Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
114
adakorn Tantisarasart, M.D.*, Sunisa Chartmongkolchart, M.D.*, Rassamee Chotipanvithayakul, M.D.**
*Department of Anesthesiology, **Department of Epidemiology, Faculty of Medicine, Prince of Songkla University, Hatyai Songkhla 90110 ailand.
Nomogram as a Predictor for Postoperative Acute
Kidney Injury in Super-Elderly Patients Undergoing
Noncardiac Surgery
ABSTRACT
Objective: ere has been increase the incidence of postoperative acute kidney injury (AKI), especially among
super-elderly patients. e study aimed to identify risk factors and develop a nomogram among super-elderly
patients undergoing noncardiac surgery.
Material and Methods: A single-center retrospective cohort study of patients aged greater than or equal to 80
years that underwent non-cardiac surgery between January 2018 to December 2020. Acute kidney injury (AKI) was
identied by Kidney Disease Improving Global Outcome (KDIGO) during seven days aer surgery. Multivariate
logistic regression was constructed from preoperative and intraoperative data with variables with P-value<0.2
included in the nal model. e performance of model function was conducted by area under the receiver curve
(AUC) and calibration curves.
Results: Eight hundred and twenty patients were included; 124 (15%) developed postoperative AKI. A multivariate
logistic regression model consisting of COPD, ASA classication, part of surgery, propofol and Succinylated gelatin
was displayed as the nomogram. e model showed good discrimination with an AUC 0.746. e cuto point of
63, which had the highest Youden index, was chosen with sensitivity and specicity of 83% and 45%, respectively.
e nomogram showed good performance by the Hosmer-Lameshow goodness-of-t test (X
2
= 6.0697 and P value
= 0.6394).
Conclusion: e nomogram predicted model for predicting postoperative AKI among super-elderly patients
showed moderate discrimination ability and was instituted. It can help physicians to detect high-risk patients early
and promptly prevent episodes of AKI.
Keywords: Postoperative acute kidney injury; super-elderly patient; nomogram (Siriraj Med J 2022; 74: 114-125)
Corresponding author: adakorn tantisarasart
E-mail: thadakorn.t@psu.ac.th
Received 1 November 2021 Revised 7 December 2021 Accepted 26 December 2021
ORCID ID: https://orcid.org/0000-0002-2448-3276
http://dx.doi.org/10.33192/Smj.2022.15
INTRODUCTION
Mortality due to surgery and anesthesia has been
reduced from >25% to 19% among elderly aged >80 years.
1
is substantially increased surgery and anesthesia among
them. AKI, a preventable condition, has been one of the
leading cause of post-operative morbidity and mortality
rates, especially among super-elderly patients. e incidence
of acute kidney injury is highest in elderly about 16.98%
2
and increase is age-dependent.
3
erefore, Postoperative
acute kidney injury was associated with 30-day unplanned
readmission, postoperative renal failure, dialysis, risk
of infection and prolonged mechanical ventilation.
4–6
Postoperative AKI is likely related to multifactorial factors
including preoperative factor which is intractable, and
intraoperative and postoperative factors which can be
improved and are preventable.
Tantisarasart et al.
All material is licensed under terms of
the Creative Commons Attribution 4.0
International (CC-BY-NC-ND 4.0)
license unless otherwise stated.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
115
Original Article
SMJ
Previous literature has reported that baseline decrease
in eGFR, renal artery involvement, radiocontrast media,
intraoperative hypotension, operative time, surgical
technique, intraoperative crystalloid use, low preoperative
Hemoglobin (Hb), intraoperative blood loss, peripheral
arterial disease, perioperative transfusion, amount of blood
loss, preoperative hypoalbuminemia, general anesthesia
and preoperative use of angiotensin-converting enzyme
inhibitor(ACEI), angiotensin II receptor blockers (ARB)
and diuretics are associated with acute kidney injury .
e most important factors in early detection of high
risk patients include enhanced invasive monitoring,
maintain uid balance and avoid nephrotoxic agent.
8-10
However, there are many predictive tools such
as preoperative GFR, preoperative proteinuria and
preoperative creatinine, but none of these tools are best
validated for prediction of AKI the incidence, risk factor
of postsurgical AKI. Studies documenting predictive score
use in postoperative acute kidney injury among elderly
people are rare. erefore, this study aims to develop a
predicting tool for postoperative acute kidney injury in
super-elderly patients.
MATERIALS AND METHODS
Ethical approval and reporting guidelines
e study was approved by the institutional review
board (IRB) of Prince of Songkla University Hospital (IRB
number: 63-408-8-1). e inform consent was waived due
to the study being observational study without medical
intervention. e study comply with the transparent
reporting of a prediction model for TRIPOD statement.
13
Study hospital and study designs
is retrospective cohort study was conducted in a
tertiary medical center in ailand from 1 January 2018
to 31 December 2020. A total 820 records of all patients
aged ≥80 years who underwent non-cardiac surgery and
received either general anesthesia or spinal anesthesia
were included.
