APACHE scoring as an indicator of mortality rate in ICU patients: a cohort study

Abstract

Introduction:
Predictive scoring systems are tools that assess the magnitude of a patient’s illness and forecast disease prognosis, usually in the form of mortality, in the ICU. We aimed to determine the mortality rate among patients admitted to ICU using the Acute Physiology and Chronic Health Evaluation II (APACHE II) scoring system correlating with lengths of stay in the ICU.

Methodology:
A cohort study using team approach to care was conducted from July 2021 through July 2022 at KRL Hospital. Five hundred fifty-two patients aged 18–40 years, admitted for medical or surgical reasons (other than cardiac) who stayed in the ICU for more than 24 h were included. The APACHE II score was determined using 12 physiological variables at the end of the first 24 h of ICU admission. Data were analyzed using IBM Corp. released in 2015 (IBM SPSS Statistics for Windows, Version 23.0, Armonk, New York).

Results:
The average age of study participants was 36.34±2.77, ranging from 18 to 40 years. Three hundred fifteen participants were males and 237 were females. Patients were categorized into four separate groups as per their respective APACHE II scores. Patients with an APACHE II score of 31–40 were assigned to group 1. Patients with an APACHE II score of 21–30 were assigned to group 2. Patients with an APACHE II score of 11–20 were assigned to group 3. Lastly, patients with an APACHE II score of 3–10 were assigned to group 4. All patients in group 1 and group 2 died and none survived. Groups 1 and 2 contained a sum of 228 patients. A total of 123 patients were assigned to group 3, out of which 88 patients (71.54%) survived and 35 patients (28.45%) died. From these observations, it is evident that a higher APACHE II score is correlated with increased mortality.

Conclusion:
APACHE II scoring serves as an early warning indication of death and prompts clinicians to upgrade their treatment protocol. This makes it a useful tool for the clinical prediction of ICU mortality.

