Independent Predictors of Mortality in ICU Patients with COVID-19

Abstract

Objective: Early identification of Coronavirus disease 2019 (COVID-19) patients at high mortality risk can improve patient care and prevent deaths. To identify prognostic predictors that increase COVID-19 patient mortality risk in the Intensive Care Unit (ICU). Methods: Retrospective analysis of clinical characteristics and serological biomarkers of ICU-COVID-19 patients was performed in a tertiary hospital from 24 March 2020 to 20 December 2020. Analysis was conducted on two groups of study participants: survivors and deceased. Multivariate logistic regression was used to determine mortality risk. In order to determine prognostic predictors, the ANOVA test was used to compare the data of serological biomarkers on the day of patients' admission to the ICU and on the 5th day of follow-up. Results: A total of 335 patients (54.65%) were in the deceased group, and 278 (45.35%) were in the survivors group. A statistically significant difference was found between the deceased and survivor groups regarding mean age (p

Full text

470
Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
Original Article / Özgün Araştırma
Independent Predictors of Mortality in ICU Patients with COVID-19
Mehmet Özel 1, Songül Araç 1, Hasan Akkoç 2, Eşref Araç 3
1 Department of Emergency Medicine, University of Health Sciences, Diyarbakır Gazi Yaşargil Training and Research Hospital, Diyarbakır, Turkey
2 Department of Pharmacology, Faculty of Medicine, Dicle University, Diyarbakır, Turkey
3 Dicle University, Faculty Of Medicine, Department Of Internal Medical Sciences, Diyarbakır, Turkey
Received: 16.10.2023; Revised: 05.12.2023; Accepted: 08.12.2023
Abstract
Objective: Early identification of Coronavirus disease 2019 (COVID-19) patients at high mortality risk can improve
patient care and prevent deaths. To identify prognostic predictors that increase COVID-19 patient mortality risk in the
Intensive Care Unit (ICU).
Methods: Retrospective analysis of clinical characteristics and serological biomarkers of ICU-COVID-19 patients was
performed in a tertiary hospital from 24 March 2020 to 20 December 2020. Analysis was conducted on two groups of
study participants: survivors and deceased. Multivariate logistic regression was used to determine mortality risk. In
order to determine prognostic predictors, the ANOVA test was used to compare the data of serological biomarkers on the
day of patients' admission to the ICU and on the 5th day of follow-up.
Results: A total of 335 patients (54.65%) were in the deceased group, and 278 (45.35%) were in the survivors group. A
statistically significant difference was found between the deceased and survivor groups regarding mean age (p<0.001).
According to multivariate analyses of patients' data, age, oxygen saturation, direct bilirubin, and ionized calcium were
independent predictors of mortality (p <0.05). According to this analysis, age (OR=1.035, p=0.002, 95%CI 1.013-1.058),
peripheral capillary oxygen saturation (SpO2) (OR=0.912, p<0.001, 95%CI 0,873-0.953), direct bilirubin (OR=6.821,
p=0.024, 95%CI 0.282-36.285), ionized calcium (OR=30.524, p=0.035, 95%CI 1.262-738.34) was found that it increased
the risk of mortality. In the multivariate logistic regression analysis, it was found that gender, age, and comorbidities had
the highest odds ratios in terms of mortality.
Conclusion: The study revealed that advanced age, low SpO2, high direct bilirubin, and elevated ionized calcium levels
were independent predictors of mortality for COVID-19 patients in the ICU.
Keywords: COVID-19, Intensive Care Unit, Mortality, Prognostic Predictors
DOI: 10.5798/dicletip.1411504
Correspondence / Yazışma Adresi: Mehmet Özel, Department of Emergency Medicine, University of Health Sciences, Diyarbakır Gazi Yasargil Training
and Research Hospital, Diyarbakır, Turkey e-mail: drmehmetozel@yahoo.com.tr

Ozel M., Arac S., Akkoc H., Arac E.
471
Yoğun Bakım Ünitesinde Yatan COVID-19 Hastalarının Bağımsız Mortalite Belirleyicileri
Öz
Amaç: Yüksek mortalite riskine sahip Yeni Koronavirüs Hastalığı (COVID-19) hastalarının erken teşhis edilmesi, hasta
bakımını artırabilir ve ölümleri önleyebilir. Bu çalışma yoğun bakım ünitesi (YBÜ)ndeki COVID-19 hastalarının mortalite
riskini artıran prognostik belirleyicileri belirlemeyi amaçlamıştır.
Yöntemler: YBÜ’de COVID-19 hastalarının klinik özelliklerinin ve serolojik belirteçlerinin retrospektif analizi, 24 Mart
2020'den 20 Aralık 2020'ye kadar olan dönemde bir üçüncü basamak hastanede gerçekleştirildi. Hastaların analizi sağ
kalanlar ve ölenler olmak üzere iki grupa ayrılarak yapıldı. Mortalite riskini belirlemek için çok değişkenli lojistik
regresyon kullanıldı. Prognostik belirleyicileri saptamak amacıyla, ANOVA testi kullanılarak hastaların YBÜ'ye kabul ve
takibin 5. günündeki serolojik belirteç verileri karşılaştırıldı.
Bulgular: Toplam 335 hasta (%54,65) ölen grup içindeyken, 278 hasta (%45,35) sağ kalanlar grubundaydı. Ortalama
yaş açısından ölenler ve sağ kalanlar grupları arasında istatistiksel olarak anlamlı bir fark bulundu (p<0.001). Hastaların
verilerine yönelik çok değişkenli lojistik regresyon analizlere göre yaş, periferik kapiller oksijen saturasyonu (SpO2),
total bilirubin ve iyonize kalsiyum mortalitenin bağımsız belirleyicileri olarak saptandı (p <0.05). Bu analize göre yaş
(OR=1.035, p=0.002, %95 CI 1.013-1.058), Spo2 (OR=0.912, p<0.001, %95 CI 0.873-0.953), total bilirubin (OR=6.821,
p=0.024, %95 CI 0.282-36.285), iyonize kalsiyum (OR=30.524, p=0.035, %95 CI 1.262-738.34) mortalite riskini artırdığı
bulundu. Çok değişkenli lojistik regresyon analizinde, cinsiyet, yaş ve komorbiditelerin, mortalite açısından en yüksek
odds oranlarına sahip olduğu bulundu.
Sonuç: Çalışma, COVID-19 hastalarında ileri yaşın, düşük SpO2, yüksek total bilirubin ve yüksek iyonize kalsiyum
seviyelerinin YBÜ'de mortalite için bağımsız belirleyiciler olduğunu ortaya koymuştur.
Anahtar kelimeler: COVID-19, Yoğun Bakım Ünitesi, Mortalite, Prognostik Belirleyiciler.
INTRODUCTION
Worldwide, 771 million cases and 6,9 million
deaths have been reported of COVID-19
according to World Health Organization (WHO)
data1. Additionally, WHO classified EG.5, known
as "Eris," as an "intriguing variant," signifying
the need for closer monitoring than other
COVID-19 subvariants due to mutations that
might enhance its infectivity or severity2. Even
though COVID-19 is associated with a high
mortality rate, the clinical and laboratory
determinants of mortality in hospitalized
COVID-19 patients remain controversial3.
Identifying mortality risk predictors early in
critical COVID-19 patients can improve
management and prevent mortality.
COVID-19 patients' clinical characteristics,
serologic biomarkers, and risk factors can be
used to determine their severity. Studies have
identified the main factors associated with
COVID-19 fatalities, such as older age, diabetes
mellitus, cardiovascular disease, hypertension,
obesity and CBC, platelet count, lymphocyte
count, IL-6, and serum ferritin, lower albumin
levels, increased D-dimer, ferritin, and serum
troponin levels4-7. It is possible to estimate the
risk of death during hospitalization in COVID-19
patients using demographic, clinical, and
laboratory parameters mentioned in the
literature. Furthermore, determining the
parameters that can be used as independent
predictors can be vital to recognizing fatality
risks early.
COVID-19 patients with severe symptoms need
to be admitted to an intensive care unit, and
death rates are very high8. The fatality rate from
COVID-19 can be reduced if effective and rapid
treatment efforts are made early after
admission to the ICU for COVID-19 patients at
high risk of mortality. The best clinical and
laboratory parameters to predict early
mortality in COVID-19 patients during ICU
admission remain unclear. We aimed to
investigate the prognostic predictors that

Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
472
increase the mortality risk of ICU COVID-19
patients.
METHODS
A retrospective study in ICU COVID-19 patients
was conducted between 24 March 2020 and 20
December 2020. The study protocol was
approved by the institutional ethical board of a
tertiary hospital (Date: 31 December 2021,
Decision No: 967).The study recruited 613
COVID-19 patients older than 18 in the ICU of a
tertiary hospital. The diagnosis of COVID-19was
confirmed by real-time (RT)- polymerase chain
reaction(PCR) assay for SARS-CoV-2 from
nasopharyngeal and oropharyngeal swabs.
COVID-19 patients treated and monitored for at
least five days after admission to the ICU were
included in this study. In addition to the
duration of hospitalization mentioned above,
electronic medical records were evaluated for
COVID-19 ICU patients. This study excluded ICU
patients with negative RT-PCR results.
A comparison of survivors versus deceased
COVID-19 patients was conducted by dividing
the patients into two groups. Individuals
displaying clinical symptoms of pneumonia
such as fever, cough, sputum production, and
dyspnea, combined with at least one of the
following criteria, were categorized as ICU
patients: a respiratory rate exceeding 30
breaths per minute, severe respiratory distress,
SpO2 below 90% when breathing room air, and
the presence of potential COVID-19 pneumonia
indicators on CT scans. On admission ICU of all
patients, demographic-clinical characteristics
(risk factors, vital parameters), ordinary
laboratory test results with the inclusion of
complete blood count (CBC) along with
differentiation, , creatinine, cardiac troponin I,
lactate dehydrogenase (LDH),
aminotransferases (AST and ALT), blood urea
nitrogen, albumin, total bilirubin (Tbil), direct
bilirubin (Dbil),D-dimer, ferritin, procalcitonin,
C-Reactive Protein (CRP), International
normalized ratio (INR), and blood gas results of
patients were recorded.
Statistical Analysis
The arithmetic mean and standard deviation
were calculated for numerical data, while
frequency and percentage were used for
categorical data. In order to perform analysis,
IBM SPSS Statistics for Windows, version 21.0
(IBM Corp., Armonk, NY, USA) was used.
Analysis of categorical data was conducted
using the Chi-square test. Shapiro-Wilks tests
were performed on the numerical data to
determine their conformity to the normal
distribution. The Mann-Whitney U test was
utilized in data analysis that did not comply
with the normal distribution, and the 95%
Confidence Interval (CI) min-max values of
these data were shown in parentheses. In
univariate logistic regression analyses, clinical
patient risk factors were recorded that were
noteworthy in standard analyses. A 95% CI was
included with the odds ratios (ORs). A logistic
regression analysis was used to count variables
that continued to be statistically significant
after univariate analysis. Mortality risk was
determined by multivariate logistic regression.
Mortality risk was defined by multivariate
logistic regression. To reveal the prognosis
predictors; the Anova test was used with
repeated measurements comparing the
biochemical data of day 1 and day 5. In this
analysis, 1st and 5th-day measurements were
taken as within-subjects, and two different
(survivors vs deceased) patient groups were
taken as between-subjects. A p-value below
0.05 was examined remarkably for whole
analyses.
RESULTS
This study included a total of 613 patients,
consisting of 270 women and 343 men. Among
all study patients, two groups were formed
survivors [278 (45,35%) patients] and
deceased [335(54.65%) patients]. A total of 613

