods can reduce pain during labor2 but also have disadvantages, such as decreased consciousness during labor, respiratory distress in the

Corresponding author: Pichita Prasongvej E-mail: pichita.pra@gmail.com

Received 27 June 2023 Revised 10 August 2023 Accepted 13 August 2023 ORCID ID:http://orcid.org/0000-0002-0485-4367 https://doi.org/10.33192/smj.v75i10.263860

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

newborn3, slowing of the fetal heart rate1, maternal respiratory suppression, nausea, vomiting, dizziness, and drowsiness.1,4 For these reasons, non-pharmacological methods such as music therapy, yoga, relaxation, and other methods are becoming more popular worldwide.5 Several studies have shown that music therapy can reduce anxiety, stress, and pain in pregnant women5-10 by stimulating the brain to produce dopamine and endogenous opioids.8 In addition, music therapy has been found to improve heart rate variability in pregnant women.9 Other studies have shown that geranium, frankincense, lavender, jasmine, and rose essential oils can relieve anxiety and labor pain.11,12 Ginger essential oil (GEO; Zingiber officinale Roscoe) is an essential oil extracted from the ginger plant and abundant in Southeast Asia. It is currently available worldwide and is used to relieve pain in various organs through various mechanisms.13 However, studies on labor pain after inhalation have not yet been conducted. These two alternative therapies are simple, inexpensive, and have no serious side effects. Therefore, we aim to investigate the efficacy of music therapy and GEO aromatherapy for pain relief in the

first stage of labor.

MATERIALS AND METHODS

This randomized controlled trial was conducted among pregnant women admitted to the delivery room of Thammasat University Hospital from May 2022 to

March 2023. A flowchart of the study according to the Consolidated Standards of Reporting Trials (CONSORT) is shown in Fig 1. The inclusion criteria were nulliparous, pregnant women aged 15 to 45 years old, singleton pregnancy, gestational age of ≥ 37 weeks, initial cervical dilation of 3 cm, use of oxytocin for labor augmentation, and ability to understand and communicate Thai language. Exclusion criteria were women who had an indication for cesarean section since admission, did not have a cephalic presentation, and had schizophrenia, anosmia, deafness, or chronic diseases that can exacerbate back pain or Covid-19 infection. Women who experienced adverse effects such as nausea and dizziness after inhaling GEO disliked the smell of GEO after using it for more than 10 minutes, experienced headaches or pain in their ears when listening to music for more than an hour, disliked the music, or were uncomfortable using headphones, or removed headphones for more than an hour were excluded from the study. Women who received opioid or regional anesthesia during delivery and women who wanted to end their participation in the study were also later excluded from the study.

After admission to the delivery room, research assistants presented and explained the research objectives to potential candidates who met the inclusion criteria. If they decided to participate in the study, their consent was obtained. Baseline data and clinical symptoms were recorded on a data collection form. Participants were

Fig 1. The CONSORT flow diagram of study participants

randomly assigned to three experimental groups: the control group, the music group, and the GEO aromatherapy group. The consecutively numbered assignments were sealed in opaque envelopes. Each group was placed in a different room. The control group received routine treatment, while the music group listened to a collection of music entitled “Musical Journey through Pregnancy”10 through headphones, consisting of 8 songs that were uplifting, gentle, and natural sounds. The aromatherapy group inhaled the GEO reed diffuser placed next to the bed, and participants in all groups were in different rooms. All three groups received interventions and were assessed for pain every 2 hours using a visual analog scale (VAS) until the end of the first stage of labor or cesarean section was indicated. Satisfaction was assessed after the end of labor. The pain scale was divided into mild (pain score 0-3), moderate (pain score 4-6), and severe (pain score 7-10) pain groups for further and more precise information.

Statistical analysis

The sample size was calculated using the formula with two independent groups, and the mean values of

3.82 and 4.39 were used from a previous study.14 With an alpha-type error rate of 1% (two-sided) and statistical power of 80% (one-sided), the estimated sample size per group was 93. With an expected sample loss of 5%, the

final size was adjusted to 100 cases per group, resulting in a total of 300 participants in the study.

Descriptive statistics were reported as mean, percentage, frequency, and standard deviation. Statistical Package for the Social Sciences version 23 (SPSS Inc., Chicago, IL, USA) was used for data analysis. One-way analysis ANOVA was performed, and a p-value < 0.05 was considered statistically significant.

RESULTS

A total of 300 nulliparous pregnant women were randomized equally to the control, music, and aromatherapy groups, with no participant withdrawal. The average age was 27.7 years old. The gestational age of each group was approximately 38 weeks. About half (145/300) of participants had a normal BMI. Almost all participants had no underlying diseases. There were no significant differences in participants’ characteristics between groups, as shown in Table 1.

Our study showed no differences in baseline VAS scores among the three groups. The mean VAS at cervical dilation of 4-5 cm were 6.1, 1.9, 2.6, and at the dilation of

6-7 cm, were 8.0, 2.8, and 3.6 for the control, music, and aromatherapy groups, respectively. The VAS score for cervical dilatations of 4-5 cm and 6-7 cm was significantly lower in both intervention groups than in the control group (p < 0.001), and the VAS score in the music group

TABLE 1. Demographic of the participants (100 case each).

*; Analysis of variance p-value <0.05 indicates statistically significant

Abbreviations: BMI: Body mass index, GDM; Gestational diabetes mellitus, PIH; Pregnancy induce hypertension

was significantly lower than in the aromatherapy group (p < 0.001).

In addition, the VAS score was further subdivided into severity levels. With cervical dilation of 4-5 cm, all participants in both intervention groups had mild pain. In contrast, most participants in the control group had moderate pain. At cervical dilation of 6-7 cm, all participants in the music group had mild pain; three- fourths of participants in the aromatherapy group also had mild pain, while most participants in the control group had severe pain, as shown in Table 2.

In both intervention groups, the first and second stages of labor were significantly shorter than in the control group. There were no differences in the rate of cesarean deliveries among the three groups, the most common indication being (15, 16, and 11 cases in the control, music, and aromatherapy groups, respectively, in that order). Another indication that led to cesarean delivery in the control group was abnormal fetal heart rate. Moreover, the estimated blood loss (EBL) for vaginal delivery was significantly lower in both intervention groups than in the control group (p < 0.001). However, no difference was found for cesarean delivery (p = 0.305). The mean birth weight was 3.0-3.1 kg in all groups. Almost all

neonates in all groups had standard Apgar scores in the first and fifth minutes of life, except for one case in the control group with an Apgar score of 5.6 in the first and fifth minutes of life, subsequently. All pregnant women in both intervention groups reported high satisfaction with a score of 7-10 out of a total of 10 points, while about four-fifths of participants in the control group were only moderately satisfied with a score of 4-6 out of a total of 10 points.

DISCUSSION

The management of labor pain is one of the significant issues in obstetrics, for which various methods of relief are being explored. Previous studies showed that noninvasive complementary and adjunctive strategies, namely music therapy and aromatherapy, significantly reduce initial labor pain7,15,19 (and other clinically relevant outcomes.10,15 In addition, many studies have found that both interventions can reduce anxiety in stressful situations.5,6,9,11,16,20 Currently, no study uses GEO inhalation and compares GEO aromatherapy with music therapy for pain relief in the first stage of labor. Therefore, we propose these two therapies for nulliparous pregnant women to treat labor pain in the first stage of labor.

TABLE 2. Pain score and severity score of the participants during various cervical dilatation.

*; Analysis of variance p-value <0.05 indicates statistically significant, §; Significant difference from the other 2 group, ¶; Significant difference from aroma group, Pain score 0-3; mild, 4-6; moderate, 7-10; severe pain

TABLE 3. Obstetrics data of the participants (100 case each).

*; Analysis of variance p-value <0.05 indicates statistically significant, †; mean± standard deviation, §; Significant difference from the other 2 group, ‡; n (%)

Abbreviations: C/D; cesarean delivery, V/DEBL; vaginal delivery estimate blood loss, BW; birth weight

In our study, participants were distributed among the research groups, and there were no significant differences in demographic characteristics. Most participants in the three groups had no underlying diseases, and there were no differences in cervical dilation, baseline pain scores, or augmentation rates.

Our study showed that participants in both intervention groups had significantly less labor pain than the control group when cervical dilation was 4-7 cm. Based on these results, we can assume that both music and aromatherapy are effective in reducing pain in the early active phase of the first stage of labor, which is consistent with previous studies in both the music7,17,18 and aromatherapy groups14,16,19 In addition, our study found that music therapy was more effective than GEO aromatherapy in reducing labor pain in the early active phase. The mechanism for pain relief in both interventions is unclear. However, music and aromatherapy may stimulate brain substances that serve as natural analgesics. However, in the late active phase of labor (cervical dilation of 8-10 cm), no significant difference in the severity of labor pain was observed among the three groups in our study. A lower focus on both interventions can explain this, as pain is more intense in the late active phase of labor.

Furthermore, we could hypothesize that music therapy is beneficial in shortening the duration of the first and second stages of labor, which is consistent with

Gonzalez’s study.10 Although this result contradicts a previous study by Guo15, which found that music can prolong the first stage of labor, these different results may be explained by differences in the type of music and the duration of music listening.

Additionally, GEO aromatherapy has significantly reduced the duration of the first and second stages of labor. This finding contrasts the results reported by Tanvisut14 and Yazdkhasti.19 It could be due to differences in the type of aromatherapy, the duration during which the participants were exposed to the fragrances, and the diffuser used.

Our study also showed that both intervention groups had a statistically significant reduction in EBL during vaginal delivery, consistent with the results reported by Guo.15 However, the mechanisms responsible for these results remained unexplained. Both intervention groups showed higher satisfaction levels than the control group, which may be due to the calming effect of these interventions, which helped to reduce stress during the waiting period.

Our study is the first to report the effectiveness of music therapy in reducing the time of the second stage of labor. In addition, our research reports for the first time that the scent of GEO can effectively reduce labor pain during the early active phase of the first stage of labor, the duration of labor, and EBL for vaginal birth.

Our results could be a milestone for future studies. At the same time, we are aware of the study’s limitations, such as the lack of blindness to interventions and the fact that it is a single-center design.

CONCLUSION

This study showed that music and GEO aromatherapy effectively reduced labor pain in the early active phase of the first stage of labor, with music therapy being an effective therapy. In addition, music and GEO aromatherapy reduced the duration of the first and second phases of labor, estimated blood loss during vaginal delivery, and had higher satisfaction levels among participants in both intervention groups. This study recommends both therapies as a cost-effective and complementary approach for treating initial labor pain in the delivery room.

ACKNOWLEDGMENTS

This study was supported by the Research Fund of the Faculty of Medicine, Thammasat University. The authors would like to thank Assoc. Prof. Dr. Dittakarn Boriboonhirunsarn for his kind help in preparing the manuscript.

Conflict of interest

Authors declare no conflict of interest for this article.

REFERENCES

1. Peter HP. Obstetric pain. In: McMahon S, Koltzenburg M, Tracey I, Turk DC, editors. Wall and Melzack's textbook of pain. 6th ed. Philadelphia: Elsevier Saunders; 2013. p.772-92.

2. Czech I, Fuchs P, Fuchs A, Lorek M, Tobolska-Lorek D, Drosdzol-Cop A, et al. Pharmacological and non-pharmacological methods of labour pain relief-establishment of effectiveness and comparison. Int J Environ Res Public Health. 2018;5:2792- 802.

3. Thomson G, Feeley C, Moran VH, Downe S, Oladapo OT. Women's experiences of pharmacological and non-pharmacological pain relief methods for labour and childbirth: a qualitative systematic review. Reprod Health. 2019;16:71-90.

4. Zanfini BA, Catarci S, Vassalli F, Laurita Longo V, Biancone M, Carducci B, et al. The effect of epidural analgesia on labour and neonatal and maternal outcomes in 1, 2a, 3, and 4a Robson's classes: a propensity score-matched analysis. J Clin Med. 2022; 11:614-32.

5. Smith CA, Levett KM, Collins CT, Armour M, Dahlen HG, Suganuma M. Relaxation techniques for pain management in

   labour. Cochrane Database Syst Rev. 2018;3:CD009514.

6. Santiváñez-Acosta R, Tapia-López ELN, Santero M. Music therapy in pain and anxiety management during labor: A systematic review and meta-analysis. Medicina (Kaunas). 2020; 56:526-36.

7. Buglione A, Saccone G, Mas M, Raffone A, Di Meglio L, Toscano P, et al. Effect of music on labor and delivery in nulliparous singleton pregnancies: a randomized clinical trial. Arch Gynecol Obstet. 2020;301:693-8.

8. Lunde SJ, Vuust P, Garza-Villarreal EA, Vase L. Music-induced analgesia: how does music relieve pain. Pain. 2019;160:989-93.

9. Teckenberg-Jansson P, Turunen S, Pölkki T, Lauri-Haikala M-J, Lipsanen J, Henelius A, et al. Effects of live music therapy on heart rate variability and self-reported stress and anxiety among hospitalized pregnant women: a randomized controlled trial. Nord J Music Ther. 2019;28:7-26.

10. Garcia Gonzalez J, Ventura Miranda MI, Requena Mullor M, Parron Carreno T, Alarcon Rodriguez R. Effects of prenatal music stimulation on state/trait anxiety in full-term pregnancy and its influence on childbirth: a randomized controlled trial. J Matern Fetal Neonatal Med. 2018;31:1058-65.

11. Tabatabaeichehr M, Mortazavi H. The effectiveness of aromatherapy in the management of labor pain and anxiety: a systematic review. Ethiop J Health Sci 2020;30:449-58.

12. Bertone AC, Dekker RL. Aromatherapy in obstetrics: a critical review of the literature. Clin Obstet Gynecol. 2021;64:572-88.

13. Andrei C, Zanfirescu A, Nițulescu GM, Negreș S. Understanding the molecular mechanisms underlying the analgesic effect of ginger. Nutraceuticals. 2022;2:384-403.

14. Tanvisut R, Traisrisilp K, Tongsong T. Efficacy of aromatherapy for reducing pain during labor: a randomized controlled trial. Arch Gynecol Obstet. 2018;297:1145-50.

15. Guo H, Que M, Shen J, Nie Q, Chen Y, Huang Q, et al. Effect of music therapy combined with free position delivery on labor pain and birth outcomes. Appl Bionics Biomech. 2022;2022:8963656.

16. Hamdamian S, Nazarpour S, Simbar M, Hajian S, Mojab F, Talebi A. Effects of aromatherapy with Rosa damascena on nulliparous women's pain and anxiety of labor during first stage of labor. J Integr Med. 2018;16:120-5.

17. Hosseini SE, Bagheri M, Honarparvaran N. Investigating the effect of music on labor pain and progress in the active stage of first labor. Eur Rev Med Pharmacol Sci. 2013;17:1479-87.

18. Xavier T, Viswanath L. Effect of music therapy on labor pain among women in active labor admitted in tertiary care hospital Kochi. Int J Intg Med Sci. 2016;3:444-8.

19. Yazdkhasti M, Pirak A. The effect of aromatherapy with lavender essence on severity of labor pain and duration of labor in primiparous women. Complement Ther Clin Pract. 2016;25 81.

20. Wongdama S, Siriussawakul A, Ratta-apha W, Suraprasit P, Kanjanapiboon K, Thanakiattiwibun C, et al. Effects of Music on Preoperative Anxiety in Patients Undergoing Hair Transplantation. Siriraj Med J. 2023;75:13-9.