Exclusion criteria were patients who had preoperative
end-stage renal disease requiring renal replacement
therapy, diagnosed acute kidney injury since the time
of admission to a day prior to the surgery, amputee
12
,
bed-ridden, patients without baseline serum creatinine or
follow-up serum creatinine levels to identify postoperative
AKI and no urine output recorded. Eligible patients’ data
were retrieved from anesthetic records in the hospital
information system (HIS) (Fig 1). An anesthesiologist
reviewed patients’ proles and clinical data including ASA
classication, vital signs, and urine outputs. Laboratory
results, details of surgery and anesthesia, and clinical
and hospital-related outcomes were also recorded.
Outcome and denition of acute kidney injury (AKI)
Postoperative AKI was assessed within 7 days aer
surgery. According to Kidney Disease-Improving Global
Outcomes (KDIGO) 2012 guideline
14
, AKI was dened
as either one of the following 3 conditions a) serum
Fig 1. Study ow chart.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
116
creatinine increased by more than 1.5-1.9 times of the
baseline level, b) increase in serum creatinine level more
than 0.3 mg/dL within 48 hours and c) urine output less
than 0.5 mL/kg/h for 6 hours continuously.
e primary outcome was the 7-days AKI rate,
according to KDIGO. Secondary outcomes were post-
operative 30-day mortality rate, period of hospital stay,
period of ICU stay and requirement of renal replacement
therapy.
Sample size determination
Estimated sample size estimated base on ability to
detect association with exposure and outcome at odds
ratio at least 1.5 with 80% power was 820. e statistical
signicance was set at P value ≤ 0.05.
Statistical analysis
Statistical analysis was performed using R program
version 4.0.2.
15
Categorical variables were presented
as percentage and proportion such as gender, Body
mass index (BMI), ASA classication, type of surgery,
choice of anesthesia, part of surgery, and anesthetic
drug use. e categorical parameters were compared
between AKI group and non-AKI group by Student
t-test or Wilcoxon Rank Sum test. In case of small size
population, Chi-squared test or Fisher’ exact test were
used to compare the two groups. Continuous variables
were presented as mean ± standard deviation (SD) and
median ± interquartile range was used to describe normal
and non-normal distribution, respectively. Categorical
variables were analyzed using Chi-square test or Fisher’
Exact test as appropriate.
Predictors of postoperative acute kidney injury
Predictive factors included preoperative and
perioperative data. Pre-operative factors included
patient’s prole, clinical condition, current medication,
and laboratory variables including Hemoglobin and
creatinine. Preoperative serum creatinine level was
analyzed within 24-48 hours prior to surgery. Baseline
creatinine clearance was computed by the standard
Cockcro Gault formula using age, lean body weight
(kg), serum creatinine (mg/dL) and sex. Perioperative
factors included anesthetic drugs, uid resuscitation,
use of vasopressors, estimated blood loss, duration of
surgery, duration of hypotension and intraoperative urine
output and intraoperative hemodynamic. (Supplement 1)
Model development
Variables included in multivariate analysis were
chosen from univariate analysis results that had p-value
< 0.2 (Tables 1 and 2). Multi-collinearity was tested by
using variance ination factor (VIF) > 5 criteria. Stepwise
logistic regression method was used to get the nal model
(Table 3). e Hosmer-Lameshow goodness-of-t test
also displayed a good performance (X
2
= 6.0697 and
P value = 0.6394). e variable with p-value< 0.05 in
multivariate logistic regression model was considered
statistically signicant.
Score derivation and validation
e prediction score variable was selected from
multivariate analysis and weight adjustment was conducted
by nomogram. For the nal model, each predictor score
was summarized in total postoperative acute kidney
injury score. Youden’s index was used to reveal maximize
specicity and sensitivity cuto values of prediction
score.
RESULTS
A total 3840 patients who underwent non cardiac
surgery were assessed for eligibility from 1 January 2018
to 31 December 2020. Eight hundred and twenty eligible
patients were retrieved in this study. e incidence of
postoperative acute kidney injury among super-elderly
was 15%.
Table 1 shows clinical characteristics at preoperative
period between patients with AKI (acute kidney injury)
and non AKI with no statistically signicant dierences
in demographic data such as age, sex, BMI, comorbidities
and preoperative blood pressure among groups. e
mean age of the super elderly was 87 years old and the
average BMI was 21 kg/m
2
. Generally, half of the AKI
group underwent emergency surgery and had a higher
proportion of ASA classication IV and V than non
AKI group. No dierences in part of surgery, choice of
anesthesia and current medication were found between
the two groups. Creatinine clearance among AKI group
was signicant lower than that in the non-AKI group.
Table 2 illustrates clinical characteristics in
intraoperative period between AKI group and non-
AKI group. According to induction agent, midazolam
use was signicantly higher among AKI group, while,
Propofol use tended to lower incidence of AKI compared
to others. Opioid drugs, muscle relaxant drugs and
inhalation agents displayed no signicant dierences
in the postoperative AKI. However, AKI group more
frequently used vasopressors than non-AKI group. e
incidence of postoperative acute kidney injury was higher
among patients with duration of hypotension more than
15 minutes. e AKI group tended to use Succinylated
gelatin for resuscitation more than non AKI group.