Ozel M., Arac S., Akkoc H., Arac E.
473
patients were aged 68 ±15 (95 CI: 66,93-
69,59%) years. A statistically significant
difference in mean age was found between the
deceased group and the survivors' group
(p<0.001). Comorbidities such as diabetes were
statistically significantly higher in the deceased
group than in the survivors (p= 0.037).
Demographic data, relevant comorbidities, and
laboratory are in Table I.
Table I: Demographic data, presence of comorbidities, vital parametres and laboratory parameters of all groups
All patients (n:613)
Survivor (n:278)
Deceased (n:335)
p
n (%) Mean±SD (%95 CI min-max) n (%) Mean±SD (%95 CI min-max) n (%) Mean±SD (%95 CI min-max)
Gender
0.838
Female
270 (44)
124 (44.6)
146 (43.6)
Male
343 (56)
154 (55.4)
189 (56.4)
Age (year)
68.26±14.74 (66.93-69.59)
6164±1631 (58.91-64.38)
7101±1311 (69.60-72.42)
0
Underlying
Comorbidities
n (%)
n (%)
n (%)
Hypertension
319 (52)
130 (46.8)
189 (56.4)
0.055
Diabetes Mellitus
178 (29)
66 (23.7)
112 (33.4)
0.037
Chronic obstructive
pulmonary disease-
asthma
82 (13.4)
42 (15.1)
40 (11.9)
0.431
Coronary Artery
Disease/ Heart
Failure
107 (17.5)
42 (15.1)
65 (19.4)
0.269
Oxygen Saturation
85.44±9.63 (84.57-86.31)
89.87±5.49 (88.94-90.79)
83.6±10.36 (82.48-84.71)
0
White Blood Cell
(4.000-
10.000/mm3)
8.29±5.24 (8.82-9.76)
8.30±4.72 (7.51-9.09)
9.7±5.4 (9.12-10.28)
0.004
Neutrophil(2.000-
7.000/mm3)
7.7±4.74 (7.27-8.12)
6.56±3.8 (5.92-7.2)
8.17±5 (7.63-8.71)
0.001
Lymphocytes (800-
4000/mm3)
1.15±1.89 (0.98-1.32)
1.32±2.53 (0.89-1.74)
1.08±1.54 (0.92-1.25)
0.005
Monocytes
0.4±0.24 (0.38-0.42)
0.4±0.24 (0.35-0.44)
0.4±0.24 (0.38-0.43)
0.695
Eosinophil
0.01±0.05 (0.002-0.01)
0.02±0.05 (0.01-0.03)
0.01±0.05 (0.008-0.02)
0.002
Basophil
0.02±0.03 (0.02-0.02)
0.01±0.01 (0.01-0.02)
0.02±0.03 (0.02-0.03)
0.014
Hemoglobin (11-16
gr/dl)
13.01±2.05 (12.83-13.2)
13.30±2.15 (12.94-13.66)
12.89±2 (12.68-13.11)
0.014
Hematocrit (37-54
%)
40.98±6.1 (40.43-41.53)
41.7±6.22 (40.65-42.74)
40.68±6.04 (40.03-41.33)
0.032
MCV
88.76±7.49 (88.08-89.43)
87.4±6.34 (86.34-88.46)
89.32±7.85 (88.47-90.16)
0.008
Platelet (150.000-
450.000/mm3)
209.13±84.07 (201.54-216.72)
218.75±80.68 (205.22-232.28)
205.14±85.24 (195.97-214.3)
0.059
RDW CW
14.3±1.78 (14.14-14.46)
13.97±1.68 (13.69-14.25)
14.44±1.8 (14.24-14.63)
0
RDW SD
47.15±6.08 (46.60-47.7)
45.43±5.69 (44.48-46.39)
47.87±6.1 (47.21-48.52)
0
Albumin
(34-48 g/L)
30.44±5.25 (29.96-30.91)
31.92±5.62 (30.98-32.87)
29.82±4.96 (29.29-30.36)
0
ALT
(0-41 U/L)
47.45±110.61 (37.46-57.44)
42.05±57.6 (32.39-51.71)
49.7±126.29 (36.1-63.29)
0.899

Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
474
AST
(0-40 U/L)
78.94±249.7 (56.38-101.5)
60.46±145.26 (36.10-84.82)
86.63±281.83 (56.29-116.97)
0.013
C-reactive protein
(0-5 mg/L)
126.84±82.98 (119.33-134.34)
109.53±84.3 (95.39-123.67)
134.06±81.46 (125.28-142.84)
0.001
Total Calcium (8.8-
10.6 mg/dl)
8.79±0.63 (8.73-8.84)
8.8±0.59 (8.7-8.9)
8.78±0.65 (8.71-8.85)
0.68
E-GFR
58.31±28.34 (55.72-60.90)
71.64±27.96 (66.95-76.33)
52.59±26.56 (49.69-55.49)
0
Glucose
169.75±92.85 (161.36-178.14)
158.25±87.32 (143.61-172.9)
174.53±94.78 (164.33-184.74)
0.008
Chlorine (98-107
mmol/l)
103.13±6.03 (102.59-103.68)
102.53±5.31 (101.64-103.43)
103.38±6.3 (102.7-104.06)
0.149
Creatinine (0.72-
1.25 mg/dL)
1.56±1.57 (1.41-1.7)
1.14±0.93 (0.98-1.29)
1.73±1.74 (1.54-1.92)
0
Creatine Kinase
467.46±1967.47 (289.32-645.60)
401.25±2396.02 (0.58-803.1)
495.18±1760.79 (305.08-685.28)
0
Lactate
Dehydrogenase
(135-225 U/l)
470.4±417.45 (432.61-508.2)
388.13±210.08 (352.9-423.3)
504.85±474.35 (453.64-556.06)
0
Potassium(3.5-5.1
mmol/L)
4.24-0.71 (4.17-4.30)
4.1±0.57 (4-4.19)
4.29±0.75 (4.21-4.38)
0.013
Sodium (134-146
mEq/L)
136.88±6.074 (136.33-137.43)
136.69±4.96 (135.85-137.52)
136.97±6.48 (136.27-137.66)
0.672
Indirect Bilirubin
0.31±0.21 (0.29-0.33)
0.3±0.18 (0.27-0.33)
0.31±0.22 0.29-0.33)
0.765
Direct Bilirubin (0-
0.3mg/dL)
0.38±0.33 (0.35-0.41)
0,29±0,2 (0,25-0,32)
0.41±0.36 (0.37-0.45)
0
Total Bilirubin
0.68±0.49 (0.63-0.72)
0,63±0,38 (0,56-0,69)
0.7±0.53 (0.64-0.76)
0.081
Urea(16-48mg/dl)
59.85±43.97 (55.87-63.82)
43.02±31.37 (37.75-48.28)
66.85±46.53 (61.84-71.86)
0
INR
1.25±0.23 (1.22-1.27)
1.23±0.2 (1.19-1.26)
1.26±0.24 (1.23-1.29)
0.268
D Dimer(0-243
ng/ml)
1375.8±4274.2 (980.7-1770.9)
535.06±743.41 (408.51-661.61)
1733.9±5040.7 (1176.9-2290.9)
0
Procalcitonin
2.51±8.88 (1,47-3,55)
0.90±2.53 (0.33-1.48)
3.12±10.24 (1.70-4.53)
0
Troponin (0-0.16
ng/ml)
0.32±1.35 (0.19-0.44)
0.15±0.32 (0.09-0.2)
0.39±1.6 (0.21-0.57)
0
Ferritin
946.25±2504.36 (712.67-1179.8)
744.10±775.21 (609.04-879.15)
1029.04±2929.14 (704.3-1353.7)
0.063
pH
7.38±0.16 (7.36-7.39)
7.4±0.06 (7.39-7.41)
7,36±0,18 (7,34-7,38)
0.01
Lactate
2.49±1.92 (2.31-2.67)
2.01±0.9 (1.85-2.16)
2.69±2,17 (2.45-2.93)
0.001
Bicarbonate(HCO3
ACT)
21.95±3.68 (21.61-22.29)
23.39±3.31 (22.82-23.96)
21.36±3.66 (20.96-21.76)
0
Base Exercise
-2.36±4.68 (-2.79-1.92)
-0.65±3.96 (-1.34-0.02)
-3.05±4.78 (-3.58-2.53)
0
Ionized Calcium
(Ca2+) (1.15-
1.35mmol/lt)
1.11±0.09 (1.10-1.12)
1.09±0.81 (1.08-1.11)
1.12±0.1 (1.11-1.13)
0.015
Categorical data are expressed as n (%) and numerical data as Mean±SD (95% CI min-max).
MCV: Mean Corpuscular Volume, E-GFR: Estimated Glomerular Filtration Rate, RDW CW: Red Cell Distribution Width - Coefficient of Variation, RDW SD:
Red Cell Distribution Width - Standard Deviation , INR: International Normalized Ratio, ALT Alanine AminotransferaseAST:Aspartate Aminotransferase
Mortality was associated with age, diabetes, low
oxygen saturation, and many hematological and
biochemical parameters. In multivariate
analyses of these data, age, oxygen saturation,
Dbil, and ionized calcium were independent
predictors of mortality (p<0.05) (Table I).Age,
gender, presence of comorbidity, oxygen
saturation value at admission ICU, CBC, and all
biochemistry and blood parameters were
evaluated with Univariate analysis. In Table I,

Ozel M., Arac S., Akkoc H., Arac E.
475
only statistically significant (p<0.05)
parameters were listed. Significant parameters
were included in the Multivariate analysis in the
second step. Multivariable logistic regression
was achieved by adjusting sex, age, and
admissible comorbidities with the highest
individual OR for mortality. According to the
Multivariable Logistic Regression, age
(OR=1.035, p=0.002, 95%CI 1.013-1.058),
SpO2; (OR=0.912, p<0.001, 95%CI 0,873-
0.953), Dbil (OR=6.821, p=0.024, 95%CI 0.282-
36.285), Ionized calcium (OR=30.524, p=0.035,
95%CI 1.262-738.34) were found that
parameters increased the risk of mortality (OR:
odds ratio) (Table II).
Table II: Univariate and multivariatelogistic regression analyses for predictors of mortality in patients with COVID-19
admitted to ICU
Univariate
Analysis
Multivariate
Analysis
95% CI 95% CI
Variable
p values
OR
Lower
Upper
p values
OR
Lower
Upper
Age
0
1.045
1.03
1.06
0.002
1.035
1.013
1.058
Diabetes Mellitus
0.038
1.613
1.027
2.535
0.07
1.76
0.955
3.246
Oxygen Saturation 0
0.9
0.87
0.931
0
0.912
0.873
0.953
White Blood Cell 0.009
1.061
1.015
1.11
0.19
0.856
0.678
1.08
Neutrophil 0.001
1.09
1.036
1.148
0.17
1.187
0.929
1.516
Basophil
0.013
0.69
0.516
0.924
0.692
0.166
0
1182.128
MCV 0.012
1.035
1.008
1.063
0.079
1.187
0.98
1.438
Platelet
0.11
0.998
0.996
1
RDW CW 0.01
1.191
1.042
1.362
0.11
2.339
0.825
6.636
RDW SD
0
1.093
1.045
1.142
0.174
0.805
0.588
1.1
Albumin 0
0.924
0.888
0.961
0.552
1.02
0.955
1.09
C-reactive protein 0.004
1.004
1.001
1,006
0.813
1.001
0.996
1.005
E-GFR 0
0.972
0.963
0.981
0.373
0.992
0.974
1.01
Creatinine
0
1.753
1.28
2.4
0.125
1.625
0.874
3.02
Lactate Dehydrogenase 0.003
1.002
1.001
1.003
0.293
1.001
0.999
1.003
Potassium 0.006
1.527
1.126
2.07
0.585
0.845
0.462
1.546
Direct Bilirubin 0
25.301
6.403
99.968
0.024
6.821
1.282
36.285
Urea
0
1.02
1.012
1.027
0.212
0,.99
0.975
1.006
D DİMER 0.003
1
1
1.001
0.194
1
1
1
pH 0.001
0.006
0
0.125
0.48
0.436
0.044
4.362
Ionized Calcium
0.016
15.178
1.658
138.938
0.035
30.524
1.262
738.335
Lactate 0
1.445
1.175
1.778
0.372
1.126
0.868
1.459
Bicarbonate (HCO3
ACT)
0
0.829
0.771
0.891
0.799
0.97
0.767
1.227
Base Exercise
0
0.872
0.825
0.923
0.757
0.971
0.806
1.17
Age, gender, presence of comorbidity, oxygen saturation, count of blood cells, and all biochemistry parameters were evaluated with Univariate analysis.
Including P values considered statistically significant (P < 0.05) OR: odds ratio MCV: Mean Corpuscular Volume, E-GFR: Estimated Glomerular Filtration
Rate, RDW CW: Red Cell Distribution Width - Coefficient of Variation, RDW SD: Red Cell Distribution Width - Standard Deviation
ANOVA tests were used for repeated
measurements on the 1st and 5th days to
compare the laboratory data of COVID-19
patients treated and followed at least 5 days
length of after admission ICU. In this analysis,
1st and 5th-day measurements were taken as
within-subjects, and two different (survivors vs
deceased) patient groups were taken as
between-subjects. p<0.05 was taken as the
statistical significance level. From all

Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
476
parameters; within subjects(*), between
subjects(**), Dbil (p:0.054, p:0), Tbil (p:0.106,
p:0.041), AST (p:0.064, p:0.009) and creatinine
(p:0.7, p:0.005)(Table 3). In spite of the fact that
the model parameter values on the 1st and 5th
days did not differ statistically significantly, the
values in the deceased’s group increased, while
those in the survivors’ group decreased (Table
III).Elevated AST levels during admission to the
ICU had more statistical significance than ALT
levels on the 5th day of admission (p=0.013).
Elevated ALT levels did not have a statistically
significant difference during admission to the
ICU and on the 5th day of admission (p=0.899).
Table III: Variance in repeated measures As measurements within subjects (*), between subjects (survivors vs
deceased**).
Total
(n:560)
Survivors
(n:266)
Deceased
(n:294)
p**
Total (n:560)
Survivors
(n:266)
Deceased
(n:294)
p**
Mean±SD Mean±SD Mean±SD Mean±SD Mean±SD Mean±SD
White Blood Cell
(4.000-
10.000/mm3)
1.
Day
9.04±5.11
8.25±4.64
9.38±5.28
0
E-GFR
1.
Day
59.34±27.88
71.99±27.76
53.39±25.92
0.01
5.
Day
11.41±6.34
8.68±4.71
12.65±6.60
5.
Day
61.37±28.60
78.38±18.98
53.38±28.89
p*
0
p*
0.01
Neutrophil
(2.000-
7.000/mm3)
1.
Day
7.44±4.57
6.52±3.68
7.86±4.86
0
Blood Glucose
1.
Day
168.21±92.05
157.66±87.78
172.98±93.67
0.17
7
5.
Day
9.74±5.46
6.88±3.42
11.03±5.72
5.
Day
157.19±85.49
136.96±68.40
166.34±90.82
p*
0
p*
0.009
Lymphocytes
(800-
4000/mm3)
1.
Day
1.16±1.97
1.33±2.59
1.08±1.62
0.68
Chlorine
1.
Day
102.91±5.48
102.46±5.36
103.11±5.53
0.00
1
5.
Day
1.15±2.66
1.29±3.06
1.09±2.46
5.
Day
107.50±10.24
104.40±5.15
108.90±11.58
p*
0.814
p*
0
Monocytes
1.
Day
0.40±0.22
0,38±0.19
0.41±0.24
0,58
7
Creatinine
(0.72-1.25
mg/dL)
1.
Day
1.47±1.39
1.08±0.56
1.65±1.60
0.00
5
5.
Day
0.44±0.30
0,43±0.24
0.44±0.33
5.
Day
1.51±1.46
0.89±0.53
1.79±1.65
p*
0.014
p*
0.7
Eosinophil
1.
Day
0.01±0.05
0.02±0.05
0.01±0.06
0.03
6
Creatine
Kinase
1.
Day
458.69±204.,5
3
409.33±2449.
47
481.26±1839.
14
0.46
5.
Day
0.03±0.08
0.05±0.09
0.02±0.07
5.
Day
279.05±541.0
8
122.66±251.7
7
350.53±617.9
4
p*
0
p*
0.048
Basophile
1.
Day
0.02±0.03
0.01±0.01
0.02±0.03
0.06
9
Lactate
Dehydrogena
se (135-225
U/l)
1.
Day
432.23±284.9
5
393.85±212.3
7
449.84±311.4
1
0.00
1
5.
Day
0.04±0.04
0.03±0.02
0.04±0.04
5.
Day
724.96±1020.
18
439.71±189.6
6
855.79±12036
2