Full text

APACHE scoring as an indicator of mortality rate
in ICU patients: a cohort study
Hassan Mumtaz, MBBS, MSPHa,*, Muhammad K. Ejaz, MBBSb, Muhammad Tayyab, MBBSa,
Laiba I. Vohra, MBBSc, Shova Sapkota, MBBSf, Mohammad Hasan, MBBSd, Muhammad Saqib, MBBSe
Introduction: Predictive scoring systems are tools that assess the magnitude of a patient’s illness and forecast disease prognosis,
usually in the form of mortality, in the ICU. We aimed to determine the mortality rate among patients admitted to ICU using the Acute
Physiology and Chronic Health Evaluation II (APACHE II) scoring system correlating with lengths of stay in the ICU.
Methodology: A cohort study using team approach to care was conducted from July 2021 through July 2022 at KRL Hospital. Five
hundred fifty-two patients aged 18–40 years, admitted for medical or surgical reasons (other than cardiac) who stayed in the ICU for
more than 24 h were included. The APACHE II score was determined using 12 physiological variables at the end of the first 24 h of
ICU admission. Data were analyzed using IBM Corp. released in 2015 (IBM SPSS Statistics for Windows, Version 23.0, Armonk,
New York).
Results: The average age of study participants was 36.34 ±2.77, ranging from 18 to 40 years. Three hundred fifteen participants
were males and 237 were females. Patients were categorized into four separate groups as per their respective APACHE II scores.
Patients with an APACHE II score of 31–40 were assigned to group 1. Patients with an APACHE II score of 21–30 were assigned to
group 2. Patients with an APACHE II score of 11–20 were assigned to group 3. Lastly, patients with an APACHE II score of 3–10 were
assigned to group 4. All patients in group 1 and group 2 died and none survived. Groups 1 and 2 contained a sum of 228 patients. A
total of 123 patients were assigned to group 3, out of which 88 patients (71.54%) survived and 35 patients (28.45%) died. From these
observations, it is evident that a higher APACHE II score is correlated with increased mortality.
Conclusion: APACHE II scoring serves as an early warning indication of death and prompts clinicians to upgrade their treatment
protocol. This makes it a useful tool for the clinical prediction of ICU mortality.
Keywords: adult intensive and critical care, APACHE II score, epidemiology
Introduction
Predictive scoring systems are tools that assess the magnitude of a
patient’s illness and forecast their prognosis, usually their mortal-
ity, in the ICU
[1]
. Clinical scoring systems are utilized to classify
risks, anticipate health outcomes, or enhance other clinical activ-
ities. They help doctors in patient care but are rarely applied reg-
ularly in clinical practice. This is because scoring systems can be
complex to use and require expertise and training. Other reasons
for the limited adoption ofthese scoring systems worldwide are but
are not limited to, reliability, inefficiency, and a scarcity of internal
or external accuracy assessment
[2]
. Scoring systems utilized in
adult ICU-admitted patients frequently are the following:
APACHE (acute physiology and chronic health evaluation), SAPS
(simplified acute physiology score), MPM (mortality prediction
model), ODIN (organ dysfunction and infection system), SOFA
(sequential organ failure assessment), MODS (multiple organs
dysfunction score), LOD (logistic organ dysfunction) model, and
TRIOS (three-day recalibrating ICU outcomes)
[3]
.
The APACHE score was introduced in 1981 by the Medical
Center of George Washington University and is conceivably the
most well-known and popular scale for determining the ser-
iousness of an acute illness
[4]
. The APACHE II has been estab-
lished for several research and clinical audit applications and is
still frequently used as a measure of disease severity in critically ill
patients confined to the ICU
[1]
. It is highly important for
HIGHLIGHTS
•This cohort study explains APACHE II (Acute Physiology
and Chronic Health Evaluation II) scoring serving as an
early warning indication of death and prompts clinicians to
upgrade the treatment protocol.
•It is a useful tool for the clinical prediction of hospital
mortality in patients.
•An effort to reduce mortality rates within the constraints of
the resources at their disposal.
•In terms of its predictive power and ease of use, the scoring
system is superior.
a
Health Services Academy, Islamabad,
b
Gujranwala Medical College, Gujranwala,
Punjab,
c
Ziauddin University,
d
Jinnah Postgraduate Medical Center, Karachi, Sindh,
e
Khyber Medical College, Peshawar, Khyber Pakhtunkhwa, Pakistan and
f
Kathmandu Medical College, Kathmandu, Nepal
Sponsorships or competing interests that may be relevant to content are disclosed at
the end of this article
Published online 24 March 2023
*Corresponding author. Address: Health Services Academy, Rawalpindi, Punjab
46000, Pakistan. E-mail address: hassanmumtaz.dr@gmail.com (H. Mumtaz).
Received 18 October 2022; Accepted 26 January 2023
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. This is an
open access article distributed under the terms of the Creative Commons
Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is
permissible to download and share the work provided it is properly cited. The work
cannot be changed in any way or used commercially without permission from the
journal.
Annals of Medicine & Surgery (2023) 85:416–421
http://dx.doi.org/10.1097/MS9.0000000000000264
’
Original Research
416
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023