Tantisarasart et al.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
117
Original Article
SMJ
Supplement 1. Univariate analysis of risk factors for acute kidney injury.
Variable OR 95%CI P-value
Age* 1.04 (1,1.09) 0.05
Male* 1.24 (0.84,1.82) 0.28
Comorbidities
HT 1.37 (0.9,2.07) 0.13
DM 1.36 (0.85,2.18) 0.19
COPD 1.84 (0.93,3.73) 0.09
PVD 1.12 (0.13,9.7) 0.09
CHF 1.88 (0.19,18.2) 0.58
Type (ref=elective)
Emergency 1.92 (1.3,2.82) <0.05
Part (ref=Neuro)
Abdomen 1.9 (0.92,3.95) 0.08
Orthopedic 2.13 (1,4.55) 0.05
Urology 3.93 (1.48,10.39) <0.05
Vascular 1.91 (0.89,4.11) 0.09
Others 1.77 (0.6,5.2) 0.30
ASA (ref=II)
III 2.19 (1.23,3.9) <0.05
IV-V 13.13 (6.37,27.08) <0.05
CrCl 0.99 (0.97,1) 0.04
Hb 0.93 (0.84,1.02) 0.10
Current medication
ACEI 0.76 (0.38,1.52) 0.43
ARB 0.96 (0.48,1.94) 0.92
Choice (ref=spinal block)
Generalanesthesia 1.02 (0.54,1.89) 0.96
Induction
Propofol* 0.38 (0.25,0.56) <0.05
Etomidate* 2.24 (1.15,4.39) 0.01
Ketamine* 2.44 (0.62,9.57) 0.02
Midazolam* 1.98 (1.34,2.92) <0.05
Narcotic
Morphine* 1.1 (0.31,3.85) 0.88
Fentanyl* 4.96 (0.67,36.83) 0.11
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
118
Supplement 1. Univariate analysis of risk factors for acute kidney injury. (Continue)
Variable OR 95%CI P-value
Muscle relaxant
Cisatracurium* 1.35 (0.67,2.7) 0.39
Roccuronium* 0.78 (0.39,1.56) 0.47
Succinylcholine* 0.94 (0.48,1.83) 0.84
Inhalation (ref=no) 0.09
Sevourane 0.84 (0.5,1.4)
Desurane 0.55 (0.3,1)
Contrastmedia 0.78 (0.47,1.29) 0.33
Hypotension 1.52 (0.97,2.39) 0.06
Durationofhypotension(>=30mins) 2.32 (1.52,3.52) <0.05
Total urine output (ref>=2)
<=0.5 6.87 (3.8,12.42) <0.05
0.51-2 1.81 (1.05,3.12) 0.03
Diureticuse 1.24 (0.65,2.4) 0.51
Vasopressoruse
Norepinephrine* 1.95 (1.33,2.88) <0.05
Ephedrine* 0.59 (0.39,0.89) 0.01
Epinephrine* 4.57 (1.67,12.5) <0.05
Dopamine* 9.58 (3.64,25.23) <0.05
Fluid
NSS* 1.74 (1.11,2.71) 0.01
Balanced salt solution* 0.51 (0.34,0.76) <0.05
Succinylatedgelatinuse* 2.55 (1.66,3.93) <0.05
Blood transfusion 1.65 (1.12,2.42) 0.01
RBC* 1.58 (1.07,2.34) 0.02
FFP* 2.04 (1.25,3.34) <0.05
PC* 2.09 (1.07,4.05) 0.03
Abbreviations: COPD: chronic obstructive lung disease, CHF: chronic heart failure, Hb: hemoglobin, Crcl: creatinine clearance, ACEI:
Angiotensin-converting enzyme inhibitors, ARB: Angiotensin receptor blockers, NSS: normal saline solution, PRC: packed red blood cell,
FFP: fresh frozen plasma, PC: platelet count *Continuous variables
Tantisarasart et al.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
119
Original Article
SMJ
TABLE 1. Clinical characteristics of preoperative period in super-elderly patients with and without Acute Kidney
Injury (AKI) (N=820).