Ozel M., Arac S., Akkoc H., Arac E.
477
p*
0
p*
0
Hemoglobin
(11-16 gr/dl)
1.
Day
13.07±1.99
13.34±2.07
12.95±1.95
0.29
5
Potassium
(3,5-5,1
mmol/L)
1.
Day
4.21±0.68
4.09±0.57
4.27±0.73
0.80
6
5.
Day
12.08±1.94
12.46±1.86
11.90±1.96
5.
Day
4.23±0.77
4.09±0.60
4.29±0.82
p*
0
p*
0.833
Hematocrit (3 7-
54 %)
1.
Day
41.12±5.95
41.83±5.99
40.80±5.91
0.90
8
Sodium (134-
146 mEq/L)
1.
Day
136.63±5.54
136.63±5.06
136.63±5.75
0
5.
Day
38.82±5.65
39.50±4.96
38.52±5.92
5.
Day
142.78±7.74
138.93±4.32
144.51±8.30
0
p*
0
MCV
1.
Day
88.49±7.36
87.40±6.45
88.99±7.69
0
Indirect
Bilirubin
1.
Day
0.31±0.21
0.31±0.18
0.32±0.22
0.34
7
5.
Day
89.17±7.43
86.99±6.45
90.16±7.64
5.
Day
1.16±12.53
0.30±0.18
1.55±15.12
p*
0.021
p*
0.351
Platelet(150.00
0-
450.000/mm3)
1.
Day
208.94±83.3
7
218.63±80.0
3
204.55±84.6
0
0.00
1
DirectBilirubi
n (0-
0.3mg/dL)
1.
Day
0.37±0.30
0.29±0.21
0.40±0.32
0
5.
Day
251.14±110.
14
282.32±117.
46
237.04±103.
83
5.
Day
0.41±0.36
0.27±0.14
0.48±0.41
p*
0
p*
0.054
RDW-CW
1.
Day
14.28±1.82
13.98±1.72
14.42±1.86
0.00
3
Total
Bilirubin
1.
Day
0.68±0.47
0.64±0.38
0.69±0.51
0.04
1
5.
Day
14.57±2.19
14.04±2.34
14.81±2.08
5.
Day
0.73±0.54
0.63±0.33
0.78±0.60
p*
0
p*
0.106
RDW-SD
1.
Day
46.96±6.01
45.45±5.78
47.64±6.00
0
Urea (16-
48mg/dl)
1.
Day
57.96±43.24
42.31±29.98
65.07±46.40
0
5.
Day
48.16±7.00
44.88±5.87
49.65±6.97
5.
Day
72.02±52.76
41.05±28.15
86.08±55.31
p*
0
p*
0
Albumin (34-48
g/L)
1.
Day
30.69±5.10
31.93±5.55
30.12±4.78
0.07
8
pH
1.
Day
7.38±0.16
7,40±0,06
7.37±0.19
0.04
3
5.
Day
25.60±4.45
27.43±4.63
24.77±4.11
5.
Day
7.34±0.12
7.39±0.06
7.32±0.13
p*
0
p*
0.004
Alanine
Aminotransfera
se (0-41 U/L)
1.
Day
40.20±56.87
42.39±58.60
39.21±56.15
0.01
3
Ionised
Calcium
1.
Day
1.11±0.09
1,09±0.08
1.12±0.10
0.73
5.
Day
91.32±290.6
6
42.48±37.11
113.42±347.
34
5.
Day
1.15±0.53
1.12±0.09
1.16±0.63
p*
0.013
p*
0.28
Aspartate
Aminotransfera
se (0-40 U/L)
1.
Day
63.66±192.2
9
61.61±148.3
7
64.59±209.3
8
0.00
9
Lactate
1.
Day
2.39±1,61
1,99±0,94
2,56±1,78
0.02
3

Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
478
5.
Day
134.40±457.
29
42.66±34.97
175.90±545.
83
5.
Day
2.87±2.53
1.95±0.83
3.24±2.87
p*
0.064
p*
0.045
C-reactive
protein (0-5
mg/L)
1.
Day
12.90±80.83
109.60±84.7
0
127.48±78.5
2
0.00
3
Bicarbonate
(HCO3 ACT)
1.
Day
22.14±3.51
23.63±3.21
21.55±3.45
0.30
5
5.
Day
112.45±81.6
5
79.48±62.08
127.42±85.1
0
5.
Day
22.03±4.85
23.95±3.64
21.28±5.05
p*
0.003
p*
0.938
Calcium (8.8-
10.6 mg/dl)
1.
Day
8.79±0.62
8.80±0.58
8.79±0.64
0.15
2
Base Exercise
1.
Da
y
-2.14±4.53
-0.49±3.98
-2.80±4.57
0.51
5
5.
Da
y
8.73±0.60
8.81±0.57
8.69±0.61
5.
Day
-2.12±5.97
-0.13±4.60
-2.91±6.27
p*
0.24
p*
0.724
MCV: Mean Corpuscular Volume, E-GFR: Estimated Glomerular Filtration Rate, RDW CW: Red Cell Distribution Width - Coefficient of Variation, RDW SD:
Red Cell Distribution Width - Standard Deviation
DISCUSSION
Several studies have found significant increases
in fatalities, respiratory failure, and ICU
admissions among hospitalized COVID-19
patients during the pandemic3,4,8,9. The poor
clinical outcomes of COVID-19 patients who are
identified early (such as mortality, intubation,
ICU admission, etc.) can be reduced through
effective, rapid treatment efforts. The clinical
characteristics, risk factors, and laboratory
parameters can be used to determine the
severity of COVID-19.
This study identified useful parameters as
independent predictors of mortality in COVID-
19 patients at ICU admission. Among COVID-19
ICU patients hospitalized in the ICU, age, oxygen
saturation, direct bilirubin, and ionized calcium
were predictive of mortality. During ICU
admission, these parameters significantly
increased mortality risk; Spo2 (OR=1,096,
95%CI 1.049-1.145, p<0.001), advanced age
(OR=1.035, 95%CI 1.013-1.058, p=0.002),
direct bilirubin (OR=6.821, 95%CI 0.282-
36.285, p=0.024) and ionized calcium
(OR=30.524, 95%CI 1.262-738.34, p=0.035).
During admission to the ICU and on the fifth day,
Dbil, Tbil, AST, and creatinine levels were found
to be remarkable predictors of mortality.
SpO2 is a parameter that we use to evaluate
respiratory functions and provides the clinician
with a noninvasive measurement opportunity
thanks to its correlation with partial arterial
oxygenation. Studies have shown that advanced
age and low oxygen saturation value increase
ICU admission and mortality10-11. SpO2 has been
found to be an important predictor of COVID-19
severity in another study12. In line with the
literature, we found that mortality was more
likely to occur in older adults and those with low
Spo2.
Several viral infections are associated with
changes in serum ionized calcium levels
according to Crespi et al13. As a result of COVID-
19, serum calcium levels decrease. The
presence of hypocalcemia was associated with a
longer hospital stay (especially in the ICU) and
higher mortality rates in studies of COVID-1913-
15.Cappellini et al16 found that the total and
ionized calcium levels of patients with COVID-
19 were lower, arguing that this was a strategy
developed by cells so that the virus would not
utilize calcium. In severe COVID-19 patients,