Heart Rate Variability and Baroreflex Sensitivity: Factors and Reference Ranges

Paveena Somboon, M.Sc., Surat Kulapatana, M.D., Wattana B. Watanapa, M.D., Ph.D., Tanchanok Chattaris, M.D., Phatadon Sirivongrangson, M.D., Panomporn Livikorn, B.Sc., Thongchai Nonthadid, M.Sc., Poungpetch Hussarin, M.Sc.

Department of Physiology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand.

ABSTRACT

Objective: Heart rate variability (HRV) and baroreflex sensitivity (BRS) are influenced by various internal and external factors, making it necessary to establish universal reference values. We aimed to identify independent factors affecting short-term HRV parameters and BRS, as well as determine appropriate reference ranges, which are lacking in Asian populations.

Materials and Methods: A total of 117 healthy Thai participants (aged 20-72 years; final n=117) were recruited based on strict criteria, including normal medical history, physical examination (ascertained by a physician), and normal blood profile. Spontaneous 5-minute HRV parameters (time and frequency domains) and spontaneous cross-correlation BRS were measured.

Results: Age and resting heart rate (HR) are independently and inversely correlated with BRS and all HRV parameters, except LFnu (normalized low-frequency component) and LF/HF ratio (ratio of low- to high-frequency component). Sex differences were observed only in LF, which was higher in men. However, multiple regression analysis showed that sex did not significantly contribute as an independent variable to either HRV or BRS. Remarkably, BRS exhibited moderate to strong correlations with all HRV values, indicating its prominent role in influencing HRV and surpassing the impact of age.

Conclusion: Age and resting HR were identified as independent factors influencing 5-minute HRV and BRS. Our findings suggest that decreased baroreflex function associated with aging may contribute to reduced HRV among the elderly. Furthermore, we established reference ranges for each 5-minute HRV parameter and BRS, categorized by age group (20-39, 40-59, and ≥60 years old). These reference values offer valuable clinical insights, particularly for Asian populations where such normative ranges were previously unavailable.

Keywords: Heart rate variability; baroreflex sensitivity; reference values; healthy subjects; cardiovascular autonomic assessment (Siriraj Med J 2023; 75: 713-724)

INTRODUCTION

Heart rate variability (HRV) and baroreflex sensitivity (BRS) are quantitative markers that represent cardiovascular autonomic functions.1-3 HRV measures the oscillation of interbeat (RR) intervals, which reflects the central sympathetic and parasympathetic modulation of the

heart.4 BRS refers to the relationship between changes in systolic blood pressure (SBP) and corresponding RR intervals. It is widely accepted that hemodynamic parameters and derivatives, including HRV and BRS, are influenced by various internal and external factors, such as age, sex, ethnicity, recording procedures, surrounding

Corresponding author: Wattana B. Watanapa E-mail: wattana.wat@mahidol.ac.th

Received 5 July 2023 Revised 10 August 2023 Accepted 20 August 2023 ORCID ID:http://orcid.org/0000-0001-6206-4980 https://doi.org/10.33192/smj.v75i10.264037

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

temperature and environment.1,2,5-11 Consequently, it is necessary to establish reference ranges for HRV and BRS that are specific to each laboratory to ensure meaningful interpretation, and facilitate inter-laboratory comparisons.12 However, there are considerable variations in HRV and BRS reference values, even among Asian individuals10,13-15, likely due to disparities in recording duration, analysis methods, and other factors.5,12,16

HRV and BRS have been increasingly employed in the clinical setting, mostly in clinical trials, e.g., as risk assessment tools in coronary artery diseases.17,18 HRV is sometimes used as an additional investigation in patients with atypical postural orthostatic tachycardia syndrome or POTS.19 Moreover, their recording practicality have made both HRV and BRS widely used for other purposes besides clinical applications. For example, HRV is used as an objective stress measurement in e-sport players20 and training monitoring parameters in sports science.21 This popularity, however, resulted in a huge variation in procedure and outcome, leading to difficulties in interpretation, particularly when comparing among studies without reference values.

Resting, short-term assessments of HRV (5 minutes) and BRS (10 minutes) are noninvasive and practical for clinical investigations.1,2,12,17-19,22 However, standardized 5-minute HRV reference values for Asian adults are currently lacking age stratification13 or fail to demonstrate age dependence.20 While a few studies in Thailand have compared HRV and BRS between patients and controls or different groups23-25, normal values for the Thai population have not yet been established, limiting the clinical use in the region. Our aim was to determine factors that influence short-term HRV parameters and BRS, and to establish reference values for healthy Thai adults.

MATERIALS AND METHODS

This study was approved by the Siriraj Institutional Review Board (COA no. Si 679/2012).

Participants

We enrolled healthy Thais, aged 20-80 years, from Bangkok. All participants provided written informed consent and met the following criteria: normal BMI (18.5-22.9 kg/m2)26, and blood profile within the last six months, including complete blood count (CBC), fasting blood glucose (FBG), blood urea nitrogen (BUN), creatinine, liver function tests, and lipid profile: triglyceride (TG) <200 mg/dL and total cholesterol <200 mg/dL, or ≥200mg/dL with calculated LDL (cal-LDL) <160 mg/dL, and HDL >50 mg/dL (women) or >40 mg/dL (men). The exclusion criteria was as follows: diabetes

mellitus; cardiovascular diseases (e.g., hypertension); abnormal electrocardiogram (ECG); pulmonary diseases, including obstructive sleep apnea; psychiatric diseases; use of drugs that affect the autonomic nervous system (ANS), with the exception of statins and baby aspirin; current smokers or individuals who had quit smoking

<1 year; heavy alcohol drinkers (>2 standard drinks per day); individuals with occupations that affect the ANS, such as security guards and athletes; acute illness; and pregnancy. All female participants were assessed outside their menstruation period.

Data collection

Participants avoided alcoholic beverages for ≥48 hours, and caffeinated beverages, vitamins, herbs, and intense exercise for ≥12 hours. Upon arrival at the Cardiovascular Autonomic Function Laboratory, Department of Physiology, participants underwent an interview and a comprehensive physical examination conducted by a physician. Recorded data comprised participant demographics, including sex, age, BMI, exercise habits, alcohol consumption, menstruation, resting blood pressure (BP), and heart rate (HR) using an automatic sphygmomanometer.

The autonomic assessment took place in a quiet room, with a temperature of 23-24°C, humidity between 50-60%, and commenced at least ≥2 hours after the last meal. Noninvasive continuous hemodynamic parameters were measured using the Finometer® Pro device with an ECG module (lead II) and analyzed with the BeatScope® Easy program (Finapres Medical System BV, Amsterdam, The Netherlands), at a sampling rate of 200 Hz.

The protocol began with a 10-minute period of rest in a supine position, with system calibrations performed between the 6th and 10th minute. Subsequently, continuous measurements of blood pressure (BP) and ECG were recorded for 12 minutes, during which participants remained awake, quiet, and relaxed.

Data and statistical analysis

HRV analysis was performed using LabChart®7Pro software (ADInstrument, Castle Hill, Australia). A 5-minute segment of stable ECG data without, ectopic beats was manually selected to derive 11 time- and frequency- domain HRV parameters and BRS, as shown in Table 2. Statistical analyses were performed using SPSS software (IBM Corporation, New York, USA). Data normality was assessed using the Kolmogorov-Smirnov test. Results were reported as mean±SD, with additional median and interquartile range (P25-P75) provided for non-normally distributed data. For comparisons between

groups, the unpaired t-test or one-way ANOVA with post hoc Tukey’s test was used for normally distributed data, while the Mann-Whitney U test or Kruskal-Wallis test with post hoc Dunn’s test was applied for non-normally distributed data, as appropriate. A P value <0.05 was considered statistically significant.

Pearson correlation analysis was utilized to determine the linear correlation between continuous data and HRV parameters or BRS. Correlation coefficients (r) falling within the range 0.36≤ |r| <0.68 or |r| ≥0.68 were regarded as indicating moderate or strong correlations, respectively. 27 Factors that exhibited significant differences in parameters between groups or significant correlations with HRV or BRS values were further included in a stepwise multiple regression analysis to identify independent determinants. Lastly, reference values for each parameter were determined based on percentile values and categorized as normal, borderline, or abnormal as described in Table 5.

RESULTS

Participant general characteristics

Of the 144 initially interviewed healthy participants, 27 individuals were excluded due to abnormal blood profiles or physical examination findings. The general characteristics of the remaining 117 participants (44 men and 73 women) are summarized in Table 1. The mean age was similar between men and women. Men exhibited higher BMI, BP, and respiratory rate (RespR), but had

significantly lower resting HR. All blood profile data were within normal limits, consistent with the inclusion criteria (data not shown). The participants were divided into three age groups: 20-39, 40-59, and 60-79 years (n=56, 40, and 21, respectively). In the two younger age groups, the average ages did not differ significantly between sexes, while in the oldest group, women were older than men (average age, women vs. men, 64.41 vs. 60.50 years, n=17 vs. 4, respectively) (Supplemental Materials, Table S.1).

Comparisons of HRV and BRS among independent groups

The comparison of HRV parameters and BRS among different age groups is presented in Table 2. Most parameters showed a declining trend with age, except for LFnu, HFnu, and LF/HF, which were not dependent on age. The 40-59 age group generally did not differ significantly from the 60-79 age group, except for SDNN, Total power, and LF, which were higher in the 40-59 age group. Interestingly, we observed no significant differences between sexes in all parameters except for LF, which was higher in men (Table S.2).

We also examined HRV and BRS differences in relation to other characteristics. However, no significant differences were found when caffeine drinkers vs. non- caffeine drinkers, non-heavy alcohol drinkers vs. non- alcohol drinkers, regular exercise vs. sedentary individuals,

TABLE 1. Participants’ general characteristics.

Characteristics Men (n = 44) Women (n = 73) P value

BMI (kg/m2)

21.29 ± 1.21,

21.65 (20.22 - 22.32)

20.75 ± 1.09

0.0127b

Age (year) 39.00 ± 12.82 44.11 ± 14.94 nsa

SBP (mmHg) 118.16 ± 10.50 109.89 ± 11.77, 0.0002b

108.00 (101.00 - 115.50)

DBP (mmHg) 73.39 ± 8.10 69.01 ± 8.69 0.0079a

HR (/min)

67.25 ± 8.20

72.10 ± 9.75,

72.00 (66.00 - 76.50)

0.0053b

MAP (mmHg) 88.31 ± 8.19 82.64 ± 8.97 0.0009a

RespR (/min) 15.98 ± 9.69, 15.38 ± 2.14, 0.0040b

13.00 (12.00 - 16.00) 16.00 (14.00 - 16.00)

Data are mean±SD. Medians (P25 - P75) are additionally shown for non-normally distributed data. aUnpaired t-test (normally distributed data); bMann-Whitney U test (non-normally distributed data); ns, not statistically significant or P ≥ 0.05.

TABLE 2. Comparison of HRV parameters and BRS among age groups.

Parameters 20-39 years (n=56) 40-59 years (n=40) 60-79 years (n=21) P value

HRV

Data are mean±SD. Medians (P25 - P75) are additionally shown for non-normally distributed data. aOne-way ANOVA followed by post hoc;

bKruskal-Wallis test followed by post hoc Dunn's test (non-normally distributed data); *,#Significant difference compared to age groups

*20-39 years or #40-59 years, respectively; ns, not statistically significant or P ≥ 0.05.

Abbreviations: SDNN, standard deviation of the NN interval (NN stands for normal-to-normal RR interval); RMSSD, the root mean square of differences between adjacent NN intervals; SDSD, standard deviation of differences between adjacent NN intervals; pNN50, proportion of NN50 to total NN intervals (NN50 stands for number of adjacent NN interval pairs with a difference greater than 50 ms); VLF, very low frequency (≤ 0.4 Hz); LF, low frequency (0.04-0.15 Hz); HF, high frequency (0.15-0.4 Hz); LFnu, normalized unit of LF; HFnu, normalized unit of HF.

women in the follicular vs. the luteal phase of the menstrual cycle, and morning vs. afternoon assessment time (Table S.3-S.7).

Correlation analysis

Correlation analysis was conducted between each HRV parameter or BRS and all factors, including hemodynamic and blood profile measures. Age showed a moderate negative correlation with all absolute HRV parameters (excluding LFnu, HFnu, and LF/HF) and BRS. Additionally, a weak negative correlation was observed between age and HFnu, while a weak positive correlation was found between age and LFnu or LF/HF. Resting HR, but not BP, exhibited a moderate inverse correlation with three time-domain HRV indices, namely SDSD, RMSSD, and

pNN50. Furthermore, BRS displayed a stronger negative correlation with SBP than resting HR (Table 3). Other factors not shown in the table, including blood profile data, either showed no significant correlation with HRV parameters or BRS (P ≥0.05), or displayed significant but weak correlations (P <0.05, |r| <0.36).

Stepwise multiple regression

To determine independent influences on each HRV or BRS parameter, stepwise multiple regression analysis was conducted. All factors with any significant correlation or differences between groups were included in the analysis, resulting in a total of 22 parameters: sex (men=0, women=1), age, BMI, weight, height, SBP, DBP, MAP, HR, RespR, ambient temperature, humidity, FBG,

TABLE 3. Correlations between HRV parameters or BRS vs age, BMI, resting BP, and HR.

Parameters Age BMI SBP DBP MAP HR

HRV

r, Pearson’s correlation coefficient; ns, not statistically significant or P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Significant correlations with |r| > 0.36 are shown in bold type.

TG, cal-LDL, ALT, Hb, Hct, RBC, WBC, neutrophil (N), and eosinophil (%E). Table 4 displays the multiple linear equations and their adjusted coefficient of determination (Adj. R2) for each HRV and BRS value. The independent variables in each equation are presented in order of their contribution size to the regression model. Age was the main contributing variable for all time-domain HRV parameters, Total power, LF, HF, and BRS, followed by resting HR. Interestingly, sex did not emerge as a significant variable in any of the equations, suggesting that any differences in HRV parameters between sexes may be an epiphenomenon. It is worth noting that none of the adjusted R2 exceeded 0.50, indicating that the factors included in the study accounted for less than

50% of the variance in each parameter. It is possible that other factors not considered in our study may play a role.

Relationship between HRV parameters and BRS

BRS exhibited a strong positive correlation with all time-domain HRV parameters. It also displayed positive correlations with most frequency-domain parameters, except for LFnu and LF/HF ratio, which showed an inverse correlation (Table S.8). Among the HRV parameters, RMSSD (r=0.86), SDSD (r=0.86), pNN50 (r=0.85), and

HF (r=0.80), which are associated with parasympathetic activity, showed the strong positive correlations with BRS. LF exhibited a moderate correlation (r=0.59). Given

TABLE 4. Stepwise multiple regression results for HRV parameters and BRS.

Regression equation Adj. R2

HRV

Adj. R2, adjusted coefficient of determination; italicized numbers in parentheses are F-to-remove of the parameter immediately above.

these robust associations, we conducted a reanalysis by incorporating HRV and BRS into the stepwise multiple regression calculation for each other (Table S.9A and S.9B). The inclusion of both variables resulted in improved predictive power (Adj. R2) in every model. Interestingly, BRS emerged as the most influential variable for predicting all HRV parameters, replacing the variance explained by age (Table S.9A). Furthermore, RMSSD, a vagal HRV parameter, contributed the most to the BRS value (Table S.9B).