Characteristics AKIgroup Non-AKIgroup P-value
(N=124) (N=696)
N (%) N (%)
Age (years) mean±SD 87.5 ±4.5 86.7±3.9 0.05
Gender 0.32
Female 53 (41.1) 334 (48)
Male 71(57.3) 362(52)
BMI (kg/m
2
)* 21.9(18.9,24.9) 21.7(19.1,24.2) 0.54
Comorbidities
Diabetes 27(21.8) 118(17.0) 0.24
Hypertension 87(70.2) 440(63.5) 0.16
COPD 11 (8.9) 35 (5) 0.13
Congestive heart failure 1 (0.8) 3 (0.4) 0.48
Type of surgery <0.05
Scheduled 59(47.6) 442(63.5)
Emergency 65(52.4) 254(36.5)
ASA type < 0.05
II 15(12.1) 197(28.3)
III 78(62.9) 468(67.2)
IV 28(22.6) 31(4.5)
V 3(2.4) 0(0.0)
PreoperativeHb(g/dL)* 10(9,11.2) 11(9,12) 0.04
PreoperativeCr(mg/dL)* 1.1(0.8,1.4) 0.9(0.8,1.2) <0.05
PreoperativeCrCl(mL/min)* 34.7(26.8,44.9) 38.8(29.6,49.7) <0.05
Part of surgery 0.14
Abdomen 40(32.3) 223(32.0)
Orthopedic 30(24.2) 149(21.4)
Neurology 10(8.1) 106(15.2)
Urology 10(8.1) 27(3.9)
Vascular 28(22.6) 155(22.3)
Others 6(4.8) 36(5.2)
Current medication
ACEI 10(8.1) 72(10.3) 0.53
ARB 10 (8.1) 58 (8.3) 1
Choice of anesthesia 1
Generalanesthesia 111(89.5) 622(89.4)
Spinal block 13 (10.5) 74 (10.6)
Preoperative blood pressure
SBP(mmHg)* 131(115.8,146.2) 130(119,146.2) 0.72
DBP(mmHg)* 69(60,77) 70(62,79) 0.18
MAP(mmHg)* 88.5(81,98.2) 90(82,99) 0.47
Abbreviations: BMI: body mass index, COPD: chronic obstructive lung disease, CHF: chronic heart failure, Hb: hemoglobin, Cr: creatinine,
Crcl: creatinine clearance, ACEI: Angiotensin-converting enzyme inhibitors, ARB: Angiotensin receptor blockers, SBP: systolic blood
pressure, DBP: diastolic blood pressure, MAP: mean arterial pressure *Continuous data were reported as median and IQR 1-3
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
120
TABLE 2. Clinical characteristics of intraoperative period in super-elderly patients with and without Acute Kidney
Injury (AKI) (N=820).
Characteristics AKI Non-AKI P-value
(N=124) (N=696)
N (%) N (%)
Inductionagent
Propofol 73(58.9) 551(79.2) <0.05
Thiopental 0(0) 12(1.9) 0.23
Etomidate 13 (11.5) 35 (5.5) <0.05
Ketamine 3(2.4) 7(1.0) 0.18
Midazolam 57(46) 209(30) <0.05
Narcotic
Morphine 120(97.6) 658(97.8) 0.74
Fentanyl 123(100) 669(99.4) 1
Muscle relaxant
Succinylcholine 11 (10.3) 67 (10.9) 0.98
Cisatracurium 97 (90.7) 539 (87.8) 0.49
Rocuronium 10(9.3) 72(11.7) 0.58
Inhalation
Sevourane 72(58.1) 371(53.3) 0.37
Desurane 29(23.4) 226(32.5) 0.05
No 23(18.5) 99(14.2) 0.26
Vasopressor
Ephedrine 38(30.6) 298(42.8) 0.01
Epinephrine 7 (5.6) 9 (1.3) <0.05
Norepinephrine 71(57.3) 283(40.7) <0.05
Durationofsurgery(minutes)mean±SD 177.5(135,256.3) 180(135,256.2) 0.88
Intraoperativehypotension 96(77.4) 482(69.3) 0.08
Durationofhypotension(minutes)* 15(5,82.5) 10(0,20) <0.05
Contrastmediause 21(16.9) 144(20.7) 0.39
Diureticuse 12(9.7) 55(7.9) 0.63
Fluidintake(mL)* 1150(500,2000) 1100(700,1675) 0.63
Balanced salt solution (mL)* 1000 (500, 1450) 1000 (700, 1500) 0.56
NSS(mL)* 500(100,1212.5) 500(0,1000) 0.07
Succinylatedgelatin* 500(450,675) 500(300,500) 0.02
Blood transfusion 0.01
PRC 54(43.5) 228(32.8) 0.02
FFP 26(21) 80(11.5) <0.05
PC 13 (10.5) 37 (5.3) 0.04
Intraoperativebloodloss(mL)* 150(50,412.5) 100(50,300) 0.06
Intraoperativeurineoutput(mL/kg/hr) <0.05
≤0.5 46(37.1) 80(11.5)
0.51-2 59(47.6) 389(55.9)
>2 19(15.3) 227(32.6)
Abbreviations: NSS: normal saline solution, PRC: packed red blood cell, FFP: fresh frozen plasma, PC: platelet count *Continuous data
were reported as median and IQR 1-3
Tantisarasart et al.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
121
Original Article
SMJ
Nonetheless, Blood transfusion consists of red blood
cells, fresh frozen plasma and platelet concentrate that
were used in AKI group more than others. Intraoperative
urine output less than 5 mL/kg occurred in 75% of AKI
group, while an intraoperative output less than 5 mL/kg
were found a higher risk in AKI group than non-AKI
group (P-value<0.05). Other independent variables were
not statistically dierent.