healthcare staff to be routinely trained and comply with specific
standards for optimal use of the APACHE II scoring system
[5]
.
Three key factors make up this scoring system: the acute phy-
siology scores, age scores, and chronic health scores, generating a
point score from 0 to 71
[4]
. However, in terms of predicting
mortality, the APACHE II score is neither extremely sensitive nor
specific. This score system’s main drawback is that many patients
have multiple and diverse comorbid diseases, making it challen-
ging to fit them under just a single primary diagnostic category
[1]
.
The APACHE score exhibits a high level of selectivity and
accuracy, and by evaluating patient groups using the standar-
dized frequency of mortality ratios, hence the APACHE score
may be utilized as an ICU standard
[6]
.
The objective of this study is to determine the mortality rate
among ICU-admitted patients (who have multiple comorbid
conditions and cannot be totally categorized under a single
APACHE II primary diagnostic category) stratified as per the
APACHE II scoring system in relation to the length of stay in the
ICU. We think that patient stratification in relation to the length
of stay in ICU as per the APACHE II scoring system is necessary
as it may better showcase this scoring systems’strengths and
weaknesses as applied to a real-world ICU setting and help in its
improvement (by highlighting areas of improvement) or better
adoption (by demonstration of its accuracy in mortality predic-
tion) by researchers and future healthcare professionals,
respectively.
Methods
From July 2021 through July 2022, researchers at KRL
Hospital’s ICU used a team approach to care for study partici-
pants. Using the Raosoft sample size calculator
[7]
with a 95% CI
and a 5% margin of error, a sample size of 552 was selected.
Nonprobability consecutive sampling technique was used for
sample size calculation in our study.
Our study is fully compliant with the STROCSS 2021
(Strengthening the reporting of cohort, cross-sectional and case-
control studies in surgery) guidelines
[8]
. A complete STROCSS
2021 checklist has been provided as a supplementary file. Our
study has been registered on Research Registry with the following
UIN: researchregistry8340
[9]
. Our study is in accordance with the
Declaration of Helsinki.
Inclusion and exclusion criteria
Patients aged 18–40 years, admitted for medical or surgical rea-
sons (other than cardiac), who stayed in the ICU for more than
24 h were included. Patients who had a missing physiological
characteristic, those who had recently undergone CABG surgery,
and those who spent less than 24 h in the ICU were not included.
Data collection and analysis
Data were collected regarding demographic information, the
reason for the patient’s admission to the ICU, and whether or not
they had a chronic disease. The APACHE II score was determined
using 12 physiological variables at the end of the first 24 h of ICU
admission. The worst values of each variable were given points in
accordance with the APACHE II scoring system’s established
protocol. Similar weightage was applied to age and chronic
Figure 1. Age distribution in years in our cohort of patients is 32.81 ±4.75 (mean ±SD) and the length of ICU stay in days in our cohort of patients is 9.06 ±5.95
(mean ±SD); N=number of patients in our cohort.
Table 1
APACHE II score with mortality and length of ICU stay in days
correlational analysis using Pearson correlation coefficient.
Correlational statistical analysis Mortality Length of ICU stay in days
APACHE II score
Pearson correlation coefficient 0.843 −0.115
Pvalue (<0.05) 0.01 0.01
Number of patients in the study 552 552
APACHE II, Acute Physiology and Chronic Health Evaluation II.
Mumtaz et al. Annals of Medicine & Surgery (2023)
417
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023