Age-specific reference values of HRV parameters and BRS (Table 5)

The reference values for HRV parameters and BRS were categorized by age group, except for LFnu and LF/ HF, which were age-independent (Table 4). These values were organized into five intervals based on percentile, as described in the Materials and Methods section. As expected, the age-dependent parameters were highest in the youngest age group.

DISCUSSION

In this study, we examined short-term HRV (5 minutes) and spontaneous cross-correlation BRS in 117 healthy Thai volunteers (44 men and 73 women, 20-72 years). Our findings revealed that BRS and most HRV parameters, namely, SDNN, SDSD, RMSSD, pNN50, Total power, LF, HF, and HFnu, were negatively and independently correlated with age and resting HR. However, factors such as sex, most blood profile components, caffeine intake, alcohol drinking, exercise, menstrual phase, and assessment time did not show any significant associations with HRV parameters and BRS.

Major determinants of HRV and BRS are age and resting heart rate.

Age and resting heart rate are major determinants of HRV parameters and BRS in healthy subjects. The negative correlation between age and all parameters, except for LFnu and LF/HF, was observed in both the correlation study and stepwise multiple analysis (Table 3 and 4). These results are consistent with several studies reporting a decline in HRV and BRS with age in healthy participants5,14,28-31 with our comparison of HRV parameters among age groups (Table 2). Our rigorous inclusion and exclusion criteria ensured the selection of a homogeneous pool of healthy subjects, indicating that the negative correlation with age represents a normal aging process rather than a pathology.32 The absence of an age association with LFnu and LF/HF was supported by the stepwise multiple regression results (Table 4), aligning with the findings

of Remaekers et al.33 Furthermore, the differences were primarily observed between the youngest group (20-39 years) and the two older groups. The average or median values of HRV and BRS in the age group 40-59 and

≥60 were generally similar, especially parasympathetic- associated HRV parameters: RMSSD, pNN50, and HF. The two older groups only differed significantly in SDNN, Total power, and LF (Table 2). Although this discrepancy could be attributed to the small sample size in the oldest age group, a study has shown that older subjects (60-77) did not differ significantly from the oldest middle-age group (40-59), which supports our results.28 Possible explanations include a faster decline in HRV values with age in younger individuals32 and a more linear correlation between sympathetic modulation, global autonomic regulation HRV (LF and SDNN, respectively) and age compared to parasympathetic modulation.14,34

Another significant independent factor affecting HRV and BRS, as observed in our data, is resting HR. While this factor is less frequently reported6,33,35, we found a negative correlation between resting HR and most HRV parameters and BRS, similar to the age effect but of smaller magnitude (Table S.9A and S.9B). Our results are generally in agreement with Ramaekers et al.33, who reported negative correlations between HR and SDNN, SDSD, pNN50, RMSSD, Total power, LF, and HF, except for LFnu and LF/HF ratio. Since resting HR is associated with cardiac autonomic tone, while HRV parameters are more related to cardiac autonomic regulation4, the negative correlation suggests reduced cardiac autonomic modulation at higher autonomic tone. Alternatively, it may reflect limited variability at narrower RR intervals.36

BMI did not show a significant association with most HRV parameters and BRS in our study, and it only weakly influenced LFnu and HFnu. In contrast, a study by Vallejo et al.37 in 30 females (21-35 years; BMI ranging from <19 to >30kg/m2) found that both age and BMI significantly affected HRV. The discrepancy may be attributed to the narrower range of BMI in our healthy participants (18.5-22.9kg/m2), or the fact that the other study did not include resting HR in their analysis (which could be a determinant variable of BMI).

HRV parameters and BRS are probably not sex dependent. Most HRV parameters and BRS were not significantly different between men and women, except for significantly higher LF in men (Table S.2). Additionally, sex did not appear in any of the stepwise multiple regression equations (Table 4), indicating that sex was not an independent factor determining HRV parameters and BRS in healthy

TABLE 5. Normal, borderline, and abnormal values for HRV parameters and BRS.

TABLE 5. Normal, borderline, and abnormal values for HRV parameters and BRS. (Continue)

Normal, P10-P90; borderline low (BL) ≥ P2.5 to < P10; borderline high (BH) > P90 to ≤ P97.5; abnormally low (AL) < P2.5; abnormally high (AH)

> P97.5.; *, not defined.

subjects. The difference in LF between men and women could be attributed to the negative correlation between LF and resting HR (Table 3), as men tend to have lower resting HR than women (Table 1). Although many reports have described sex differences in HRV, especially in young adults12,13, most of these studies did not utilize multiple regression analysis in their investigations.

BRS is highly correlated with HRV.

BRS has been found to have a positive correlation with HRV in both normotensive and hypertensive subjects. It is also the most important determinant in predicting HRV in multiple regression analysis38,39, which is consistent with our findings. Although BRS and HRV both reflect cardiovascular autonomic functions, they represent

different aspects and are not redundant. Therefore, it is reasonable to consider their correlations.17

The parasympathetic components of HRV, such as RMSSD, SDSD, pNN50, and HF, exhibited higher correlations with BRS compared to sympathetic or global autonomic components like LF, SDNN, and Total power (Table S.8). These distinct correlations support the notion that activation and deactivation of sympathetic or parasympathetic control during blood pressure fluctuation are not mirror images. While the SA node is controlled by both parasympathetic and sympathetic cardiac fibers, the vagal response is much faster, leading to immediate changes in RR intervals in response to blood pressure fluctuations.2 Furthermore, it is important to note that our study, despite the ‘baroreflex

sensitivity’ name, specifically focused on the cardiovagal branch of the baroreflex arc, as we derived responses from RR intervals, while excluding changes in sympathetic vasomotor tone. Thus, it is not surprising that our results align with other studies showing a high correlation between BRS and cardiac vagal control.40

The high correlations observed between HRV parameters and BRS can provide insights into the mechanisms underlying reduced HRV with aging (Table 4). Studies associating HR and BP oscillations have suggested a close relationship between HRV and baroreceptor function.41,42 In our study, we attempted to integrate these two observations by including BRS in the multiple regression analysis to predict each HRV parameter and vice versa. The inclusion of BRS replaced all HRV variance previously accounted for by age and emerged as the most influential factor, exhibiting the highest proportion of squared semi-partial correlation for each HRV parameter (Table S.9A). In other words, the age-related decline in HRV can also be explained by the reduced d BRS. Therefore, a decline in baroreceptor or baroreflex function may contribute to the diminished HRV observed in older individuals.

Age-specific reference values of HRV parameters and BRS

The present study presents reference values for 5-minute HRV and cross-correlation BRS in a sample of healthy Thai individuals (Table 5). We identified age and resting HR as independent factors influencing HRV and BRS, while sex did not show a significant association. Therefore, we categorized the reference values into three age groups without considering sex stratification. The reference values were presented based on percentiles since most of the HRV and BRS data exhibited nonparametric distribution.

While many studies in Asian populations defined normal ranges for short-term HRV in terms of mean

± SD, with or without age specification13,14,43, the only other study reporting age-adjusted reference values in percentile in Asian individuals (in Brazil) was conducted by Dantas et al.44 Comparing our reference values in in the frequency domain with the Asian Brazilian data, we observed higher values. These differences could be attributed to variations in analytical methods or environmental factors among the study subjects.

We provided age-adjusted reference values for BRS using the spontaneous cross-correlation sequence method, which has been suggested to be more reliable than the sequential method.22 To the best of our knowledge, our reference values for cross-correlation BRS may be the

only ones reported for an Asian population. However, Tang et al.15 recently published spectral BRS reference values for a Chinese population. When compared to the ATRAMI study17, which reported BRS values in post-myocardial infarction patients, our lower limit for abnormal BRS (AL) was higher than the cut-point associated with increased cardiac mortality (3ms/mmHg) in those patients. This disparity can be explained by the ATRAMI study inclusion of non-healthy subjects, with a higher average age (57 years), and the use of phenylephrine injection to determine BRS.

Limitations

Our sampling rate for BP and ECG signal provided by the Finometer® Pro device was limited to 200 Hz maximum45, which was lower than the Task Force’s recommendation at 250-500 Hz.1 However, there is evidence that 200 Hz is acceptable for HRV and BRS analysis, especially for healthy subjects.46

We were unable to provide reference values based on age and resting HR, even though they are both independent variables in HRV and BRS regression equations, due to limited sample size, especially in the oldest age group. Another limitation of our study is that the reference values are specific to the spontaneous 5-minute HRV recording and cross-correlation BRS method. While these methods are valuable and practical in clinical settings, further research is needed to establish reference values for 24-hour HRV recording and spectral analysis BRS, specifically for the Thai population.

CONCLUSION

Our study provides age-dependent reference values for spontaneous 5-minute HRV and cross-correlation BRS in healthy Thai adults, aged 20-72, along with an understanding of the factors influencing them. We found that HRV parameters, except LFnu and LF/HF ratio, as well as BRS, were independently and negatively correlated with age and resting HR, while sex did not show a significant association. Moreover, our results suggest that the decline in HRV with age may be partially attributed to age-related reductions in baroreflex function. The reference values and insights gained from our data analysis will serve as a foundation for the interpretation of HRV and BRS values in autonomic research and clinical practice.

ACKNOWLEDGEMENTS

This study was supported by the Siriraj Graduate Scholarship (P.S.) and R2R grant (W.B.W.), Faculty of Medicine Siriraj Hospital. We appreciate valuable inputs from Prof. Rungroj Krittayaphong, Drs. Chulaluk

Komoltri, Sompol Tapechum, and Akarin Nimmannit. Participant screening was assisted by Drs. Jindapa Srikajon, Sirikan Luangthamma, Nattachat Namsawat, Rinlada Pongratanakul, and Anutr Khummongkol. Finally, we thank the Siriraj Elder Association, Nursing Elder Association, and Departments of Clinical Pathology, Preventive and Social Medicine, and Physiology, Faculty of Medicine Siriraj Hospital.

Conflict of interest statement

None to declare.

Abbreviations: %E, percent eosinophil; Adj. R2, adjusted coefficient of determination; AH, abnormally high; AL, abnormally low; ALP, alkaline phosphatase; ALT, alanine aminotransferase; ANOVA, analysis of variance; BH, borderline high; BL, borderline low; BMI, body mass index; BP, blood pressure; bpm, beat per minute; BRS, baroreflex sensitivity; BUN, blood urea nitrogen; cal-LDL, calculated low-density lipoprotein cholesterol; CBC, complete blood count; DBP, diastolic blood pressure; dl, deciliter; ECG, electrocardiogram; FBG, fasting blood glucose; h, hour; Hb, hemoglobin; Hct, hematocrit; HDL, high-density lipoprotein cholesterol; HF, high frequency (0.15-0.4 Hz); HFnu, high frequency power in normalized unit; HR, heart rate; HRV, heart rate variability; Hz, Hertz; kg, kilogram; LF, low frequency (0.04-0.15 Hz); LFnu, low frequency power in normalized units; m2, square meter; MAP, mean arterial pressure; mg, milligram; mmHg, millimeter of mercury; ms, millisecond; N, neutrophil count; NN interval, the interval between two successive normal R waves; NN50, number of adjacent NN interval pairs with a difference greater than 50 ms; pNN50, proportion of NN50 to total NN intervals; r, Pearson’s correlation coefficient; R2, coefficient of determination; RBC, red blood cell count; RespR, respiratory rate; RMSSD, root mean square of differences between adjacent NN intervals; RR interval, the interval between two successive R waves; SA node, sinoatrial node; SBP, systolic blood pressure; SD, standard deviation; SDNN, standard deviation of the NN interval; SDSD, standard deviation of differences between adjacent NN intervals; TG, triglyceride; VLF, very low frequency (≤0.4 Hz); WBC, white blood cell count

REFERENCES

1. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation. 1996;93(5):1043-65.

2. La Rovere MT, Pinna GD, Raczak G. Baroreflex sensitivity: measurement and clinical implications. Ann Noninvasive Electrocardiol. 2008;13(2):191-207.

3. Sztajzel J. Heart rate variability: a noninvasive electrocardiographic method to measure the autonomic nervous system. Swiss Med Wkly. 2004;134(35-36):514-22.

4. Malik M, Camm AJ. Components of heart rate variability-- what they really mean and what we really measure. Am J Cardiol. 1993;72(11):821-2.

5. Tank J, Baevski RM, Fender A, Baevski AR, Graves KF, Ploewka K, et al. Reference values of indices of spontaneous baroreceptor reflex sensitivity. Am J Hypertens. 2000;13(3):268-75.

6. Koskinen T, Kähönen M, Jula A, Laitinen T, Keltikangas- Järvinen L, Viikari J, et al. Short-term heart rate variability in healthy young adults: the Cardiovascular Risk in Young Finns Study. Auton Neurosci. 2009;145(1-2):81-88.

7. Schnell I, Potchter O, Epstein Y, Yaakov Y, Hermesh H, Brenner S, et al. The effects of exposure to environmental factors on Heart Rate Variability: an ecological perspective. Environ Pollut. 2013;183:7-13.

8. Latchman P, Gates G, Axtell R, Pereira J, Bartels M, De Meersman RE. Spontaneous baroreflex sensitivity in young normotensive African-American women. Clin Auton Res. 2013;23(4):209-13.

9. Reed KE, Warburton DE, Whitney CL, McKay HA. Differences in heart rate variability between Asian and Caucasian children living in the same Canadian community. Appl Physiol Nutr Metab. 2006;31(3):277-82.

10. Khan M, Hussain M, Bin Aleem S. Heart rate variability in healthy population. Pak J Physiol. 2010;6:40-2.

11. Chotruangnapa C, Thongdang P. The benefit of unattended automated office blood pressure measurement on the white- coat effect: a cross-sectional study. Siriraj Med J. 2023;75(6): 454-65.

12. Shaffer F, Ginsberg JP. An overview of heart rate variability metrics and norms. Front Public Health. 2017;5:258.

13. Kuo TB, Lin T, Yang CC, Li CL, Chen CF, Chou P. Effect of aging on gender differences in neural control of heart rate. Am J Physiol. 1999;277(6):H2233-9.

14. Park SB, Lee BC, Jeong KS. Standardized tests of heart rate variability for autonomic function tests in healthy Koreans. Int J Neurosci. 2007;117(12):1707-17.

15. Tang ZH, Zeng F, Ye K, Yu X, Zhou L. The analysis of a reference value for baroreflex sensitivity and cardiovascular autonomic neuropathy prevalence in a Chinese population. Eur J Med Res. 2014;19(1):8.

16. Nunan D, Sandercock GR, Brodie DA. A quantitative systematic review of normal values for short-term heart rate variability in healthy adults. Pacing Clin Electrophysiol. 2010;33(11): 1407-17.

17. La Rovere MT, Bigger JT, Jr., Marcus FI, Mortara A, Schwartz PJ. Baroreflex sensitivity and heart-rate variability in prediction of total cardiac mortality after myocardial infarction. ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) Investigators. Lancet. 1998;351(9101):478-84.

18. Goldenberg I, Goldkorn R, Shlomo N, Einhorn M, Levitan J, Kuperstein R, et al. Heart rate variability for risk assessment of myocardial ischemia in patients without known coronary artery disease: The HRV-DETECT (Heart Rate Variability for the Detection of Myocardial Ischemia) Study. J Am Heart Assoc. 2019;8(24):e014540.