Table 3 presents multivariate logistic regression model
adjusted for patient’s status and anesthetic drug. Independent
variables signicantly associated with postoperative AKI
were COPD, ASA classication, part of surgery, Propofol
and Succinylated gelatin. ere was greater frequency of
AKI with higher ASA classication with ASA III and ASA
IV with V [OR 2.58 (1.41,4.73), OR 13.73 (5.94,31.76),
P-value <0.05], respectively. Part of surgery diered
signicantly between AKI and non-AKI group (P-value
<0.05). e risk of AKI was associated with abdominal
surgery, orthopedic surgery and urology surgery [OR
2.85 (1.27,6.38), OR 3.22(1.4,7.4), OR 8.31(2.86,24.12)],
respectively, excluding patients that underwent vascular
surgery. In contrast, Propofol conserved renal function
[OR 1.95 (1.21,3.16), P-value<0.05]. ere was a statistically
signicant risk for AKI with Succinylated gelatin use
[(OR 1.94(1.2,3.14), P-value <0.05)].
e nomogram comprised of COPD, ASA classication,
Part of surgery, Type of surgery, Propofol and Succinylated
gelatin (Fig 2). e score ranged from 20 to 100. e
nomogram has a good discriminative ability to identify
patients with postoperative AKI with AUC 0.746. e
calibrate curve shows that the apparent value is almost
the same as bias-corrected value; mean squared error
= 0.00043 (Fig 3). Cuto point at 63 was chosen, which
had the highest Youden index. e sensitivity, specicity,
positive predictive value (PPV) and negative predictive
value (NPV) of the prediction score was 83, 45, 33, and
89%, respectively.
e rate of renal replacement was higher in patients
with postoperative AKI compared with patients without
AKI (P-value<0.05). Period of ICU stay and hospital
stay was signicantly longer in AKI group than other
groups (P-value<0.05). Furthermore, 30-day mortality
was signicantly greater in AKI group (P-value <0.05).
TABLE 3. Best predictive score revealed by multivariate logistic regression.
Riskfactor Ref CrudeOR(95%CI) Adj.OR(95%CI) P(LR-test) RiskScore
COPD No 1.84(0.91,3..73) 2.02(0.93,4.38) 0.07 24
Type Scheduled
Emergency 1.92(1.3,2.82) 1.43(0.89,2.31) 0.14 14
ASA
II <0.05
III 2.19(1.23,3.9) 2.58(1.41,4.73) 37
IV-V 13.13(6.37,27.08) 13.73(5.94,31.76) 100
Part of surgery Neuro <0.05
Abdomen 1.9(0.92,3.95) 2.85(1.27,6.38) <0.05 42
Orthopedic 2.13(1,4.55) 3.22(1.4,7.4) <0.05 48
Urology 3.93(1.48,10.39) 8.31(2.86,24.12) <0.05 90
Vascular 1.91(0.89,4.11) 1.8(0.78,4.12) 0.16 25
Others 1.77(0.6,5.2) 2.35(0.73,7.57) 0.15 35
Propofol Yes 2.65(1.78,3.97) 1.63(1,2.66) <0.05 19
Succinylatedgelatin no 2.55(1.66,3.93) 1.94(1.2,3.14) <0.05 26
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
122
Fig 2. e nomogram explicated by logistic regression. Each predictor variable is presented by a line in the gure. e prediction scores
ranged from 0 to 100. e variables were gathered with point and paired with the probability of postoperative AKI. A drawing from was
made from each variable to the “Points” axis to indicate the points of the variable. e scores for all variables were summed and placed on
the “Total score” line. e predicted risk of postoperative AKI ranged from 5 to 80%.
Fig 3. e AKI nomogram and its performance. e Apparent value has a closer t to Bias-corrected value, which means the nomogram
had a good performance. Mean absolute error=0.014 Mean squared error=0.00036 Quantile of absolute error=0.03.
Tantisarasart et al.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
123
Original Article
SMJ
DISCUSSION
Nowadays, there is no eective equipment to identify
the probability of postoperative AKI in surgical patients,
particularly in super-elderly patients. AKI is a serious
consequence aer surgery. A prior study demonstrated
that approximately 6.8% of patients undergoing abdominal
surgery developed an episode of AKI.
16
In the present
study, almost 15% of AKI among super-elderly and
24.2% with AKI died in 30 days. e incidence of acute
kidney injury is highest in elderly and increases with
age.
3,17
ere are many reasons for the comparatively high
probability of AKI in super elderly patients, for instance,
comorbidities that produce AKI, comorbidities that
need intervention, medication or surgery that disarrange
kidney function, and structure and function that alter
with time.
3,9
According to meta-analysis, elderly patients
fail to recover from AKI and develop to chronic kidney
disease.
18
e risk of postoperative acute kidney injury is
associated with patient factors as age, type of surgery,
emergency surgery, part of surgery, ASA classication,
preoperative hemoglobin (Hb), preoperative creatinine,
preoperative creatinine clearance, and medical comorbidities
such as chronic obstructive lung disease (COPD). It is
not surprising that emergency surgery was found as a
part of the AKI prediction score, as in this study. A US
national study revealed that not only cardiac surgery, but
also urology, thoracic, orthopedic and malignancy were
associated with AKI.