Ozel M., Arac S., Akkoc H., Arac E.
479
Crespi et al14 suggest that ionized hypocalcemia
is a host defense as a result of pathogen
adaptation. Based on our findings, we believe
that deceased patients are unable to develop
their host defense mechanisms as a result of
their adaptation to pathogens. In accordance
with the literature, we found that ionized
calcium levels were statistically significantly
higher in deceased patients than in survivors.
The deceased and survivors showed no
significant difference in total calcium levels, but
the survivors had lower ionized calcium levels.
A large multicenter retrospective study by Fu et
al17 found elevated liver biochemistry levels in
COVID-19 patients. It was also recommended
that abnormal total bilirubin at admission is
associated with poor prognosis17. Among bile
duct cells, type II alveolar epithelial cells, and
ACE-2-expressing bile duct cells, Wentao et al18
reported that ACE2 significantly affects COVID-
19 infection. In recent studies on COVID-19, it
was realized that expression of ACE2 in
cholangiocytes can directly damage the bile
ducts and cause a potent mechanism of
infection by the virus by using ACE2 as a
receptor. It has also been suggested that high
total bilirubin may be associated with bile duct
cell damage rather than direct hepatic cell
damage caused by SARS-CoV-219,20.
As an antioxidant, anti-inflammatory, and other
vital physiological properties, bilirubin is
widely recognized as a preventive bioactive
particle21,22. A recent study found that bilirubin
not only defends against inflammation but also
has potent antiviral properties that may be
useful in fighting COVID-1923. According to
Patel et al24, elevated serum bilirubin levels
were associated with poor outcomes, such as
death, in septic patients. Researchers
discovered that the prognosis of septic patients
was dependent on direct bilirubin rather than
total bilirubin. Additionally, direct bilirubin has
a better predictive value than total bilirubin in
patients with septicemia25. In this study, we
determined that direct bilirubin and total
bilirubin increase the risk of mortality and that
direct bilirubin can also be used as a prognostic
predictor in the follow-up process. The direct
bilirubin levels during admission to the ICU and
on the fifth day were statistically significant
parameters. This revealed the importance of
direct bilirubin both during admission and
follow-up. In addition, although total bilirubin
was not detected as a mortality predictor during
admission, and also stands out in clinical follow-
up in this study.Bilirubin, an indicator of
mortality, can be a useful tool for predicting
death from COVID-19 pneumonia, according to
this study, which is relevant to the literature.
A common side effect of SARS-CoV-2 infection is
liver damage. The damage may have been
caused by three different factors, including;
direct cytopathic effect on hepatocytes, hypoxic
damage caused by the disease, and hepatocyte
damage caused by drugs used in treatment18.
This study found that elevated AST levels were
statistically significant during admission to the
ICU, whereas elevated ALT levels had no
statistically significant difference during
admission and 5th-day measurements. While
AST levels declined in survivors, they increased
in the deceased group during follow-ups. It
shows that AST levels can be used to predict
prognoses. In a study by Wisniewska et al26 AST
levels were found to be remarkably higher in
severe COVID-19 patients, but no difference was
found in ALT levels. They suggested that liver
injury is secondary to inflammatory response
and hypoxia rather than viral injury and that the
contribution of AST from sources other than the
liver, particularly muscle, should also be
considered. As Fang et al27 noted in their study,
early elevated AST levels were associated with
indicators of disease severity, suggesting that
immune-mediated inflammation might be
crucial to liver impairment in COVID-19
patients. Since AST levels were statistically
significant during admission, decreased in the

Dicle Tıp Dergisi / Dicle Med J (2023) 50 (4) : 470-481
480
surviving group, and increased in the deceased
group in our study, we found that AST levels are
more useful than ALT levels.
COVID-19, classified as a severe respiratory
disease, exerts its effects on multiple organ
systems, including the heart, brain, vessel
endothelium, and kidneys28. Patient prognosis
was efficiently predicted by creatinine levels in
COVID-1928,29. According to a meta-analysis by
Kazemi et al29, it indicates a direct correlation
between the severity of COVID-19 and the levels
of creatinine. According to a meta-analysis,
creatinine levels are remarkably related to raised
disease severity and might be a useful prognostic
factor30. Consistent with existing literature, our
study identified a significant correlation between
disease severity and creatinine levels. Notably, the
creatinine levels during ICU admission and on the
fifth day did not exhibit statistically significant
differences among patients. However, a highly
significant disparity in creatinine levels was
observed between survivors and deceased
patients.
As a result of considering risk factors, clinical
features, and serological tests, it is possible to
identify risk factors associated with fatal
outcomes. In addition, it is possible to identify
severe COVID-19 patients early and improve
COVID-19 patients' outcomes. Studies in the
literature may not be able to examine all possible
risk factors comprehensively. Each study
considers a different number and type of risk
factors. As a result of these disadvantages, we
used multivariate analysis of patient data with
models in our study, and we found that ionized
calcium, oxygen saturation, and direct bilirubin
were independent predictors of mortality.
Limitations
The study had some limitations. Retrospective
research was conducted in a single ICU with a
relatively small group of patients. However,
imaging features and ICU treatments for COVID-
19 lung involvement were not recorded, along
with risk factors, clinical features, and serological
tests. It is estimated that new information and
articles on COVID-19 ICU mortality are published
almost daily; therefore, the results of our study
cannot be regarded as comprehensive. To confirm
the reliability of the independent prediction
parameters for fatality in COVID-19 patients,
large-scale, long-term, and prospective studies
are needed.
CONCLUSION
This study showed that low saturation, advanced
age, high direct bilirubin, and high ionized calcium
levels are associated with higher mortality rates
during ICU admission. Additionally, it found that
direct bilirubin, total bilirubin, AST, and
creatinine parameters could be used as mortality
predictors during the ICU follow-up period.
Ethics Committee Approval:The study protocol was
approved by the institutional ethical board of a tertiary
hospital (Date: 31 December 2021, Decision No: 967).
Conflict of Interest: The authors declared no conflicts of
interest.
Financial Disclosure: The authors declared that this
study has received no financial support.
REFERENCES
1. WHO Coronavirus (COVID-19) Dashboard | WHO
Coronavirus (COVID-19) Dashboard With Vaccination
Data. Accessed 2023 October 5. Available from:
https://covid19.who.int/
2. Word Health Organization. EG.5 Initial Risk Evaluation,
9 August 2023. Accessed 2023 October 5. Available:
https://www.who.int/docs/default-
source/coronaviruse/09082023eg.5_ire_final.pdf?sfvrsn
=2aa2daee_1
3. A
djei S, Hong K, Molinari NM, et al. Mortality Risk
Among Patients Hospitalized Primarily for COVID-19
During the Omicron and Delta Variant Pandemic Periods
-United States, April 2020-June 2022. MMWR Morb
Mortal Wkly Rep. 2022; 71(37):1182-9.
4. Zanella A, Florio G, Antonelli M, et al. Time course of risk
factors associated with mortality of 1260 critically ill
patients with COVID-19 admitted to 24 Italian intensive
care units. Intensive Care Med. 2021; 47(9):995-1008.
5. Silverio A, Di Maio M, Citro R, et al. Cardiovascular risk
factors and mortality in hospitalized patients with COVID-
19: systematic review and meta-analysis of 45 studies
and 18,300 patients. BMC Cardiovasc Disord.
2021;21(1):23.