19. Bryarly M, Phillips LT, Fu Q, Vernino S, Levine BD. Postural Orthostatic Tachycardia Syndrome: JACC Focus Seminar. J Am Coll Cardiol. 2019;73(10):1207-28.

    
    

20. Gündoğdu S, Çolak Ö H, Doğan EA, Gülbetekin E, Polat Ö. Assessment of mental fatigue and stress on electronic sport players with data fusion. Med Biol Eng Comput. 2021;59(9):1691- 707.

21. Dong JG. The role of heart rate variability in sports physiology. Exp Ther Med. 2016;11(5):1531-6.

22. Westerhof BE, Gisolf J, Stok WJ, Wesseling KH, Karemaker JM. Time-domain cross-correlation baroreflex sensitivity: performance on the EUROBAVAR data set. J Hypertens. 2004;22(7):1371- 80.

23. Krittayaphong R, Veerakul G, Nademanee K, Kangkagate C. Heart rate variability in patients with Brugada syndrome in Thailand. Eur Heart J. 2003;24(19):1771-8.

24. Rutjanaprom W, Kanlop N, Charoenkwan P, Sittiwangkul R, Srichairatanakool S, Tantiworawit A, et al. Heart rate variability in beta-thalassemia patients. Eur J Haematol. 2009;83(5):483-9.

25. Chanavirut R, Tong-Un T, Jirakulsomchok D, Wyss JM, Roysommuti S. Abnormal autonomic nervous system function in rural Thai men: A potential contributor to their high risk of sudden unexplained nocturnal death syndrome. Int J Cardiol. 2017;226:87-92.

26. World Health Organization. Regional Office for the Western

    P. The Asia-Pacific perspective: redefining obesity and its treatment. Sydney: Health Communications Australia; 2000.

27. Taylor R. Interpretation of the Correlation coefficient: a basic review. Journal of Diagnostic Medical Sonography. 1990;6(1): 35-9.

28. Laitinen T, Hartikainen J, Vanninen E, Niskanen L, Geelen G, Länsimies E. Age and gender dependency of baroreflex sensitivity in healthy subjects. J Appl Physiol (1985). 1998;84(2):576-83.

29. Gerritsen J, TenVoorde BJ, Dekker JM, Kingma R, Kostense PJ, Bouter LM, et al. Measures of cardiovascular autonomic nervous function: agreement, reproducibility, and reference values in middle age and elderly subjects. Diabetologia. 2003; 46(3):330-8.

30. Zhang J. Effect of age and sex on heart rate variability in healthy subjects. J Manipulative Physiol Ther. 2007;30(5):374-9.

31. Abhishekh HA, Nisarga P, Kisan R, Meghana A, Chandran S, Trichur R, et al. Influence of age and gender on autonomic regulation of heart. J Clin Monit Comput. 2013;27(3):259-64.

32. Jandackova VK, Scholes S, Britton A, Steptoe A. Are changes in heart rate variability in middle-aged and older people normative or caused by pathological conditions? Findings from a large population-based longitudinal cohort study. J Am Heart Assoc. 2016;5(2):e002365.

33. Ramaekers D, Ector H, Aubert AE, Rubens A, Van de Werf F. Heart rate variability and heart rate in healthy volunteers. Is the female autonomic nervous system cardioprotective? Eur

    Heart J. 1998;19(9):1334-41.

34. Almeida-Santos MA, Barreto-Filho JA, Oliveira JL, Reis FP, da Cunha Oliveira CC, Sousa AC. Aging, heart rate variability and patterns of autonomic regulation of the heart. Arch Gerontol Geriatr. 2016;63:1-8.

35. Tsuji H, Venditti FJ, Jr., Manders ES, Evans JC, Larson MG, Feldman CL, et al. Determinants of heart rate variability. J Am Coll Cardiol. 1996;28(6):1539-46.

36. Coumel P, Maison-Blanche P, Catuli D. Heart rate and heart rate variability in normal young adults. J Cardiovasc Electrophysiol. 1994;5(11):899-911.

37. Vallejo M, Márquez MF, Borja-Aburto VH, Cárdenas M, Hermosillo AG. Age, body mass index, and menstrual cycle influence young women’s heart rate variability --a multivariable analysis. Clin Auton Res. 2005;15(4):292-8.

38. Mancia G, Parati G, Pomidossi G, Casadei R, Di Rienzo M, Zanchetti A. Arterial baroreflexes and blood pressure and heart rate variabilities in humans. Hypertension. 1986;8(2):147- 53.

39. Hesse C, Charkoudian N, Liu Z, Joyner MJ, Eisenach JH. Baroreflex sensitivity inversely correlates with ambulatory blood pressure in healthy normotensive humans. Hypertension. 2007;50(1): 41-6.

40. van de Vooren H, Gademan MG, Swenne CA, TenVoorde BJ, Schalij MJ, Van der Wall EE. Baroreflex sensitivity, blood pressure buffering, and resonance: what are the links? Computer simulation of healthy subjects and heart failure patients. J Appl Physiol (1985). 2007;102(4):1348-56.

41. Karemaker JM. Autonomic integration: the physiological basis of cardiovascular variability. J Physiol. 1999;517 ( Pt 2)(Pt 2): 316.

42. Cohen MA, Taylor JA. Short-term cardiovascular oscillations in man: measuring and modelling the physiologies. J Physiol. 2002;542(Pt 3):669-83.

43. Choi J, Cha W, Park MG. Declining trends of heart rate variability according to aging in healthy asian adults. Front Aging Neurosci. 2020;12:610626.

44. Dantas EM, Kemp AH, Andreão RV, da Silva VJD, Brunoni AR, Hoshi RA, et al. Reference values for short-term resting- state heart rate variability in healthy adults: Results from the Brazilian Longitudinal Study of Adult Health-ELSA-Brasil study. Psychophysiology. 2018;55(6):e13052.

45. BeatScope Easy User’s Guide. Amsterdam, The Netherlands: Finapres Medical Systems B.V.; 2012.

46. Ziemssen T, Gasch J, Ruediger H. Influence of ECG sampling frequency on spectral analysis of RR intervals and baroreflex sensitivity using the EUROBAVAR data set. J Clin Monit Comput. 2008;22(2):159-68.

Cost-Effectiveness Analysis of Long-acting Injectable Once-monthly of Aripiprazole Compared with Long-acting Injectable Once- monthly Paliperidone Palmitate for the Treatment of Stable Schizophrenia Patients in Thailand

Piyanut Ueapanjasin, M. Pharm*,**, Wiwat Thavornwattanayong, B.Sc.(Pharm), M.A., FCCP.*,***, Jadesada Lertsirimunkong, Pharm.D.*,****, Kamolpat Chaiyakittisopon, Ph.D.*, *****

*Health Economic and Research Outcomes (HERO) team, Thailand, **Faculty of Pharmacy, Silpakorn University, Nakhon Pathom, Thailand,

***Department of Pharmaceutical Care, Faculty of Pharmacy, Silpakorn University, Nakhon Pathom, Thailand, ****Department of Pharmacy Administration, College of Pharmacy, Rangsit University, Pathum Thani, Thailand, *****Department of Health Consumer Protection and Pharmacy Administration, Faculty of Pharmacy, Silpakorn University, Nakhon Pathom, Thailand.

ABSTRACT

Objective: Long-acting injectable (LAI)-aripiprazole and LAI-paliperidone palmitate are both second-generation antipsychotics that have been introduced to increase drug compliance in patients. These attributes are expected to enhance drug compliance, particularly in stable patients. The previous studies demonstrated that the efficacy of LAI-aripiprazole and LAI-paliperidone palmitate is controversial. Nevertheless, the costs of treatments and adverse events of both LAI-aripiprazole and LAI-paliperidone palmitate are unlikeness. As there had been no previous cost-effectiveness studies comparing the use of LAI-aripiprazole and LAI-paliperidone palmitate in Thailand, this study was carried out to investigate the matter.

Materials and Methods: This study analysed the cost-effectiveness of LAI-aripiprazole compared with LAI-paliperidone palmitate in the treatment of stable schizophrenia, by using the Markov model from a societal perspective.

Results: The total cost of treatment with LAI-aripiprazole and LAI-paliperidone palmitate was 1,334,919.05 baht and 1,329,818.79 baht, respectively, while the quality-adjusted life years (QALYs) were both 16.35 years. Life-year of the treatment with LAI-aripiprazole and LAI-paliperidone was 24.27 years and 24.25 years, respectively. The cost- effectiveness ratios (CER) of the treatment with LAI-aripiprazole and LAI-paliperidone palmitate were 81,652.85 baht/QALY gained and 81,330.94 baht/QALY gained, respectively.

Conclusion: In Thailand, the treatment of stable schizophrenia with LAI-aripiprazole was shown to provide similar benefits to LAI-paliperidone palmitate in terms of QALYs, despite being more costly. Comparatively, LAI-aripiprazole exhibited better clinical efficacy and led to a longer average life expectancy than LAI-paliperidone. Treatment with LAI-aripiprazole may be dominant strategy, especially with a 2% reduction in drug cost. The results could contribute to appropriate decision-making by policymakers.

Keywords: Aripiprazole; cost-effectiveness; long-acting injectable antipsychotics; paliperidone; stable schizophrenia patients (Siriraj Med J 2023; 75: 725-735)

Corresponding author: Kamolpat Chaiyakittisopon E-mail: chaiyakittiso_k@su.ac.th

Received 18 August 2023 Revised 3 September 2023 Accepted 5 September 2023 ORCID ID:http://orcid.org/0000-0003-3603-7027 https://doi.org/10.33192/smj.v75i10.264770

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

INTRODUCTION

Schizophrenia is a chronic mental illness affecting cognition, emotion, and behaviour. Over 24 million people around the world are impacted by the illness. With about 5.5 million sufferers, China has the highest rate of schizophrenia in the entire globe.1 Schizophrenia incidence rates range from 8 to 43 per 100,000 individuals.2,3 It is a considerable economic and social burden worldwide since the afflicted patients are hardly able to work. Nowadays, schizophrenia is one of the 25 leading causes of disability worldwide4 and one of the top ten causes of Disability-Adjusted Life Years (DALYs).4,5 In Thailand, approximately 8.8 per 1000 of the population suffer from schizophrenia, and was the eighth or ninth leading cause of years lived with disability.6 It also consumed a large amount of annual cost of treatment (i.e., estimated to be THB 87,000 (USD 2,600) per person or THB 31,000 million (USD 925 million) for the entire schizophrenic population in Thailand.7

Treatment of schizophrenia includes both pharmacological and non-pharmacological therapies.8 The goal of treatment is to improve the patient’s quality of life by reducing the symptoms and preventing relapse of patients. For example, conventional and atypical antipsychotics are commonly used for the treatment of schizophrenia9, but conventional antipsychotics produces extrapyramidal symptoms (EPS) while atypical antipsychotics have been reported as metabolic adverse events, e.g., weight gain, hyperlipidemia, and hyperglycemia. Both adverse events have affected patients’ compliance.9,10 Therefore, for some patients, even if they control their symptoms and are discharged from the hospital, 30-40% of discharged patients relapse within 1-2 years.11 Patient compliance is one of the important factors associated with relapse.12-15 Non-adherent patients were approximately three times more likely to be hospitalized in a given year, according to a study conducted in the United States.16

Recently, long-acting injection (LAI) antipsychotics have been suggested to be used17, particularly in the prevention of relapse for non-compliant patients, according to the most evidence-based guidelines for the maintenance treatment of schizophrenia.18 LAIs are recommended as a first-line treatment.18 Thai clinical guidelines19 have been updated to include LAI first-generation antipsychotics (e.g., Fluphenazine decanoate and Haloperidol decanoate) as well as LAI second-generation antipsychotics (e.g., LAI-aripiprazole and LAI-paliperidone palmitate) to improve medication compliance in stable schizophrenia patients. Currently, LAI second-generation have been used more than first-generation antipsychotics because they have lower EPS side effects.19

LAI-aripiprazole and LAI-paliperidone are both second-generation antipsychotics that have been introduced to increase drug compliance20 since they are LAIs and administered once a month. These attributes are expected to enhance drug compliance, particularly for stable patients who are not inpatients. However, they have different pharmacological mechanisms.21 Aripiprazole is a partial agonist at dopamine D2 and serotonin 5-HT1A receptors and an antagonist at 5HT2A receptors while Paliperidone is an antagonist at D2 and 5HT2A receptors. The different mechanisms may contribute to different effectiveness and tolerability.22 The previous comparative studies23,24 demonstrated that the efficacy of LAI-aripiprazole and LAI-paliperidone is controversial. Nevertheless, the costs of treatments and adverse events of both LAI-aripiprazole and LAI-paliperidone are unlikeness.22,23 Utilizing the economic evaluation as a tool for selecting the optimal LAI strategy would be reasonable.

Previously, no cost-effectiveness studies have been carried out that compared the use of LAI-aripiprazole and LAI-paliperidone palmitate in Thailand from the societal perspective. A cost-effectiveness study is crucial in order to evaluate and compare these antipsychotics. Therefore, the cost-effectiveness of LAI-aripiprazole and LAI-paliperidone palmitate was investigated in this study.

MATERIALS AND METHODS

Study design

This study was health economic evaluation using a Markov model to compare the cost-effectiveness of LAI-aripiprazole with LAI- Paliperidone palmitate for the treatment of stable schizophrenia. The analysis was carried out using cost-effectiveness ratios (CER) and presented humanistic outcomes in Quality-Adjusted Life Years (QALYs). The perspective of this study was societal. Future costs and utilities were discounted at 3 percent per year.24 This study has been reviewed and approved by the Human Research Ethics Committee of Silpakorn University (COE 65.1007-165)

Treatments

This study compared monthly dosages of LAI- aripiprazole 400 mg and LAI-paliperidone palmitate 156 mg (equivalent dose of Paliperidone 100 mg).25-28 Patients with schizophrenia who did not respond to LAI- aripiprazole or LAI-paliperidone palmitate treatment were switched to 300 mg per day of clozapine29, which is the only antipsychotic medicine approved by the FDA for treatment-resistant schizophrenia.30

Decision model

The decision model was developed based on previously studies of antipsychotics used in the treatment of schizophrenia patients.31 A Markov model was used to perform decision analysis through Microsoft Excel 2020. The model and assumptions were validated for the disease sequence to ensure its appropriateness for the treatment of stable schizophrenia in Thailand by three psychiatrists.

The model comprises three main health states including remission under the first antipsychotic, relapse, and death, see Fig 1. A ‘death state’ is a state where a patient dies for any reason. ‘Relapse state’ denotes patients who have suffered an exacerbation of their condition and hospital admission, due to non-compliance or the inefficacy of LAI-aripiprazole or LAI-paliperidone. Patients who are not in the death or relapse state are in the ‘remission state’ Transition probabilities and health state utilities were reviewed and derived from published literature.