19
In patients following non-cardiac
surgery, ASA classication III and IV were shown to
comprise a higher proportion of AKI than those with
ASA classication I and II.
4
Other preoperative laboratory data, preoperative
anemia
7
high creatinine
20
, low creatinine clearance
6
and
low eGFR
2
were shown to be similar with previous studies.
In cardiac surgery, preoperative anemia increased AKI
rate from 1.8% to 3.2%. Meanwhile, patients receiving
blood transfusion had relative risk of AKI more than two
times compared without transfusion.
21,22
e mechanism
of blood transfusion that could cause AKI is still unclear.
Although, It was suggested that pathophysiology can
aggravate tissue oxygen delivery and stimulate inammatory
process and oxidative stress.
23-25
Previous study showed
that higher baseline preoperative creatinine was signicant
risk factor for AKI.
26
Apart from urine output, Mizota T
et al. found perioperative urine output less than 0.3 mL/
kg/h increased probability of AKI.
27
Even if perioperative
urine output was dened dierently, the result remains
the same.
Neither the univariate analysis nor the nomogram
of the study found a signicant dierence with diuretic
use, which is a known AKI stimulator. It is probably due
to the reason that we are not given in lower creatinine
clearance patient. Nonetheless, The use of diuretic,
particularly furosemide has been demonstrated deleterious
in prevention and treatment of AKI.
28,29
In this study, the use of norepinephrine and
epinephrine was essential factor. Norepinephrine, is
frequently used to restore MAP from 60 to 75 mmHg
and increase renal oxygen delivery (RDO
2
), Glomerular
ltration rate (GFR) and renal oxygenation.
30
On the
other hand, It has been shown that norepinephrine
decreases renal blood flow (RBF) and renal oxygen
delivery, which provoke renal ischemia.
31
Epinephrine
has the same ecacy as norepinephrine, but also causes
hyperglycemia, hyperlactatemia and acidosis.
31
Based
on current evidence, norepinephrine should be used
to restore blood pressure within autoregulatory values
TABLE 4. Postoperative consequence in patients with and without Acute Kidney Injury (AKI).
Characteristics AKI Non-AKI P-value
(N=124) (N=696)
Renalreplacementn(%) 15(12.1) 1(0.1) <0.05
PeriodofICUstay(days)* 2.5(0,12) 0(0,1) <0.05
Periodofhospital(days)* 14.5(8.8,28) 10(7,18)
Death30n(%) 30(24.2) 11(1.6) <0.05
*Continuous data were reported as median and IQR 1-3
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
124
in hypotensive vasodilated patients with acute kidney
injury.
31
According to previous knowledge, propofol is
down regulated by nitric oxidase synthase to preserve
renal function
32
and protect renal tissue via peroxidation
of lipid membrane.
33
Lee et al. reported that propofol
reduced the new onset of chronic kidney disease aer
nephrectomy.
34
erefore, the anti-inammatory eect
of propofol might be signicant in the use of TIVA
technique, whereas propofol in this study was used as
a sole induction agent.
Resuscitation uid, such as Succinylated gelatin
another factor which impacts postoperative acute kidney
injury. Recent guidelines suggest avoiding the use of
gelatin, as systemic colloids gather in proximal renal
tubules and disturb renal function, increasing the risk
of anaphylaxis, mortality, renal failure and bleeding.
35
AKI Predictor in surgical super-elderly patients
consist of baseline characteristics and perioperative data.
We have found a relationship AKI: COPD, type of surgery,
ASA classication, part of surgery and Succinylated
gelatin. Propofol was used as a defensive factor in this
study. Prior study on AKI prediction score in a dierent
setting, displayed a good discrimination which AUC
range above 0.7. For instance, for the eGFRpreSurg as
a predictive role among very elderly patients, AUC was
at 0.703.
2
is represented eGFRpreSurg at cut o point
70mL/min/1.73 m
2
as a risk factor of postoperative AKI.
Another study by Hong et al. reported the relationship
between HUGE formula and mortality in elderly patients
from hip fracture surgery with highest AUC of 0.78
(95%CI 0.667-0.892).
26
However, both of the predictors
were in super-elderly patient settings and the score
obtained from only preoperative data.
is nomogram score in the study is easily understood
and handly to adopt in preoperative period. Accordingly,
AKI is a preventable condition, nomogram may help
physicians to early detect AKI and handle it in proper
time. According to resource-limited countries such
as ailand, invasive monitoring in every case might
be impossible. e most eective prevention is early
detection in high risk, special monitoring, optimal uid
administration and nephrotoxic agent avoidance. From
the research principle, it is expected that the nomogram
will eectively assist in selection high risk AKI patients
for prevention and aggressive intervention.