Ozel M., Arac S., Akkoc H., Arac E.
481
6. Zhou F, Yu T, Du R, et al. Clinical course and risk factors
for mortality of adult inpatients with COVID-19 in Wuhan,
China: a retrospective cohort study. Lancet.
2020;395(10229):1054-62.
7. Dessie ZG, Zewotir T. Mortality-related risk factors of
COVID-19: a systematic review and meta-analysis of 42
studies and 423,117 patients. BMC Infect Dis.
2021;21(1):855.
8. Armstrong RA, Kane AD, Kursumovic E, Oglesby FC,
Cook TM. Mortality in patients admitted to intensive care
with COVID-19: an updated systematic review and meta-
analysis of observational studies. Anaesthesia. 2021;
76(4):537-48.
9. Armstrong RA, Kane AD, Cook TM. Decreasing mortality
rates in ICU during the COVID-19 pandemic. Anaesthesia.
2021:76 (3):10.
10. Petrilli CM, Jones SA, Yang J, et al. Factors associated
with hospital admission and critical illness among 5279
people with coronavirus disease 2019 in New York City:
prospective cohort study. BMJ. 2020; 369:m1966.
11. Zhao Z, Chen A, Hou W, et al. Prediction model and risk
scores of ICU admission and mortality in COVID-19. PLoS
One. 2020; 15(7): e0236618
12. Li X, Marmar T, Xu Q, et al. Predictive indicators of
severe COVID-19 independent of comorbidities and
advanced age: a nested case-control study. Epidemiol
Infect. 2020; 148: e255.
13. Crespi B, Alcock J. Conflicts over calcium and the
treatment of covid-19. Evol Med Public Heal. 2020;
2019:149–56.
14. Martha JW, Wibowo A, Pranata R. Hypocalcemia is
associated with severe COVID-19: A systematic review
and meta-analysis. Diabetes Metab Syndr. 2021;
15(1):337–42.
15. Di Filippo L, Formenti AM, Rovere-Querini P, et al.
Hypocalcemia is highly prevalent and predicts
hospitalization in patients with COVID-19. Endocrine.
2020; 68(3): 475-8.
16. Cappellini F, Brivio R, Casati M, et al. Low levels of total
and ionized calcium in blood of COVID-19 patients. Clin
Chem Lab Med. 2020; 58(9): e171–3.
17. Fu Y, Zhu R, Bai T, et al. Clinical Features of COVID-19-
Infected Patients With Elevated Liver Biochemistries: A
Multicenter, Retrospective Study. Hepatology. 2021;
73(4):1509-20.
18. Ni W, Yang X, Yang D, et al. Role of angiotensin-
converting enzyme 2 (ACE2) in COVID-19. Crit Care.
2020; 24(1): 422.
19. Zhao Y, Zhao Z, Wang Y, et al. Single-Cell RNA
Expression Profiling of ACE2, the Receptor of SARS-CoV-
2. Am J Respir Crit Care Med. 2021; 203(6):782.
20. Zhang H, Kang Z, Gong H, et al. The digestive system is
a potential route of 2019-nCov infection: a bioinformatics
analysis based on single-cell transcriptomes. Gut 2020;
69:1010-18.
21. Huang FY, Peng Y, Huang BT, et al. The correlation
between serum total bilirubin and outcomes in patients
with different subtypes of coronary artery disease. Clin
Chim Acta. 2017; 465: 101–5.
22. Boon AC, Bulmer AC, Coombes JS, Fassett RG.
Circulating bilirubin and defense against kidney disease
and cardiovascular mortality: Mechanisms contributing
to protection in clinical investigations. Am J Physiol Renal
Physiol. 2014 ; 307(2): 123-36.
23. Wagener FADTG, Pickkers P, Peterson SJ,
Immenschuh S, Abraham NG. Targeting the Heme-Heme
Oxygenase System to Prevent Severe Complications
Following COVID-19 Infections. Antioxidants. 2020; 9(6):
540.
24. Patel JJ, Taneja A, Niccum D, et al. The association of
serum bilirubin levels on the outcomes of severe sepsis. J
Intensive Care Med. 2015; 30(1): 23–9.
25. Wu Y, Ren J, Wang G, et al. Direct bilirubin as a
prognostic biomarker in enteric fistula patients
complicated with sepsis: A case-control study. Int J Clin
Exp Med. 2014; 7(12): 5134–45.
26. Wiśniewska H, Skonieczna-Żydecka K, Parczewski M,
et al. Hepatotropic Properties of SARS-CoV-2-Preliminary
Results of Cross-Sectional Observational Study from the
First Wave COVID-19 Pandemic. J Clin Med. 2021; 10(4):
672.
27. Lei F, Liu YM, Zhou F, et al. Longitudinal Association
Between Markers of Liver Injury and Mortality in COVID-
19 in China. Hepatology. 2020; 72(2): 389–98.
28. Wu J, Shi L, Zhang P, Wang Y, Yang H. Is creatinine an
independent risk factor for predicting adverse outcomes
in COVID-19 patients? Transpl Infect Dis. 2021 ; 23(2):
e13539
29. Kazemi E, Soldoozi Nejat R, Ashkan F, Sheibani H. The
laboratory findings and different COVID-19 severities: a
systematic review and meta-analysis. Ann Clin Microbiol
Antimicrob. 2021; 20(1): 17.
30. Danwang C, Endomba FT, Nkeck JR, et al. A meta-
analysis of potential biomarkers associated with severity
of coronavirus disease 2019 (COVID-19). Biomark Res.
2020; 8(1): 37.