health status. A patient’s APACHE II score was the sum of their
grades in categories A, B, and C. The levels of consciousness were
measured using the Glasgow Coma Scale. Assessments were
performed after the patient had recovered from the anesthetic
effects of surgery. The assessment was based on the intubated
patients’overall understanding, not just their capacity to com-
municate verbally. Both the patient’sfinal outcome (either shift
out or death) and the entire period of their ICU stay were docu-
mented. The APACHE II score was recorded on a proforma
created by the principal investigator.
SPSS version 23 by IBM was used for the statistical analysis. Age
and length of ICU stay are displayed as means ±SD shown in
Figure 1. The point biserial correlation test was used to determine
the statistical significance of the relationship between APACHE II
score and the patient’s ultimate outcome (i.e. mortality), shown in
detail in Table 1 and as a boxplot in Figure 3. Correlations between
APACHE II and age and length of hospital stay were calculated
using Pearson’s correlation coefficient shown in Table 1. For sta-
tistical significance, a Pvalue of less than 0.05 was used (Figure 2).
Results
Five hundred fifty-two patients were included in the study; 351
were males and 201 were females. Two hundred ninety-three
patients were admitted to the surgical ICU and two hundred fifty-
nine patients were admitted to the medical ICU. The mean age of
study participants was 32.81 ±4.75, ranging from 18 to 40 years,
as shown in a histogram in Figure 1 and in a tabular format in
Table 2. The mean length of days the patients stayed in ICU in our
cohort of patients was 9.06 ±5.95 days (mean ±SD). The mean
APACHE II score in our cohort of patients was 31.7 ±17.94
(mean ±SD). We conducted a correlational analysis using our
data, and the Pearson correlation coefficient showed a sig-
nificantly strong positive correlation of 0.84 at a Pvalue of 0.05
between APACHE II score and mortality, as seen in Table 1 and
Figure 3. Using our data, we also performed a correlational
analysis between APACHE II score and length of stay in ICU in
days and found out that the length of stay in the ICU was sig-
nificantly inversely correlated with a value of −0.115 to the
APACHE II score, as highlighted in Table 1.
For further analysis, we decided to categorize our patients into
four separate groups as per their respective APACHE II scores.
Table 2
Age and sex distribution of study participants based on groups
A and B.
Gender
Age (years) Males Females Total
Group A
18–30 0 105 105
Group B
31–40 315 132 447
Total 315 237 552
Figure 3. Box and whiskers plot expressing the correlation between APACHE II
(Acute Physiology and Chronic Health Evaluation II) score and mortality.
Figure 2. Acute Physiology and Chronic Health Evaluation II score (APACHE II)
distribution in our cohort of patients; N=number of patients in our cohort.
Table 3
Group categories as per APACHE II scores and patient outcomes.
Outcome
Group (APACHE II scoring) Discharged Died
Group 1 (31–40) 0 93
Group 2 (21–30) 0 135
Group 3 (11–20) 88 35
Group 4 (3–10) 201 0
APACHE II, Acute Physiology and Chronic Health Evaluation II.
Mumtaz et al. Annals of Medicine & Surgery (2023) Annals of Medicine & Surgery
418
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023

Patients with an APACHE II score of 31–40 were assigned to
group 1. Patients with an APACHE II score of 21–30 were
assigned to group 2.
Patients with an APACHE II score of 11–20 were assigned to
group 3. Lastly, patients with an APACHE II score of 3–10 were
assigned to group 4. Patients were also stratified based on age,
with those aged 30 years or below categorized into group A and
those aged 31–40 years categorized into group B. We found that
our cohort asymmetrically contained only females in the younger
age group and a quantitatively dominant portion of males in the
older age group, that is group B as tabulated in Table 2.
There were 93 patients in group 1 (APACHE II score of 31–40).
A total of 135 patients were categorized into group 2 (APACHE II
score of 21–30). All patients in both groups1 and 2 died. A total of
123 patients were assigned to group 3 (APACHE II score 11–20),
out of which 88 patients (71.54%) survived and 35 patients
(28.45%) died. There were 201 patients in group 4 (APACHE II
score of 3–10), and all survived. This shows that there is an
increased chance of mortality in patients with a high APACHE II
score and increased chances of survival with a lower APACHE II
score, as shown in Table 3. And figuratively from Figure 3.
In patients aged 18–30 (group A), 91 patients stayed in the ICU
for 1 day, 14 patients for 2 days, and no patient was admitted for
3 days. Similarly, in patients aged 31–40 (group B), 152 patients
stayed in the ICU for a single day, 253 patients for 2 days, and 42
patients for 3 days, as shown in Table 4. We did not see any
significant variation in the correlation between the APACHE II
score and mortality between men and women.
Table 4
Age distribution as per the length of stay in the ICU (in days).
Length of stay in the ICU (in days)
Age groups (years) 1 2 3 Total
18–30 91 14 0 105
31–40 152 253 42 447
Total 243 267 42 552
Figure 4. Graph denoting the trend of APACHE II (Acute Physiology and Chronic Health Evaluation II) score in relation to mortality between those who died and
those that survived.
Mumtaz et al. Annals of Medicine & Surgery (2023)
419
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023