Strength and limitation
e incidence of AKI might be precise because AKI
in this study was determined by both urine output and
serum creatinine. e result may be in indubitable in
the occurrence of AKI.
ere are some limitations of this study. Firstly,
although all super-elderly patients who underwent surgery
in the study period were retrieved to reduce selection
bias, patient bias was unavoidable due to retrospective
study. Secondly, this study could not analyze the cause
of AKI categorized into three groups: prerenal, renal and
postrenal due to lack of information in our electronic
hospital system. e rate of renal replacement therapy
was compared between AKI and non-AKI patients;
nonetheless there was a lack of information to identify
the reason for renal replacement therapy.
Finally, due to the lack of external validation of the
scoring system since this is a single center study and
there could have been sampling bias. Future studies
performed on the scoring system using data of multiple
centers can validate it.
CONCLUSION
is study found an increase incidence of postoperative
AKI among super-elderly patients and relationship
between AKI and morbidity and mortality. Despite, AKI
prediction score not being a denitive tool for observation
and monitoring, it can help physicians consider various
clinical risk factors in evaluating the chance of AKI.
is nomogram can help clinical physicians improve
the prognosis among super-elderly patients undergoing
surgery.
REFERENCE
1. Khan-Kheil AM, Khan HN. Surgical mortality in patients more
than 80 years of age. Ann R Coll Surg Engl 2016 Mar;98(3):177-
80.
2. Wu Q, Yang H, Bo H, Fu M, Zhong X, Liang G, et al. Predictive
role of estimated glomerular ltration rate prior to surgery
in postsurgical acute kidney injury among very elderly patients:
a retrospective cohort study. Ren Fail. 2019 Nov;41(1):866-74.
3. Coca SG. Acute kidney injury in elderly persons. Am J Kidney
Dis O J Natl Kidney Found. 2010 Jul;56(1):122-31.
4. Li N, Qiao H, Guo J-F, Yang H-Y, Li X-Y, Li S-L, et al. Preoperative
hypoalbuminemia was associated with acute kidney injury
in high-risk patients following non-cardiac surgery: a retrospective
cohort study. BMC Anesthesiol. 2019 Sep 2;19(1):171.
5. Kim H-J, Koh W-U, Kim S-G, Park H-S, Song J-G, Ro Y-J, et al.
Early postoperative albumin level following total knee arthroplasty
is associated with acute kidney injury: A retrospective analysis
of 1309 consecutive patients based on kidney disease improving
global outcomes criteria. Medicine (Baltimore). 2016 Aug;95(31):
e4489.
6. Noyez L, Plesiewicz I, Verheugt FWA. Estimated creatinine
clearance instead of plasma creatinine level as prognostic test for
postoperative renal function in patients undergoing coronary
artery bypass surgery. Eur J Cardio-orac Surg. 2006 Apr;29(4):
461-5.
7. Jang WY, Jung J-K, Lee DK, Han S-B. Intraoperative hypotension
is a risk factor for postoperative acute kidney injury aer femoral
Tantisarasart et al.
Volume 74, No.2: 2022 Siriraj Medical Journal
https://he02.tci-thaijo.org/index.php/sirirajmedj/index
125
Original Article
SMJ
neck fracture surgery: a retrospective study. BMC Musculoskelet
Disord. 2019 Mar 27;20(1):131.
8. An Y, Shen K, Ye Y. Risk factors for and the prevention of
acute kidney injury aer abdominal surgery. Surg Today. 2018
Jun;48(6):573-83.
9. Chaudery H, MacDonald N, Ahmad T, Chandra S, Tantri A,
Sivasakthi V, et al. Acute Kidney Injury and Risk of Death Aer
Elective Surgery: Prospective Analysis of Data From an
International Cohort Study. Anesth Analg. 2019 May;128(5):
1022-9.
10. Kang W, Wu X. Pre-, Intra-, and Post-Operative Factors for
Kidney Injury of Patients Underwent Cardiac Surgery: A
Retrospective Cohort Study. Med Sci Monit Int Med J Exp
Clin Res. 2019 Aug 6;25:5841-9.
11. Swedko PJ, Clark HD, Paramsothy K, Akbari A. Serum Creatinine
Is an Inadequate Screening Test for Renal Failure in Elderly
Patients. Arch Intern Med. 2003 Feb 10;163(3):35-60.
12. McPherson RA, Pincus MR, editors. Henry’s clinical diagnosis
and management by laboratory methods. 23
rd
edition. St. Louis,
Missouri: Elsevier; 2017. 1565 p.
13. Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent
reporting of a multivariable prediction model for individual
prognosis or diagnosis (TRIPOD): the TRIPOD Statement.
BMC Med [Internet]. 2015 Jan 6;13(1):1.
14. Kellum JA, Lameire N. Diagnosis, evaluation, and management
of acute kidney injury: a KDIGO summary (Part 1). Crit Care
Lond Engl. 2013 Feb 4;17(1):204.
15. R Core Team [Internet]. Vienna; 2020. Available from: https://
www.R-project.org/
16. Long TE, Helgason D, Helgadottir S, Palsson R, Gudbjartsson
T, Sigurdsson GH, et al. Acute Kidney Injury Aer Abdominal
Surgery: Incidence, Risk Factors, and Outcome. Anesth Analg.