The model simulated stable schizophrenia patients over the period of their lifetime. Patients were assigned 400 mg LAI-aripiprazole every 4 weeks or 156 mg LAI- paliperidone palmitate (equivalent dose of Paliperidone 100 mg) every 4 weeks. All stable patients under first LAI antipsychotic entered the model in the ‘remission state’ at the beginning of the simulation. It was assumed that the health transition state cycle was 4 weeks. At the end of each cycle, mortality rate, the probability of relapse status, and any adverse events occurrence (i.e., akathisia, dystonia, parkinsonism, dyskinesia, diabetes mellitus, hyperprolactinemia, and weight gain) were assessed for each group of LAI antipsychotics until all the patients died. Patients with ‘remission state’ were not changed health state if they did not experience any

adverse events. Patients who did not respond to LAI due to inefficacy of the treatment were switched to 300 mg per day of clozapine as a second antipsychotic. Patients who experienced relapse due to one of two conditions: (1) Non-compliance with, or inefficacy of LAI-aripiprazole or LAI-paliperidone palmitate, or (2) Relapse from remission with taking clozapine were moved to the relapse state and were switched to 300 mg per day of clozapine as a second antipsychotic. Assumingly, patients who switched to clozapine would continue receiving it until the end of the study. Patients in all states could be moved to the death state throughout the study period according to the probability of death.

Assumptions of the model

1. Patients did not withdraw from the treatment during the study and remained until the end of study.

2. Patients who received LAI-aripiprazole were initially administered with a single injection of 400 mg aripiprazole followed by 20 mg of oral aripiprazole for the first 14 days (i.e., concurrent with the first dose of LAI). This was followed by monthly injections of 400 mg aripiprazole thereafter.

3. Patients who received LAI-paliperidone palmitate were initially administered with injection of 400 mg (equivalent dose of Paliperidone 250 mg) in the first month. This was followed by monthly injections of 156 mg paliperidone palmitate (equivalent dose of Paliperidone 100 mg) thereafter.

4. Patients administered with LAI-aripiprazole or LAI-paliperidone could potentially experience common adverse drug reactions, i.e., akathisia, dystonia, parkinsonism, dyskinesia, diabetes mellitus, hyperprolactinemia, and weight gain.

   
   

   
   

   Fig 1. Markov model structure of Schizophrenia disorder.

   
   

5. Hyperprolactinemia was defined as a serum prolactin level greater than or equal to 25 ng/mL and patients who experienced hyperprolactinemia were switched to clozapine 300 mg per day (or 100 mg three times per day).

6. Weight gain was defined as body weight increased to greater than or equal to 7 percent of basal body weight, and patients who experienced weight gain were provided with counselling.

7. Diabetes mellitus was defined as plasma glucose levels greater than or equal to 200 mg/dl or fasting blood glucose levels greater than or equal to 126 mg/dl which could be reversed when treated with 1,000 mg of metformin daily to control diabetes.

8. Patients who had akathisia were treated with 40 mg per day of propranolol.

9. Patients who had dyskinesia were switched to clozapine 300 mg per day (or 100 mg three times per day).

10. Patients who had dystonia were assumed to be suffering from only mild dystonia and were treated with 6 mg of trihexyphenidyl per day (or 2 mg three times per day).

11. Patients who had parkinsonism were treated with 6 mg of trihexyphenidyl per day (or 2 mg three times per day).

12. Patients who switched to clozapine and experienced a form of agranulocytosis, which is defined as ANC <500 cells per microlith, received 300 micrograms of filgrastim daily, for 7 days.

13. Patients who did not respond to LAI due to inefficacy were switched to 300 mg per day of clozapine as a second antipsychotic.

14. Patients in a relapsed state of health relapsed due to one of two conditions: (1) Non-compliance with, or inefficacy of LAI-aripiprazole or LAI-paliperidone palmitate, or (2) Relapse from remission while taking clozapine.

15. Patients had not received other antipsychotics or other co-interventions.

16. Patients in all health states who received LAI- aripiprazole or LAI-paliperidone palmitate could be moved to the death state based on the normal mortality rate of the Thai population.32 In the case of patients who received clozapine, the mortality rate was based on the mortality rate of schizophrenia patients.33

Time horizon

A Markov model was developed to imitate the treatment of adult schizophrenia patients over a lifetime period from the age of 18 until death with a life expectancy

of not more than 75.7 years.34 A cycle length of 4 weeks was considered appropriate to capture both the clinical treatment and associated events such as relapses and adverse drug reactions from a survey of treatment in Thailand.7

Probability of clinical outcomes

A systematic search was conducted in Medline, SCOPUS and Cochrane databases. The keywords were “schizophrenia, Paliperidone, and Aripiprazole”, with filtering by randomized controlled trial, meta-analysis, systematic reviews, full text and English published literature. Two reviewers independently reviewed abstracts, and articles sequentially to select studies for data abstraction based on the study eligibility criteria. All searched literature was evaluated and given a JADAD score. All probabilities were converted into risks over 4 weeks because of the cycle length and are shown in Table 1.

Studies were identified as eligible for inclusion if they were published as full papers and in the English language. All transition probabilities were obtained from studies involving schizophrenia patients who have used LAI-aripiprazole or LAI-paliperidone. The utility of health states was obtained from studies involving Thai schizophrenia patients who used LAI-aripiprazole or LAI- paliperidone with/or without any adverse events. Where search results were inconclusive, the study proceeded as follows: (i) involving schizophrenia patients controlled by antipsychotics drug and whether they had side effects, or (ii) other patients who had utility of health state and side effects, or (iii) utility was retrieved from international studies due to the limited data in Thailand. Articles were excluded from the review if they met any of the following criteria: (i) non-full text papers, (ii) editorials and opinions, letters, research protocols, conference abstracts, duplicate reports of the same study, and notes and books.

Costs

All costs are expressed in Thai baht and are shown in Table 1. Drug treatment costs were derived from the Drug and Medical Supply Information Center (DMSIC) and the Ministry of Public Health, Thailand.35

All drug costs (i.e., LAI-aripiprazole, LAI- paliperidone palmitate, Clozapine, metformin, propranolol, trihexyphenidyl, and filgrastim) were obtained from the drug’s median price in Thailand. Costs of meals, nursing care costs, and laboratory costs including tests for FBS, haemoglobin A1c, serum prolactin, and complete blood counts were obtained from the mean cost per unit of secondary care by standard cost lists for health

TABLE 1. All parameters used in the Markov model.

43

28, 46

47

47

47

TABLE 1. All parameters used in the Markov model. (Continue)

Parameters Distribution Mean ± SE References

Laboratory costs

48

48

Treatments and Additional Procedures

Hospitalization (per admission) OPD service (per visit) Psychoeducation (per year)

Gamma Gamma

Gamma

25,610.53 ± 2,561.05

592.53 ± 59.25

3,143.11 ± 314.31

7

7

7

48

Direct non-medical costs

Travel (per visit) Gamma 256.11 ± 25.61 7

Utility

Remission state Relapse state Hyperprolactinemia Akathisia Dyskinesia Dystonia Parkinsonism Weight gain Diabetes Agranulocytosis

Beta Beta Beta Beta Beta Beta Beta Beta Beta

Beta

0.690 ± 0.026

0.578 ± 0.028

0.618 ± 0.027

0.639 ± 0.056

0.608 ± 0.040

0.449 ± 0.054

0.626 ± 0.011

0.664 ± 0.027

0.664 ± 0.027

0.460 ± 0.059

40, 41

40, 41

38, 41

39-41

39-41

40-42

39-41

40, 41

40, 41

49

Meal (per visit) Gamma 61.66 ± 6.17 48

technology assessment in Thailand.35 Costs of Out- Patient Department (OPD) services, hospitalization, psychoeducation, travel expenses and family time were obtained from previous studies carried out in Thailand.7 The frequencies of outpatient visits and admissions per year from surveys of mental illness in Thailand were 7.7 and 0.5, respectively.36

All costs were adjusted to 2021 values using the consumer price index from the Bureau of Trade and Economic indices, The Ministry of Commerce, Thailand.37

Utility values

Quality-adjusted life-years (QALYs) were used for outcomes measurement. The humanistic outcomes were measured in utility weights for different health states and side effects, ranging from 0 (death) to 1 (perfect health). Utility weights were multiplied by life expectancies to generate QALYs.

Utility values of remission states were estimated

based on the disability weights according to previous studies.31 Utility values of other health states were obtained from previous studies.38-42 All utility values are shown in Table 1.

One-way sensitivity and probabilistic sensitivity analysis

Random Monte Carlo Simulation was applied for probabilistic sensitivity analyses using Microsoft Excel 2020. All variables were randomized and run 1,000 times to generate the probability distribution and the ICER estimation. Beta distribution was used for transition probabilities and utility value, and gamma distribution was used for costs. The results are shown as a cost-effectiveness plane between incremental costs and incremental QALYs. One-way sensitivity analysis was performed using Microsoft Excel 2020. The parameter values were changed individually and regularly to the lowest and highest values. The results of one-way sensitivity analyses were presented in tornado diagram.

RESULTS

Cost-effectiveness analysis

The cost-effectiveness analysis results (presented in Table 2) showed that the total cost of treatment with LAI-aripiprazole and LAI-paliperidone palmitate was LAI-aripiprazole 1,334,919.05 baht and 1,329,818.79 baht, respectively, while the QALYs were 16.35 years for both. Life years of the treatment with LAI-aripiprazole and LAI-paliperidone was 24.27 years and 24.25 years, respectively.

The CER of the treatment with LAI-aripiprazole and LAI-paliperidone palmitate was 81,652.85 baht/QALY gained and 81,330.94 baht/QALY gained, respectively. Due to the significantly lower CER, treatment with LAI-paliperidone palmitate is more cost-effective than treatment with LAI-aripiprazole.

One-way sensitivity and probabilistic sensitivity analysis Fig 2 presented the one-way sensitivity analysis result in a tornado diagram. According to the findings, the probability of LAI-aripiprazole-induced dyskinesia was the variable with the greatest influence on the ICER. The probabilistic sensitivity analysis result is presented in Fig 3 as a cost-effectiveness plane between the incremental cost and the incremental QALYs of treatment with LAI- aripiprazole compared with LAI-paliperidone palmitate. The Monte Carlo simulations randomized each variable 1,000 times. The red point represents the base-case ICER.

DISCUSSION

A previous economic evaluation study of aripiprazole in Thailand31 suggested that oral aripiprazole was the dominant strategy, showing greater QALYs and lower

TABLE 2. Cost-effectiveness results.

Fig 2. The tornado diagram depicts the results of a one-way sensitivity analysis.

Fig 3. The cost-effectiveness plane of treatment with LAI-aripiprazole compared with LAI paliperidone palmitate.

cost than risperidone in acute schizophrenia patients. However, no previous studies have investigated the economic evaluation of LAI-aripiprazole’s use in treating patients with stable schizophrenia. This analysis represented the first economic evaluation comparing LAI-aripiprazole with LAI-paliperidone palmitate for the treatment of stable schizophrenia patients in Thailand. The study revealed that both LAI-aripiprazole and LAI-paliperidone palmitate were similarly beneficial in terms of QALYs, despite the fact that the drug cost of LAI-aripiprazole was cheaper than LAI-paliperidone palmitate. This notion may be considered LAI-aripiprazole as a cost-effective strategy for the treatment of stable schizophrenia patients in Thailand.

Regarding the possible adverse events associated with LAI-aripiprazole or LAI-paliperidone palmitate treatment, such as hyperprolactinemia, akathisia, dyskinesia, dystonia, parkinsonism, diabetes, and weight gain, previous studies conducted in the United States25,50, the United Kingdom51, Finland52 and France45, only considered and included some of these events in their economic models. In contrast, this study took into consideration all potential adverse events that could impact both costs and utilities, providing a more realistic model that aligns with clinical practices. However, it’s worth noting that adverse drug reactions were assumed to be mild, as they were closely monitored every 4 weeks in accordance

with the cycle length in the model. Nevertheless, it is important to recognize that despite such monitoring, some adverse events can still occur and may be severe, especially in the short term. The probability of patients experiencing relapse symptoms with once-monthly LAI-aripiprazole treatment was lower than that of those receiving once-monthly LAI-paliperidone palmitate. This could lead to savings in long-term treatment costs. However, from an economic perspective, efficacy is not the only consideration; overall costs, including the treatment of adverse drug events, must also be taken into account. The use of LAI-aripiprazole may result in higher total costs due to the management of side effects, leading to a higher CER. Considering that both LAIs have similar QALYs, LAI-aripiprazole’s superior clinical efficacy and longer life expectancy compared to LAI-paliperidone. It’s possible that LAI-aripiprazole justifies a 2% price decrease and might be offered LAI- aripiprazole as a dominating strategy for compensating the costs associated with treating side effects, particularly dyskinesia management.

This study encountered some limitations regarding data availability. Specifically, there were few randomized controlled trials (RCTs) that directly compared the efficacy and adverse drug reactions of LAI-aripiprazole with LAI- paliperidone palmitate, and no previous studies were conducted in Thailand. Consequently, the probabilities

of transitioning between health states and experiencing adverse drug reactions were derived from international resources. To enhance the probabilities’ validity and minimize the effects of confounding factors, sensitivity analysis was performed based on global data. Additionally, utility values were obtained and recalculated into utility weights specifically for Thai schizophrenia patients. The use of different values from various data sources and diverse populations resulted in a notable variation in the likelihood of side effects and efficacy for each patient receiving LAI treatment. This variability holds the potential to significantly impact the overall treatment outcome, as manifested by the conspicuous amplitude of sensitivity dispersion observed among the sensitivity results. Moreover, this study did not include sexual dysfunction and neuroleptic malignant syndrome (NMS) as adverse drug reactions of antipsychotics due to their rarity and the limited availability of relevant data.

Further economic evaluation studies are required to evaluated the cost-effectiveness of LAI-aripiprazole and LAI-paliperidone palmitate using the real-word data in Thai stable schizophrenia patients to provide a more accurate and reliable evaluation.

CONCLUSION

In Thailand, the treatment of stable schizophrenia with LAI-aripiprazole was found to yield similarly beneficial results in terms of QALYs when compared to treatment with LAI-paliperidone palmitate, despite being more costly. Comparatively, LAI-aripiprazole exhibited better clinical efficacy and led to a longer average life expectancy than LAI-paliperidone. If the drug cost of LAI-aripiprazole were decreased by 2%, treatment with LAI-aripiprazole would become a dominant cost-effectiveness strategy. The results of this study could contribute to informed decision-making by policymakers.

ACKNOWLEDGMENTS

This study received the specific grant from Thai Otsuka Pharmaceutical Co., Ltd. However, the study sponsor had no influence on the literature review, developing model, data analysis and the decision of publication.

We would like to thank Mr. Paul Mines, Faculty of Pharmacy, Silpakorn University, for proofreading the article.

REFERENCES

1. Institute for Health Metrics and Evaluation (IHME). GBD Compare Data Visualization. Seattle, WA: IHME, University of Washington, 2020. Available from http://vizhub.healthdata. org/gbd-compare.

2. McGrath J, Saha S, Chant D, Welham J. Schizophrenia: a

   concise overview of incidence, prevalence, and mortality. Epidemiol Rev. 2008;30(1):67-76.

3. McGrath J, Saha S, Welham J, El Saadi O, MacCauley C, Chant D. A systematic review of the incidence of schizophrenia: the distribution of rates and the influence of sex, urbanicity, migrant status and methodology. BMC Med. 2004;2:13.

4. Chong HY, Teoh SL, Wu DB-C, Kotirum S, Chiou C-F, Chaiyakunapruk N. Global economic burden of schizophrenia: a systematic review. Neuropsychiatr Dis Treat. 2016;12:357-73.