Analysis of the APACHE II score after further stratification
shows that a score of 35 or above on the APACHE II score scale
almost invariably portends death, except for a few outliers.
Maximum deaths were seen in patients with an APACHE II score
of 39 in our cohort, and a maximum number of patients that
survived during ICU stay had an APACHE II score of 17. This can
be deduced from Figure 4.
Discussion
In addition to predicting outcomes, scoring systems like
APACHE II are also used to evaluate clinical performance, the
standard of care in the ICU, and to compare the effectiveness of
ICUs with one another
[10]
. Compared to other scoring systems,
APACHE II has a sensitivity of 89.9% and specificity of 97.6%;
SOFA has 90.1% sensitivity and 96.6% specificity; while
mNUTRIC score has 97.2% sensitivity and 74.0% specificity
[11]
.
However, the mortality risk is frequently overstated based on
APACHE II values. This is mainly because of the lack of proper
standard implementation as well as poor scoring system skills of
medical workers. To mitigate these issues and improve adoption,
strict clinical standards must be followed, and medical workers
utilizing these scores must get frequent training, in order for the
APACHE II scoring system to be used properly
[10]
.
Our study shows that there is an increased chance of survival in
patients with a lower APACHE II score and increased chances of
death with a higher APACHE II score. It is possible that in our
cohort of patients, death was preceded by therapeutic restrictions
brought on by a poor projected prognosis or hemodynamic insuf-
ficiency, or both. Renal failure, infection, and cardiac arrest were
indicators of death in a study reported by Berkel et al.
[12]
from the
Netherlands. We also had similar indicators of death in our study.
According to research conducted at Beth Israel Deaconess
Medical Center in Boston, Massachusetts, USA, patients with an
APACHE II score of 17 or higher on day 3 of their ICU stay are
best defined as being at high risk of fatality
[13]
. The study, con-
ducted at the Imam Khomeini Medical Center, found that
APACHE II was a more accurate predictor of mortality among
very ill patients. Grading is required that takes into account
prognostic indicators and incorporates continuous monitoring of
clinical status
[14]
.
Since the primary factor affecting ICU costs and resource use is
the length of stay, we noticed in our study that patients who were
admitted to the ICU for a lesser time survived and were dis-
charged, compared to those who stayed longer.
Our study had a few limitations, namely a smaller sample of
cases making it inefficient to draw more reliable conclusions, and
the presence of patients with complicated comorbidities in the
ICU not exactly reflected by the APACHE II score. In addition,
this study only took place in one location. Due to variations in
therapy, ending of treatment, and admission regulations depen-
dent on institutions, we could not generalize our study to other
centers. Nevertheless, these results can provide general informa-
tive data for the study of mortality prediction in critical patients in
the future.
Conclusion
APACHE II scoring serves as an early warning indication of death
and prompts clinicians to upgrade the treatment protocol,
making it a useful tool for the clinical prediction of ICU mortality
in patients. Patients who have been properly triaged may be given
intensive treatment in an effort to reduce their mortality rates
within the constraints of the resources at their disposal. In terms
of its predictive power and ease of use, the scoring system is
superior.
Ethical approval
Ethical approval granted by KRL Hospital.
Patient consent
According to the Declaration of Helsinki.
Sources of funding
No funding was received.
Author contribution
M.H.: determined the main concept; M.K.E.: collection of data;
H.M.: analyzed and interpreted data; M.T.: writing of the
manuscript; L.I.V. and S.S.: manuscript editing.
Conflicts of interest disclosure
No conflicts of interest were declared.
Research registration unique identifying number
(UIN)
1. Name of the registry: Research Registry.
2. Unique identifying number or registration ID: research-
registry8340.
3. Hyperlink to your specific registration (must be publicly
accessible and will be checked): https://www.researchregis
try.com/browse-theregistry#home/registrationdetails/
632bf6090414f80021027ddb/
Guarantor
Hassan Mumtaz.
Provenance and peer review
Not commissioned, externally peer-reviewed.
Acknowledgments
None.
References
[1] Mehrzad B, Seyed Sajjad E, Nasrollah M, et al. Use of the APACHE II
score for the assessment of outcome and mortality prediction in an
Iranian Medical–Surgical Intensive Care Unit. Arch Anesth Crit Care
2018;4:521-526.
Mumtaz et al. Annals of Medicine & Surgery (2023) Annals of Medicine & Surgery
420
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023