2016 Jun;122(6):1912-20.
17. Sirikun J. Does Acute Kidney Injury Condition Aect Revised
BAUX Score in Predicting Mortality in Major Burn Patients?
Siriraj Med J. 2019 Apr 4;71(2):150-7.
18. Schmitt R, Coca S, Kanbay M, Tinetti ME, Cantley LG, Parikh
CR. Recovery of kidney function aer acute kidney injury in
the elderly: a systematic review and meta-analysis. Am J Kidney
Dis O J Natl Kidney Found. 2008 Aug;52(2):262-71.
19. Kheterpal S, Tremper KK, Heung M, Rosenberg AL, Englesbe
M, Shanks AM, et al. Development and Validation of an Acute
Kidney Injury Risk Index for Patients Undergoing General
Surgery: Results from a National Data Set. Anesthesiology.
2009 Mar 1;110(3):505-15. Available from: https://doi.org/10.1097/
ALN.0b013e3181979440
20. Chen EY, Michel G, Zhou B, Dai F, Akhtar S, Schonberger
RB. An Analysis of Anesthesia Induction Dosing in Female
Older Adults. Drugs Aging. 2020 Jun;37(6):435-46.
21. Karkouti K, Grocott HP, Hall R, Jessen ME, Kruger C, Lerner
AB, et al. Interrelationship of preoperative anemia, intraoperative
anemia, and red blood cell transfusion as potentially modiable
risk factors for acute kidney injury in cardiac surgery: a historical
multicentre cohort study. Can J Anaesth. 2015 Apr;62(4):377-
84.
22. Karkouti K, Wijeysundera DN, Yau TM, McCluskey SA, Chan
CT, Wong P-Y, et al. Inuence of erythrocyte transfusion on
the risk of acute kidney injury aer cardiac surgery diers in
anemic and nonanemic patients. Anesthesiology. 2011 Sep;115(3):
523-30.
23. James SL, Castle CD, Dingels ZV, Fox JT, Hamilton EB, Liu Z,
et al. Global injury morbidity and mortality from 1990 to 2017:
results from the Global Burden of Disease Study 2017. Inj
Prev. 2020 Oct;26(Supp 1):i96-i114.
24. Cockcro DW, Gault MH. Prediction of creatinine clearance
from serum creatinine. Nephron. 1976;16(1):31-41.
25. Funk I, Seibert E, Markau S, Girndt M. Clinical Course of
Acute Kidney Injury in Elderly Individuals Above 80 Years.
Kidney Blood Press Res. 2016;41(6):947-55.
26. Hong SE, Kim T-Y, Yoo J-H, Kim J-K, Kim SG, Kim HJ, et al.
Acute kidney injury can predict in-hospital and long-term
mortality in elderly patients undergoing hip fracture surgery.
PloS One. 2017 Apr 20;12(4):e0176259–e0176259.
27. Mizota T, Yamamoto Y, Hamada M, Matsukawa S, Shimizu
S, Kai S. Intraoperative oliguria predicts acute kidney injury
aer major abdominal surgery. Br J Anaesth. 2017 Dec 1;119(6):
1127-34.
28. Zhang Y, Jiang L, Wang B, Xi X. Epidemiological characteristics
of and risk factors for patients with postoperative acute kidney
injury: a multicenter prospective study in 30 Chinese intensive
care units. Int Urol Nephrol. 2018 Jul;50(7):1319-28.
29. Ejaz AA, Mohandas R. Are diuretics harmful in the management of
acute kidney injury? Curr Opin Nephrol Hypertens. 2014 Mar;
23(2):155-60.
30. Redfors B, Bragadottir G, Sellgren J, Swärd K, Ricksten S-E.
Eects of norepinephrine on renal perfusion, ltration and
oxygenation in vasodilatory shock and acute kidney injury.
Intensive Care Med. 2011 Jan;37(1):60-7.
31. Bellomo R, Wan L, May C. Vasoactive drugs and acute kidney
injury. Crit Care Med. 2008 Apr;36(4 Suppl):S179-186.
32. Dikmen B, Yagmurdur H, Akgul T, Astarci M, Ustun H,
Germiyanoglu C. Preventive eects of propofol and ketamine
on renal injury in unilateral ureteral obstruction. J Anesth.
2010 Feb;24(1):73-80.
33. Li H, Weng Y, Yuan S, Liu W, Yu H, Yu W. Eect of sevourane
and propofol on acute kidney injury in pediatric living donor
liver transplantation. Ann Transl Med. 2019 Jul;7(14):340.
34. Lee H-J, Bae J, Kwon Y, Jang HS, Yoo S, Jeong CW, et al.
General Anesthetic Agents and Renal Function aer Nephrectomy.
J Clin Med. 2019 Sep 24;8(10).
35. Brown RM, Semler MW. Fluid Management in Sepsis. J Intensive
Care Med. 2019 May;34(5):364-73.