5. James SL, Abate D, Abate KH, Abay SM, Abbafati C, Abbasi N, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet. 2018;392(10159): 1789-858.

6. Phanthunane P, Vos T, Whiteford H, Bertram M, Udomratn P. Schizophrenia in Thailand: prevalence and burden of disease. Popul Health Metr. 2010;8(1):24.

7. Phanthunane P, Whiteford H, Vos T, Bertram M. Economic burden on schizophrenia: Empirical analyses from a survey in Thailand. J Ment Health Policy Econ. 2012;15(1):25-32.

8. Keepers GA, Fochtmann LJ, Anzia JM, Benjamin S, Lyness JM, Mojtabai R, et.al. The American Psychiatric Association Practice Guideline for the Treatment of Patients With Schizophrenia. Am J Psychiatry. 2020 Sep 1;177(9):868-72.

9. Crossley NA, Constante M, McGuire P, Power P. Efficacy of atypical v. typical antipsychotics in the treatment of early psychosis: meta-analysis. Br J Psychiatry. 2010;196(6):434-9.

10. Geddes J. Prevention of relapse in schizophrenia. N Engl J Med. 2002;346:56-58.

11. Olfson M, Mechanic D, Hansell S, Boyer CA, Walkup J, Weiden PJ. Predicting medication noncompliance after hospital discharge among patients with schizophrenia. Psychiatr Serv. 2000;51(2): 216-22.

12. Alvarez-Jimenez M, Priede A, Hetrick S, Bendall S, Killackey E, Parker A, et al. Risk factors for relapse following treatment for first episode psychosis: a systematic review and meta-analysis of longitudinal studies. Schizophr Res. 2012;139(1-3):116-28.

13. Coldham E, Addington J, Addington D. Medication adherence of individuals with a first episode of psychosis. Acta Psychiatr Scand. 2002;106(4):286-90.

14. Pothimas N, Tungpunkom P, Chanprasit C, Kitsumban V. A Cross-sectional Study of Factors Predicting Relapse in People with Schizophrenia. Pacific Rim Int J Nurs Res. 2020;24(4): 448-59.

15. Pitanupong J, Ratanaapiromyakij P, Teetharatkul T. Factors Associated with Readmission Among Individuals with One Previous Episode of Schizophrenia in Southern Thailand: A University Hospital-based Retrospective Study. J Hlth Sci Res. 2022;40(6):657-70.

16. Gilmer TP, Dolder CR, Lacro JP, Folsom DP, Lindamer L, Garcia P, et al. Adherence to treatment with antipsychotic medication and health care costs among Medicaid beneficiaries with schizophrenia. Am J Psychiatry. 2004;161(4):692-9.

17. Pitanupong J, Karakate A, Tepsuan L, Sritrangnant G. Attitudes Toward Long-Acting Injectable Antipsychotics among Schizophrenia Patients in Southern Thailand: A Multihospital- Based Cross-Sectional Survey. Siriraj Med J. 2022;74(3):193- 201.

18. Kittipeerachon M, Sirivech P, Suraaroonsamrit B, Mekwilai W.

    
    

    Clinical Guideline for Schizophrenia Patients in Hospitals (Medical Edition). 2nd ed. Nonthaburi: Department of Mental Health, Ministry of Public Health.; 2017.

19. Mantana Kittipeerachon, Pattama Sirivech, Burin Suraaroonsamrit, Wee Mekwilai. Clinical Guideline for Schizophrenia Patients in Hospitals (Medical Edition). 2nd ed. Nonthaburi: Department of Mental Health, Ministry of Public Health.; 2017.

20. Kishimoto T, Nitta M, Borenstein M, Kane JM, Correll CU. Long-acting injectable versus oral antipsychotics in schizophrenia: a systematic review and meta-analysis of mirror-image studies. J Clin Psychiatry. 2013;74(10):957-65.

21. Li H, Rui Q, Ning X, Xu H, Gu N. A comparative study of paliperidone palmitate and risperidone long-acting injectable therapy in schizophrenia. Prog Neuro-psychopharmacol Biol Psychiatry. 2011;35(4):1002-8.

22. Naber D, Hansen K, Forray C, Baker RA, Sapin C, Beillat M, et al. Qualify: a randomized head-to-head study of aripiprazole once- monthly and paliperidone palmitate in the treatment of schizophrenia. Schizophr Res. 2015;168(1-2):498-504.

23. Ceraso A, Lin JJ, Schneider-Thoma J, Siafis S, Tardy M, Komossa K, et al. Maintenance treatment with antipsychotic drugs for schizophrenia. Cochrane Database Syst Rev. 2020; 8(8):CD008016.

24. Subcommittee of Development of National Drug List Thailand. Handbook of Health Technology Assessment for Thailand. 2nd ed. Nonthaburi: Watcharin P.P. Printing; 2013.

25. Di Lorenzo R, Ferri P, Cameli M, Rovesti S, Piemonte C. Effectiveness of 1-year treatment with long-acting formulation of aripiprazole, haloperidol, or paliperidone in patients with schizophrenia: retrospective study in a real-world clinical setting. Neuropsychiatr Dis Treat. 2019;15:183-98.

26. Majer IM, Gaughran F, Sapin C, Beillat M, Treur M. Efficacy, tolerability, and safety of aripiprazole once-monthly versus other long-acting injectable antipsychotic therapies in the maintenance treatment of schizophrenia: a mixed treatment comparison of double-blind randomized clinical trials. J Mark Access Health Policy. 2015;3.

27. Taipale H, Mehtälä J, Tanskanen A, Tiihonen J. Comparative effectiveness of antipsychotic drugs for rehospitalization in schizophrenia—a nationwide study with 20-year follow-up. Schizophr Bull. 2018;44(6):1381-7.

28. Citrome L, Kamat SA, Sapin C, Baker RA, Eramo A, Ortendahl J, et al. Cost-effectiveness of aripiprazole once-monthly compared with paliperidone palmitate once-monthly injectable for the treatment of schizophrenia in the United States. J Med Econ. 2014;17(8):567-76.

29. National Institute for Health and Care Excellence. Psychosis and schizophrenia in adults: prevention and management Clinical guideline. London, United Kingdom. 2014.

30. Columbia University Medical Center. Better options for people with treatment-resistant schizophrenia: Research on the comparative effectiveness of antipsychotic medications [cited 2022 Aug 30]. Available from: https://www.sciencedaily.com/ releases/ 2015/11/151106062315.htm.

31. Thavornwattanayong W, Lertsirimunkong J, Thongkerd N, Pitakthanin N, Wettayanon P, Pongjakpanit H. Cost-effectiveness analysis of aripiprazole compared with risperidone in the treatment of acute schizophrenia patients in Thailand. TJPS. 2018;42(3): 169-75.

32. Health Information System Development Office. Trend of

    health in Thailand 2017 [cited 2022 Aug 30]. Available from: https://www.hiso.or.th/health/data/html/search1.php?menu=1&i.

33. Bitter I, Czobor P, Borsi A, Fehér L, Nagy B, Bacskai M, et al. Mortality and the relationship of somatic comorbidities to mortality in schizophrenia. A nationwide matched-cohort study. Eur Psychiatry. 2017;45:97-103.

34. World Health Organization. World health statistics 2016: monitoring health for the SDGs sustainable development goals: World Health Organization; 2016.

35. Drug and Medical Supply Information Center. In: Ministry of Public Health Thailand. Reference of Drugs Costs [cited 2022 Aug 30]. Available from: http://dmsic.moph.go.th/index/ drugsearch/1.

36. Kongsakon R, Leelahanaj T, Price N, Birinyi-Strachan L, Davey

    P. Cost analysis of the treatment of schizophrenia in Thailand: a simulation model comparing olanzapine, risperidone, quetiapine, ziprasidone and haloperidol. J Med Assoc Thai. 2005;88(9):1267.

37. Trade Policy and Strategy Office. Economic and Trade Indices Database: ETID [cited 2022 Aug 30]. Available from: http:// www.price.moc.go.th

38. Davies A, Vardeva K, Loze J-Y, L’Italien GJ, Sennfalt K, Baardewijk M. Cost-effectiveness of atypical antipsychotics for the management of schizophrenia in the UK. Curr Med Res Opin. 2008;24(11): 3275-85.

39. Lenert LA, Sturley AP, Rapaport MH, Chavez S, Mohr PE, Rupnow M. Public preferences for health states with schizophrenia and a mapping function to estimate utilities from positive and negative symptom scale scores. Schizophr Res. 2004;71(1):155- 65.

40. Lin L, Zhao YJ, Zhou HJ, Khoo AL, Teng M, Soh LB, et al. Comparative cost-effectiveness of 11 oral antipsychotics for relapse prevention in schizophrenia within Singapore using effectiveness estimates from a network meta-analysis. Int Clin Psychopharmacol. 2016;31(2):84-92.

41. Phanthunane P, Vos T, Whiteford H, Bertram M. Health outcomes of schizophrenia in Thailand: Health care provider and patient perspectives. Asian J Psychiatr. 2010;3(4):200-5.

42. Skogseid I, Malt U, Røislien J, Kerty E. Determinants and status of quality of life after long‐term botulinum toxin therapy for cervical dystonia. Eur J Neurol. 2007;14(10):1129-37.

43. Druais S, Doutriaux A, Cognet M, Godet A, Lançon C, Levy P, et al. Cost effectiveness of paliperidone long-acting injectable versus other antipsychotics for the maintenance treatment of schizophrenia in France. Pharmacoeconomics. 2016;34(4):363-91.

44. Seppälä A, Pylvänäinen J, Lehtiniemi H, Hirvonen N, Corripio I, Koponen H, et al. Predictors of response to pharmacological treatments in treatment-resistant schizophrenia–A systematic review and meta-analysis. Schizophr Res. 2021;236:123-34.

45. Fleischhacker WW, Gopal S, Lane R, Gassmann-Mayer C, Lim P, Hough D, et al. A randomized trial of paliperidone palmitate and risperidone long-acting injectable in schizophrenia. Int J Neuropsychopharmacol. 2012;15(1):107-18.

46. Masuda T, Misawa F, Takase M, Kane JM, Correll CU. Association with hospitalization and all-cause discontinuation among patients with schizophrenia on clozapine vs other oral second- generation antipsychotics: a systematic review and meta-analysis of cohort studies. JAMA Psychiatry. 2019;76(10):1052-62.

47. Treur M, Baca E, Bobes J, Canas F, Salvador L, Gonzalez B, et al. The cost-effectiveness of paliperidone extended release in Spain. J Med Econ. 2012;15(Suppl 1):26-34.

    
    

48. Health Intervention and Technology Assessment: HITAP Ministry of Public Health. Standard Cost Lists for Health Technology Assessment [cited 2022 Aug 30]. Available from: https://costingmenu.hitap.net/

49. Weeks JC, Tierney MR, Weinstein MC. Cost effectiveness of prophylactic intravenous immune globulin in chronic lymphocytic leukemia. NEJM. 1991;325(2):81-6.

50. Sapin C, Hartry A, Kamat SA, Beillat M, Baker RA, Eramo A. Pharmacoeconomic comparison of aripiprazole once-monthly and paliperidone palmitate from a head-to-head clinical trial

    in schizophrenia: a US analysis. Drugs Context. 2016;5:212301.

51. Hodgson RE. Evaluating the cost and clinical effectiveness of long-acting, injectable aripiprazole and paliperidone palmitate once a month in a real-world setting. ClinicoEconomics and Outcomes Research: Clinicoecon Outcomes Res. 2019;11:517- 24.

52. Einarson TR, Pudas H, Goswami P, van Impe K, Bereza BG. Pharmacoeconomics of long-acting atypical antipsychotics for acutely relapsed chronic schizophrenia in Finland. J Med Econ. 2016;19(2):121-30.

Burnout and Associated Factors among Thai Anesthesiology Residents

Neranchala Soonthornkes, M.D.*, Orattha Sakornwattananon, M.D.*, Jitpanu Jirapitakkul, M.D.*, Winitra Nuallaong, M.D.**

*Department of Anesthesiology, **Department of Psychology, Faculty of Medicine, Thammasat University, Pathum Thani 10120, Thailand.

ABSTRACT

Objective: The study evaluates the prevalence of burnout and the associated risk factors among Thai anesthesiology residents.

Materials and Methods: This cross-sectional study was conducted by an online questionnaire and sent 385 anesthesiology residents in Thailand. The questionnaire consisted of a consent form, personal information, potential risk factors of burnout, and the Thai version of the Copenhagen Burnout Inventory-Student Survey (CBI-SS).

Results: A total of 248 respondents completed the questionnaire. The overall prevalence of burnout among Thai anesthesiology residents was 35.5%. Sleeping for 5-6 hours (odds ratio 3.68, 95% CI 1.40 to 9.68, p = 0.008), sleeping less than 5 hours (odds ratio 6.89, 95% CI 1.90 to 24.92, p = 0.003), along with dissatisfaction and the idea of discontinuation of training, were associated with higher burnout (odds ratio 8.38, 95% CI 3.65 to 19.25, p < 0.001 and odds ratio 3.11, 95% CI 1.57 to 6.18, p = 0.001, respectively). Compared to first-year residents, second-year residents exhibited a lower risk of burnout (odds ratio 0.32, 95% CI 0.134 to 0.76, p = 0.009) while no significant difference among third-year residents (odds ratio 0.54, p = 0.17, 95% CI 0.25-1.17).

Conclusion: The prevalence of burnout among Thai anesthesiology residents was high. Sleeping less than 7 hours, experiencing dissatisfaction, and contemplating discontinuation of training were associated with a higher risk of burnout, while being a second-year resident lowered the risk.

Keywords: Burnout; anesthesiology; resident; Thailand (Siriraj Med J 2023; 75: 736-743)

INTRODUCTION

Burnout is a work-related syndrome characterized by emotional exhaustion, depersonalization, and a sense of reduced personal accomplishment.1 It responds to chronic job-related stressors and is exclusively associated with the work environment. Burnout has been associated with strained relationships among team members2, leading to adverse effects on health and overall quality of life. Physicians have reported increasing levels of stress and burnout3, which can have detrimental effects on the quality of patient care and potentially contribute to medical errors.4,5

Anesthesiologists exhibit a relatively high prevalence of

burnout. The prevalence of burnout among anesthesiology varies across studies, ranging from 13.8% for burnout syndrome to as high as a 59% risk of burnout.6,7 Of particular concern is the high risk of burnout during residency training. A survey conducted among anesthesiology residents in the United States found that 41% were at high risk of burnout.8 The precise reasons of anesthesia residents’ burnout remain unclear. However, it is plausible that they lack experience and need to acquire knowledge, cognitive abilities, and technical skills, as well as adapt to efficient perioperative teamwork, which contributes to this heightened risk.

Corresponding author: Neranchala Soonthornkes E-mail: neransoon@gmail.com

Received 9 August 2023 Revised 5 September 2023 Accepted 12 September 2023 ORCID ID:http://orcid.org/0000-0002-7884-695X https://doi.org/10.33192/smj.v75i10.264582

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

Objective: No study has reported on the prevalence of burnout among Thai anesthesia residents. The study objective is to evaluate the prevalence, and possible factors associated with burnout among Thai anesthesia residents.

MATERIALS AND METHODS

Cross-sectional study

Following the approval from the Ethics Committee of Thammasat University (COA No.023/2566), the data collection period took place from January 22, 2023, to February 12, 2023.