[2] Balkan B, Essay P, Subbian V, et al. Evaluating ICU clinical severity
scoring systems and machine learning applications: APACHE IV/IVa case
study. Annu Int Conf IEEE Eng Med Biol Soc 2018;2018:4073–6.
[3] Keegan MT, Gajic O, Afessa B. Severity of illness scoring systems in the
intensive care unit. Crit Care Med 2011;39:163–9.
[4] Farajzadeh M, Nasrollahi E, Bahramvand Y, et al. The use of APACHE II
Scoring System for predicting clinical outcome of patients admitted to the
intensive care unit: a report from a resource-limited center. Shraz E-Med J
2021;22:e102858.
[5] Polderman KH, Girbes AR, Thijs LG, et al. Accuracy and reliability of
APACHE II scoring in two intensive care units problems and pitfalls in the use
of APACHE II and suggestions for improvement. Anaesthesia 2001;56:47–50.
[6] Widyastuti Y, Zaki WA, Widodo U, et al. Predictive accuracy of the
APACHE IV scores on mortality and prolonged stay in the intensive care
unit of Dr Sardjito Hospital. Med J Malaysia 2022;77(Suppl 1):53–8.
[7] Raosoft. Sample Size Calculator, 2016. Accessed 1 September 2022.
http://www.raosoft.com/samplesize.html
[8] Mathew G, Agha R, Albrecht J, et al. STROCSS 2021: strengthening the
reporting of cohort, cross-sectional and case–control studies in surgery.
Int J Surg 2021;96:106165.
[9] APACHE Score as an Indicator of Mortality Rate in Intensive Care Unit
Patients: A Cohort Study [Electronic]. Research registry; 2022. Study
registration details in the Research Registry. Accessed 1 September 2022.
https://www.researchregistry.com/browse-the-registry#home/regis
trationdetails/632bf6090414f80021027ddb/
[10] Polderman KH, Girbes ARJ, Thijs LG, et al. Accuracy and reliability of
APACHE II scoring in two intensive care units problems and pitfalls in the use
of APACHE II and suggestions for improvement. Anaesthesia 2001;56:47–50.
[11] Kumar S, Gattani SC, Baheti AH, et al. Comparison of the performance
of APACHE II, SOFA, and mNUTRIC scoring systems in critically ill
patients: a 2-year cross-sectional study. Indian J Crit Care Med 2020;24:
1057–61.
[12] van Berkel A, van Lieshout J, Hellegering J, et al. Causes of death in
intensive care patients with a low APACHE II score. Neth J Med
2012;70:455–9.
[13] Tian Y, Yao Y, Zhou J, et al. Dynamic APACHE II score to predict the
outcome of intensive care unit patients. Front Med 2021;8:744907.
[14] Beigmohammadi MT, Amoozadeh L, Rezaei Motlagh F, et al. Mortality
predictive value of APACHE II and SOFA scores in COVID-19 patients in
the intensive care unit. Can Respir J 2022;2022:5129314.
Mumtaz et al. Annals of Medicine & Surgery (2023)
421
Downloaded from http://journals.lww.com/annals-of-medicine-and-surgery by BhDMf5ePHKav1zEoum1tQfN4a+
kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 04/01/2023