Participants

Inclusion criteria were Thai anesthesiology residents who were in the Thai anesthesia training program during the study period. Anesthesia residents who declined to answer the questionnaire were excluded.

Questionnaire

The Questionnaires consisted of four sections: a consent form, personal data, potential risk factors of burnout, and the Thai version of the Copenhagen Burnout Inventory Student Survey (Thai CBI-SS). The questionnaire was developed using the online Google Forms application. A research assistant reached out to the representatives of 14 institutions throughout Thailand that host anesthesia resident training programs to share the online questionnaire with all anesthesia residents. Participants could accept or decline answering it. All responses were collected anonymously. To ensure maximum participation, the research assistant sent reminders once a week for three consecutive weeks.

Participants were asked to furnish details about their demographic characteristics (including gender, age, marital status, years of resident training, underlying disease, income, smoking habits, alcohol consumption, and exercise habits). The potential risk factors of burnout were reviewed in relevant literature.7, 8-10 These factors included working hours, shift hours, adequacy of consultation, sleep duration, workplace resources, job support at work and at home, satisfaction with resident training, an idea to discontinue training, as well as demographic characteristics data. The final section was the Thai version of the Copenhagen Burnout Inventory Student Survey (Thai CBI-SS). The Thai CBI-SS consisted of a comprehensive set of twenty-five questions categorized into four domains: personal burnout (six items), study-related burnout (seven items), colleague-related burnout (six items), and teacher-related burnout (six items). Each item was rated on a five-point Likert scale, ranging from ‘100 (always)’ to ‘0 (never/almost never)’. Scores between 50 and 74

were categorized as ‘moderate’ burnout, scores between 75 and 99 were classified as ‘high’ burnout and a score of 100 indicated severe burnout.

The Copenhagen Burnout Inventory (CBI), developed by Kristensen TS11, has demonstrated good reliability and validity as a measure of burnout. Winwood and Winefield12 confirmed that the CBI effectively conceptualizes burnout, and exhibits strong reliability and validity. Moreover, the Copenhagen Burnout Inventory-Student Survey (CBI-SS)13 has been acknowledged as a reliable and valid instrument for evaluating burnout among medical students. Correlational analyses between CBI-SS and the Maslach Burnout Inventory-Student Survey (MBI-SS), a widely-used standardized tool, have revealed moderate to strong correlations. The CBI-SS has undergone linguistic translation and cultural adaptation into several languages, including Thai. Research has validated the Thai version of the Copenhagen Burnout Inventory-Student Survey (CBI-SS) and demonstrated its reliability as an effective tool for evaluating burnout syndrome among preclinical medical students in Thailand.14

Although anesthesia residents are part of a postgraduate training program, in the Thai cultural educational system, Thai residents interact with the staff in a teacher/student relationship. The decision was made to utilize the CBI- SS instead of the CBI because the CBI-ss has a teacher- related burnout question domain.

Statistical analysis

The sample size was calculated based on a previous report of the prevalence of burnout (41%), with an acceptable margin of error (0.05). The calculated sample size was 372. The decision was made to recruit all Thai anesthesia residents at the time of the study, which totaled 385 considering the survey dropout rate. The data were analyzed using StataCorp. Version 17. College Station, TX: StataCorp LLC; 2021. Descriptive statistics summarized the characteristics. Categorical variables were depicted in the form of frequencies and percentages. Continuous variables were described using either means with standard deviations or medians with interquartile ranges. For comparing continuous variables, t-tests or Wilcoxon rank-sum tests were employed. Categorical variables were compared using either chi-square tests or Fisher exact tests. Logistic regression analysis identified predictors of burnout. Univariate associations were initially examined, and factors with a p-value < 0.1 were included in the multivariable logistic regression model to identify independent risk factors. Results were reported as odds ratios with 95% confidence intervals. A p-value

< 0.05 indicated statistical significance.

RESULTS

Out of the 385 Thai anesthesia residents from 14 institutions who were invited to participate in the online questionnaire, a total of 250 individuals responded, resulting in a response rate of 64.94%. Among the respondents, 248 participants completed all the questions, while 2 residents declined to answer the questionnaire.

Physician characteristic

The majority of respondents, 200 were female (80.65%). The median age was 29 years old, with a range of 27 to 30 years. Among the respondents, there were 77 first-year residents (31.05%), 77 second-year residents (31.05%), and 94 third-year residents (37.90%) who completed the questionnaire.

Regarding underlying disease, 15 individuals (6.05%), reported underlying diseases including allergic rhinitis, polycystic ovary syndrome, gastritis, obstructive sleep apnea, migraine, and Glucose-6-Phosphate Dehydrogenase deficiency. Five residents (2.02%) had a major depressive disorder. Further details of participant characteristics are shown in Table 1.

Prevalence

The prevalence of burnout among Thai anesthesia residents who responded to the questionnaire was 35.48% (95% CI 0.30 to 0.42), with 33.06% experiencing moderate burnout and 2.42% experiencing high burnout. No participants reported severe burnout. These findings indicate that burnout is a significant concern among Thai anesthesia residents.

Factors associated with burnout

The univariate analysis was conducted to screen factors associated with burnout and found ten factors associated with burnout. The results are presented in Table 2. Following the screening procedure, a multivariable logistic regression analysis was conducted to validate the findings, identifying four significant factors which are sleep duration, year of resident training, dissatisfaction, and having an idea of discontinuation of resident training associated with burnout. The result is shown in Table 3. Sleeping less than 7-8 hours was identified as a significant risk factor for burnout, particularly among residents who slept less than 5 hours (odds ratio 6.89, 95% CI 1.90 to 24.92, P = 0.003). Additionally, residents who slept for 5-6 hours also had a higher risk of burnout

(odds ratio 3.68, 95% CI 1.40 to 9.68, P = 0.008).

Dissatisfaction with resident training and having an idea of stopping resident training were found to be significant risk factors for burnout (odds ratio 8.38, 95%

CI 3.65 to 19.25, P = 0.00, and odds ratio 3.11, 95% CI

1.57 to 6.18, P = 0.001, respectively).

Compared to first-year residents, second-year residents had a significantly lower risk of burnout (odds ratio 0.32, 95% CI 0.134 to 0.76, P = 0.009). However, no significant difference was observed between third-year residents and first-year residents (odds ratio 0.54, P = 0.17, 95% CI 0.25 to 1.17).

DISCUSSION

The prevalence of burnout among Thai anesthesiology residents is high at 35.48%, although lower than reported in the United States by Oliveira et al. (41%)8 and Huaping Sun (51%).9 Various factors may contribute to this difference, including individual variations, cultural influences, and differences in training programs between countries. Additionally, the study’s sample size may have influenced the prevalence rate, as only 66.67% (248 of 372) of the target sample size was recruited. However, this online response rate of 64.94% is higher than the average reported in recent meta-analyses of online surveys which was around 44.1%.15

The multivariable logistic regression analyses identified factors including sleeping time, year of training, satisfaction with the training program, and the idea of discontinuation of training that were significantly associated with burnout among Thai anesthesiology residents. The results align with previous research10,16 that has found an association between insufficient sleep and increased risk of burnout. In this study, sleeping less than 7 hours increased the burnout risk. Insufficient sleep can have various physical, cognitive, and emotional effects that increase stress levels and make individuals more susceptible to burnout. While this study focused on sleep duration, Hannah K. Allen17 has found an association between the poor quality of sleep which is characterized by difficulties falling asleep or frequent awakenings, and burnout. Additionally, Chatlaong T found an association between sleep quality and emotional exhaustion, which is a component of burnout among resident trainees.18 However, our questionnaire evaluated only the duration, not the quality, of sleep. It would be valuable for future studies to consider evaluating both sleep duration and quality to provide more understanding of the relationship between sleep and burnout.

According to the research conducted by Huaping Sun and De Oliveira GS8,9, it was discovered that an increase in the age of physician trainees corresponded to a reduced risk of experiencing burnout. The second- year residents had a significantly lower risk of burnout compared to first-year residents, while there was no

TABLE 1. Descriptive statistics of respondent characteristics. TABLE 1. Descriptive statistics of respondent characteristics.

(Continue)

Characteristic N 248

Sex

Male 48 (19.35)

Year of resident training 3rd year

2nd year

1st year

94 (37.90)

77 (31.05)

77 (31.05)

Female 200 (80.65)

Marital status Single Married

In relationship

139 (56.05)

17 (6.85)

92 (37.10)

Age 29 (27-30)

Have children 5 (2.02)

Have underlying disease 15 (6.05)

Smoking 2 (0.81)

Have Psychological disease 5 (2.02)

Exercise Routine Sometimes

No

16 (6.45)

125 (50.4)

107 (43.15)

Alcohol drinking 128 (51.61)

Official working hours per month

>160 119 (47.98)

128-160 115 (46.37)

Shifts hours per month

>160

128-160

88-120

40-80

<40

13 (5.24)

54 (21.77)

73 (29.44)

97 (39.11)

11 (4.44)

80-120 14 (5.65)

Income

Enough to save 8 (3.23)

Enough to spend 75 (30.24)

Sleeping time (hours) 7-8

5-6

<5

31 (12.50)

169 (68.15)

48 (19.35)

Not enough 165 (66.53)

Dissatisfaction with resident training 62 (25)

Data are presented as number (%) or median (range). *One piece of data is missing from the answers of 248 participants.

significant difference among third-year residents. A possible explanation for this finding is that second-year residents have had more time to adapt to their work, studies, and interactions with colleagues and the work environment. They may have gained more experience and developed better coping strategies, which could contribute to a lower risk of burnout. On the other hand, first-year residents are faced with the challenges of acquiring new knowledge and navigating stressful situations which could increase their risk of burnout. Interestingly, our study did not find a significant difference in burnout risk between first-year and third-year residents. This could be attributed to the fact that third-year residents carry higher responsibilities, face greater expectations, and need to maintain high levels of concentration as they prepare for the board-certified examination. These factors may offset the additional experience they have gained.

Having thought about discontinuing training* 114 (45.97)

Residents who expressed dissatisfaction with the training program and considered discontinuing their training had a significantly higher risk of burnout. The association between burnout and job satisfaction has been observed in previous studies, such as Govardhan LM’s study, which found an inverse correlation between burnout

TABLE 2. Univariate analysis for factors associated with burnout.

TABLE 2. Univariate analysis for factors associated with burnout. (Continue)

CI = confidence interval, OR = odds ratio, significant at P ≤0.0, risk of burnout are presented as number (%)

and job satisfaction among obstetrics and gynecology residents.19 Factors contributing to job satisfaction are influenced by various details. Unfortunately, our questionnaire did not explore the specific reasons behind residents’ dissatisfaction or desire to discontinue training. Further research is needed to investigate these causes for improvement in training programs.

Regarding other factors related to burnout, some previous studies have reported factors to be associated with an increased risk of burnout, such as younger age20- 22, female gender8,20, marital status21, having children22, high alcohol consumption8, lack of job support7,21, and high workload.7-9,22 Conversely, having sufficient work resources has been found to decrease the risk of burnout9.

TABLE 3. Multivariable logistic regression analysis for factors associated with burnout.

power to detect significant associations. Thirdly, the unclear wording of questions regarding working hours may adversely impact the result. Lastly, the study did not explore certain potential risk factors in depth, such as the quality of sleep or the specific reasons for dissatisfaction with the training program. Future research could be conducted to explore the factors contributing to burnout. This research would help in developing suitable support

Conflict of interest statement

The authors declare no conflict of interest

Funding

This study was supported by a grant from the Faculty of Medicine Thammasat Hospital, Thammasat University, Pathum Thani, Thailand.

REFERENCES

1. Maslach C, Jackson S E, Leiter M. Maslach Burnout Inventory. In book: Evaluating Stress: A Book of Resources Publisher: The Scarecrow Press, Editors. C. P. Zalaquett, R. J. Wood 1, January 1997. p.191-218

2. Devi S. Doctors in distress. Lancet. 2011;377(9764):454-5.

3. Shanafelt TD, Boone S, Tan L, Dyrbye LN, Sotile W, Satele D, West CP, Sloan J, Oreskovich MR. Burnout and satisfaction with work-life balance among US physicians relative to the general US population. Arch Intern Med. 2012;172(18):1377- 85.

4. Shanafelt TD, Balch CM, Bechamps G, Russell T, Dyrbye L, Satele D, Collicott P, Novotny PJ, Sloan J, Freischlag J. Burnout and medical errors among American surgeons. Ann Surg. 2010; 251(6):995-1000.

5. Firth-Cozens J, Greenhalgh J. Doctors’ perceptions of the links between stress and lowered clinical care. Soc Sci Med. 1997; 44(7):1017-22.

6. Sanfilippo F, Noto A, Foresta G, Santonocito C, Palumbo GJ, Arcadipane A, Maybauer DM, Maybauer MO. Incidence and Factors Associated with Burnout in Anesthesiology: A Systematic Review. Biomed Res Int. 2017;2017:8648925.

7. Afonso AM, Cadwell JB, Staffa SJ, Zurakowski D, Vinson AE. Burnout Rate and Risk Factors among Anesthesiologists in the United States. Anesthesiology. 2021;134(5):683-96.

8. de Oliveira GS Jr, Chang R, Fitzgerald PC, Almeida MD, Castro-Alves LS, Ahmad S, McCarthy RJ. The prevalence of burnout and depression and their association with adherence to safety and practice standards: a survey of United States anesthesiology trainees. Anesth Analg. 2013;117(1):182-93.

9. Sun H, Warner DO, Macario A, Zhou Y, Culley DJ, Keegan MT. Repeated Cross-sectional Surveys of Burnout, Distress, and Depression among Anesthesiology Residents and First- year Graduates. Anesthesiology. 2019;131(3):668-77.

10. Chin W, Guo YL, Hung YJ, Yang CY, Shiao JS. Short sleep duration is dose-dependently related to job strain and burnout in nurses: a cross sectional survey. Int J Nurs Stud. 2015;52(1): 297-306.

11. Kristensen TS, Borritz M, Villadsen E, Christensen KB. The Copenhagen Burnout Inventory: A new tool for the assessment of burnout. Work & Stress. 2005;19(3):192-207.

12. Winwood PC, Winefield AH. Comparing Two Measures of Burnout Among Dentists in Australia. Int J Stress Manag. 2004;11(3): 282-9.

13. Campos JADB, Carlotto MS, Marôco J. Copenhagen Burnout Inventory - student version: adaptation and transcultural validation for Portugal and Brazil. Psicol Reflex Crit. 2013;26(1): 87-97.

14. Dangprapai Y, Saisavoey N, Sa-Nguanpanich N. Reliability and validity study of the Thai adaptation of the Copenhagen Burnout Inventory-Student Survey (CBI-SS) among preclinical medical students at the Faculty of Medicine Siriraj Hospital, Mahidol University, Thailand. PLoS One. 2021;16(12):e0261887.

15. Wu MJ, Zhao K, Fils-Aime F. Response Rates of Online Surveys in Published research: a meta-analysis. Computers in Human Behavior Reports [Internet]. 2022 [cited 2023 Apr 02] May;7:100206. Available from: https://www.sciencedirect. com/science/article/pii/S2451958822000409

16. Söderström M, Jeding K, Ekstedt M, Perski A, Akerstedt T. Insufficient sleep predicts clinical burnout. J Occup Health Psychol. 2012;17(2):175-83.

17. Allen HK, Barrall AL, Vincent KB, Arria AM. Stress and Burnout Among Graduate Students: Moderation by Sleep Duration and Quality. Int J Behav Med. 2021;28(1):21-28.

18. Chatlaong T, Pitanupong J, Wiwattanaworaset P. Sleep Quality and Burnout Syndrome among Residents in Training at the Faculty of Medicine, Prince of Songkla University. Siriraj Med J [Internet]. 2020 Apr. 5 [cited 2023 Sep. 3];72(4):307-14. Available from: https://he02.tci-thaijo.org/index.php/sirirajmedj/ article/view/240944

19. Govardhan LM, Pinelli V, Schnatz PF. Burnout, depression and job satisfaction in obstetrics and gynecology residents. Conn Med. 2012;76(7):389-95..

20. Chiron B, Michinov E, Olivier-Chiron E, Laffon M, Rusch E. Job satisfaction, life satisfaction and burnout in French anaesthetists. J Health Psychol. 2010;15(6):948-58

21. Shams T, El-Masry R. Job Stress and Burnout among Academic Career Anaesthesiologists at an Egyptian University Hospital. Sultan Qaboos Univ Med J. 2013;13(2):287-95.

22. Shanafelt TD, Balch CM, Bechamps GJ, Russell T, Dyrbye L, Satele D, et al. Burnout and career satisfaction among American surgeons. Ann Surg. 2009;250(3):463-71.

Associated Factors of Quality of Life in Adult Female Acne Coexisting with Hyperandrogenism and Polycystic Ovarian Syndrome using the Dermatology Life Quality Index

Pattriya Jirawattanadon, M.D.*, Leena Chularojanamontri, M.D.*, Yanisorn Nanchaipruek, M.D.*, Panicha Chantrapanichkul, M.D.**, Papapit Tuchinda, M.D.*, Kanokvalai Kulthanan, M.D.*

*Department of Dermatology, **Department of Obstetrics and Gynecology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok 10700, Thailand.

ABSTRACT

Objective: This study aimed to assess the effects of AFA and its associated factors, especially those coexisting with hyperandrogenism or PCOS, on QoL using the Dermatology Life Quality Index (DLQI).

Materials and Methods: A retrospective cross-sectional study was conducted on AFA patients who had been treated between May 2018 and January 2020. Dermatologists had performed history taking and determined the clinical severity of the acne, and gynecologists had identified PCOS. Aggravating factors and DLQI were self-reported by the patients.

Results: A total of 208 AFA patients, aged 31.8±7.1 years on average, were included. The mean DLQI score was 8.0±5.4 (range: 0-23). Patients with moderate to severe acne had significantly higher total DLQI scores compared to those with almost clear to mild acne (P=0.043). Similarly, patients who reported acne aggravated by diet, androgenetic alopecia, or perioral and chin lesions had significantly higher total DLQI scores (P=0.025, P=0.049, and P=0.014, respectively). However, PCOS and hirsutism did not significantly impact QoL. The aspect most affected was feeling embarrassed and self-conscious, with significantly greater impacts observed in patients with moderate to severe acne compared to mild acne. The daily activities of patients with androgenetic alopecia were significantly impacted in terms of their choice of clothes and sexual difficulties.

Conclusion: AFA mostly had a moderate to extremely large effect on patients’ QoL. Knowing the factors influencing QoL, such as acne severity, dietary aggravation of acne, and androgenetic alopecia, may enable physicians to improve the QoL of patients.

Keywords: Acne; adult female acne; DLQI; quality of life; hyperandrogenism; androgenetic alopecia (Siriraj Med J 2023; 75: 744-751)

INTRODUCTION

Acnevulgarisisoneofthemostprevalentdermatological diseases. It is commonly found in adolescents but can sometimes persist through adulthood.1 It is generally accepted that acne negatively impacts patients’ lives, including embarrassment and limitations to daily and

social activities.2 The impact of acne is evident in the form of decreases in the quality of life (QoL) that are as severe as those found with other debilitating diseases, such as asthma, epilepsy, diabetes mellitus, back pain, and arthritis.3 “Adult female acne” (AFA) is defined as acne in women over the age of 25.4-6 Previous reports

Corresponding author: Leena Chularojanamontri E-mail: leenajim@gmail.com

Received 24 July 2023 Revised 29 August 2023 Accepted 12 September 2023 ORCID ID:http://orcid.org/0000-0001-6625-6445 https://doi.org/10.33192/smj.v75i10.264327

All material is licensed under terms of the Creative Commons Attribution 4.0 International (CC-BY-NC-ND 4.0) license unless otherwise stated.

have mentioned that AFA affects mental and emotional health, social and personal behavior, and QoL.7,8

Acnevulgarisisalsooneofthesignsofhyperandrogenism and polycystic ovary syndrome (PCOS), which also negatively affect patients’ QoL.9 QoL is a crucial aspect of the morbidity related to acne, and it is an essential consideration when determining treatments. The number of studies specifically exploring QoL in AFA is limited, especially in AFA with PCOS, and few factors associated with QoL have been identified.10,11 Several dermatological questionnaires are available to measure the impact of skin disorders on QoL. One of these is the Dermatology Life Quality Index (DLQI), a 10-item self-administered questionnaire to measure the effects of skin diseases on patients’ QoL. It has been translated and validated in many languages, including Thai.12,13

This study aimed to assess the effects on patients’ QoL of AFA (especially AFA coexisting with hyperandrogenism or PCOS) and associated factors. Assessments were made using the Thai version of the DLQI.

MATERIALS AND METHODS

This retrospective cross-sectional study was approved by the Siriraj Institutional Review Board of the Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand (COA no. Si 1043/2021). We enrolled female patients aged 25 or older who presented with AFA disease at the Siriraj Dermatology Clinic between May 2018 and January 2020 and had DLQI scores in their medical records. Patients were excluded if they had been previously treated with any systemic therapy for acne or if no DLQI information was available.

Demographic characteristics and acne-related information were gathered. The data comprised patient age, onset subtype, precipitating cause of acne flare, type and sites of lesions, extent of acne lesions, blood pressure, body weight, height, and waist circumference. Acne presenting in women aged 25 or older was defined as AFA,4,5 whereas acne in women older than 45 was referred to as “perimenopausal acne.”14 The onset of AFA is classified into 3 subtypes: persistent, late-onset, and relapsed.5 Acne developing in adolescence and continuing into adulthood is deemed “persistent acne,” whereas acne that first occurs after age 25 is called “late-onset acne.” Acne that occurs during adolescence for a few years and then reoccurs after 25 is termed “relapsed acne.”

All patients were evaluated for acne severity by dermatologists using the Investigator’s Global Assessment. The severities were classified as follows:15

• “Almost clear:” rare comedones or 1 inflammatory lesion

• “Mild:” some comedones with few inflammatory lesions without nodular lesions

• “Moderate:” up to many comedones with some inflammatory papules but no more than 1 small nodular lesion

• “Severe:” many comedones and inflammatory papules, and few nodular lesions

• “Very severe:” many comedones, inflammatory papules, and nodular lesions

The skin signs of hyperandrogenism, including androgenetic alopecia and hirsutism, were recorded. “Androgenetic alopecia” (AGA) was defined as grade I or more under Ludwig’s classification.16 In Asian populations, a Modified Ferriman–Gallwey score of ≥ 6 indicates hirsutism,16 and a body mass index (BMI) ≥ 23 kg/m2 denotes an overweight or obese status.17

Gynecologists evaluated PCOS with the consent of the AFA patients. Diagnoses of this condition were based on recommendations from international evidence- based guidelines for assessing and managing PCOS.16 QoL impairment was recorded using the validated Thai version of the DLQI.13 The questionnaire has 10 items that evaluate 6 dermatological aspects of life: symptoms and feelings, daily activities, leisure activities, work or school, personal relationships, and treatment. Each item has a score ranging from 1 to 3. The total DLQI score is calculated by summing the scores for the 10 items, giving a possible maximum of 30 and a minimum of 0. Higher scores signify greater degrees of impairment of QoL. The total DLQI scores are interpreted as follows: 0–1, no effect on QoL; 2–5, small effect; 6–10, moderate effect; 11–20, very large effect; and 21–30, extremely large effect.12,13

The statistical analysis was undertaken using IBM SPSS Statistics for Macintosh, version 27.0 (IBM Corp., Armonk, NY, USA). Data are presented as numbers and percentages, means and standard deviations, and ranges. Mean differences between continuous variables were calculated by an independent t-test or 1-way ANOVA. Differences in categorical data were calculated by the chi- squared test or Fisher’s exact test. Statistical significance was set at P < 0.05, and the results are reported with a 95% confidence interval.

RESULTS

Two hundred and eight AFA patients were included. Their mean age was 31.8 ± 7.1 years, ranging from 25.2 to 59.4 years. Perimenopausal acne was identified in 14 patients (6.7%). Most patients presented with persistent acne (47.1%); late-onset and relapsed acne accounted for 26.9% and 26.0% of cases, respectively. Half of the

patients (51%) had moderate severity of acne, followed by mild acne (38%), severe acne (8.1%), and almost clear acne (2.9%). Twenty-four patients (11.5%) had hirsutism, and 12 (5.8%) had AGA. Fifty-four patients consented to undergo evaluations by gynecologists, of whom 26 (48.1%) were diagnosed with PCOS.

All patients completed all 10 questions of the DLQI questionnaire. The overall mean DLQI score was 8.0 (standard deviation, 5.4; range, 0–23). No, small, moderate, very large, and extremely large effects on QoL were found in 8.2%, 32.7%, 29.3%, 27.4%, and 2.4% of patients, respectively. The mean total DLQI scores for each characteristic are detailed in Tables 1 & 2. Perimenopausal acne had lower mean total DLQI scores than non-perimenopausal acne, with no significant differences noted. All acne subtypes had a moderate effect on QoL, whereas late-onset acne had the lowest mean total DLQI scores. Patients with moderate to severe acne had significantly higher mean total DLQI scores than patients with almost clear to mild acne (P = 0.043).

Regarding promoting factors, patients who reported that diet was related to acne had a significantly higher mean total DLQI score than those who did not (P = 0.025). Stress, sleep deprivation, pre-menstruation, sunlight, dust, and exercise did not significantly affect the mean total DLQI scores. Patients with cosmetic products promoting acne had significantly lower mean total DLQI scores (P = 0.018).

As for the risk for metabolic syndrome, patients with moderate to severe acne had a significantly higher mean BMI (22.6 ± 4.4) than patients with almost clear to mild acne (21.4 ± 3.1) (P = 0.031). However, the mean DLQI score of patients at risk of metabolic syndrome did not differ from that of patients with AFA who did not have the risk. In addition, AFA with AGA had a significantly greater mean total DLQI score (11.0 ± 7.4) than AFA without AGA (P = 0.049). However, significant differences were not found between AFA with and without hirsutism, or between AFA with and without PCOS.

A comparison of the DLQI scores for the various acne characteristics is presented in Table 2. Regarding the sites of acne lesions, patients with acne in the perioral area and chin had significantly greater mean total DLQI scores than those without such lesions (P = 0.014). More extensive acne resulted in higher total DLQI scores, especially when comparing the mean total DLQI score for 1 site versus the involvement of ≥ 2 sites (P = 0.05). Mixed comedones and inflammatory papules had the highest mean total DLQI scores.

Concerning DLQI, the dermatological aspect of life

most frequently affecting patients was feeling embarrassed and self-conscious. This factor had the highest mean ± SD score (1.7 ± 1.0) and was followed by symptoms, leisure activities, and daily activities (Table 3). Regarding the severity of acne, patients with moderate to severe acne reported greater impacts on feeling embarrassed and self-conscious, daily activities, and leisure activities than those with mild acne. As mentioned earlier, patients with AGA experienced a marked impact on their QoL compared with patients without AGA. In particular, the daily activities of patients with AGA were significantly impacted in terms of their choice of clothes and sexual difficulties. As for age, most aspects of DLQI did not vary by age; the exception was perimenopausal acne, which had relatively less effect on the choice of clothes.

DISCUSSION

This study highlights that AFA can have moderate to substantial effects on QoL. Patients with AFA in the perimenopausal group tended to have a better QoL than those in the non-perimenopausal group. This difference may result from younger patients paying more attention to their appearance, thereby negatively affecting their social and personal lives. However, Lasek et al. reported that acne had greater effects on the QoL of older patients,18 which may be due to differences in self-perception related to ethnicity. A previous study showed that even mild acne could enormously affect patients’ QoL.19, 20 In the current investigation, 17.6% of the patients with almost clear to mild acne reported very large effects on their QoL. Although previous studies reported no correlation between acne clinical severity and QoL scores,3,21 our work revealed a correlation between the impact of acne on QoL and the severity of acne. The clinical severity of acne among AFA patients affected their QoL more than the subtype of acne.

According to the effects of aggravating factors and the QoL, the patients with diet-aggravated acne stated that they had a worse QoL than those who did not report that diet exacerbated the condition. In addition, in our study on adult acne patients, those who reported the use of cosmetic products that exacerbated their acne also exhibited a significantly lower impact on their quality of life, possibly suggesting a potential link between the use of cosmetics and the role of cosmetic camouflage, although further investigation is required to elucidate this relationship.

Feeling embarrassed was the aspect that mainly affected QoL in cases of AFA, which is consistent with earlier studies on acne patients in Thailand.19,20 An extensive study on different racial groups also reported that AFA

TABLE 1. Demographic, clinical characteristics, and the Dermatology Life Quality Index of adult female patients with acne (N = 208).

*A P value <0.05 indicates statistical significance.

Abbreviations BMI, body mass index; DBP, diastolic blood pressure; DLQI, Dermatology Life Quality Index; PCOS, polycystic ovary syndrome; SBP, systolic blood pressure; SD, standard deviation

TABLE 2. The Dermatology Life Quality Index scores for different characteristics of acne.

*A P value <0.05 indicates statistical significance.

Abbreviations: DLQI, Dermatology Life Quality Index; SD, standard deviation

had the greatest impact on self-esteem and perceptions across all population groups.22 It has also been reported that AGA can affect psychological and social experiences as well as QoL, especially in female patients,23,24 which is similar to our findings. Our investigation determined that AFA with AGA had substantial and statistically significant adverse effects on QoL, especially regarding the choice of clothes and sexual difficulties.

This study has a few limitations. First, the number of patients was insufficient to draw some significant associated factors. In addition, the aggravating factors were subjective, given that patients reported them. Last, evaluating QoL with only 1 tool means that not all aspects of the QoL of the patients may have been captured.

In conclusion, our study demonstrated that AFA had moderate to very large effects on patients’ QoL. There was a correlation between the clinical severity of acne and QoL scores. The greatest impacts on QoL were associated with AGA, acne on the perioral area and chin, and patients’ perception that their diet aggravated AFA. PCOS and hirsutism did not affect patients’ QoL. These data highlight the impact of acne and the associated factors in AFA patients. This underscores the multifaceted nature of adult acne and highlights the need for a holistic approach in understanding and managing this dermatological condition and it presents a compelling avenue for future research and potential interventions to enhance the quality of life for individuals affected by acne.

Original Article SMJ

TABLE 3. Percentage of affected patients and mean score for each aspect of the Dermatology Life Quality Index, compared by the severity of acne, the presence of androgenetic alopecia, and the status of perimenopausal acne.