Journal of Postgraduate Medicine
 Open access journal indexed with Index Medicus & ISI's SCI  
Users online: 8545  
Home | Subscribe | Feedback | Login 
About Latest Articles Back-Issues Articlesmenu-bullet Search Instructions Online Submission Subscribe Etcetera Contact
 ::  Similar in PUBMED
 ::  Search Pubmed for
 ::  Search in Google Scholar for
 ::Related articles
 ::  Article in PDF (1,121 KB)
 ::  Citation Manager
 ::  Access Statistics
 ::  Reader Comments
 ::  Email Alert *
 ::  Add to My List *
* Registration required (free) 

  IN THIS Article
 ::  Abstract
  ::  Introduction
  ::  Subjects and Methods
  ::  Results
  ::  Discussion
 ::  References
 ::  Article Figures
 ::  Article Tables

 Article Access Statistics
    PDF Downloaded33    
    Comments [Add]    

Recommend this journal


  Table of Contents     
Year : 2022  |  Volume : 68  |  Issue : 4  |  Page : 199-206

Performance of prognostic scores in prediction of 30-day postoperative mortality in COVID-19 patients after emergency surgery: A retrospective cohort study

1 Department of Anaesthesiology and Critical Care, All India Institute of Medical Sciences, Bhopal, Madhya Pradesh, India
2 Department of Community and Family Medicine, All India Institute of Medical Sciences, Bhopal, Madhya Pradesh, India

Date of Submission28-Dec-2021
Date of Decision11-Mar-2022
Date of Acceptance26-Mar-2022
Date of Web Publication06-Oct-2022

Correspondence Address:
S T Karna
Department of Anaesthesiology and Critical Care, All India Institute of Medical Sciences, Bhopal, Madhya Pradesh
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jpgm.jpgm_1197_21

Rights and Permissions

 :: Abstract 

Background: Risk assessment with prognostic scoring, though important, is scarcely studied in emergency surgical patients with COVID-19 infection.
Methods and Material: We conducted a retrospective cohort study on adult emergency surgical patients with COVID-19 infection in our institute from 1 May 2020 to 31 October 2021 to find the 30-day postoperative mortality and predictive accuracy of prognostic scores. We assessed the demographic data, prognostic risk scores (American Society of Anesthesiologists-Physical Classification (ASA-PS), Sequential Organ Failure Assessment (SOFA), Quick SOFA (qSOFA), Physiologic and Operative Severity Score for the enUmeration of Mortality and Morbidity (POSSUM) and Portsmouth-POSSUM (P-POSSUM) scores), surgical and anesthetic factors. We assessed the postoperative morbidity using the Clavien-Dindo scale and recorded the 30-day mortality. Correlation of prognostic scores and mortality was evaluated using Univariate Cox proportional hazards regression, receiver operating characteristic curve (ROC), Youden's index and Hosmer- Lemeshow goodness of fit model.
Results: Emergency surgery was performed in 67 COVID-19 patients with postoperative complication and 30-day mortality rate of 33% and 19%, respectively. A positive qSOFA and ASAPS IIIE/IVE had a 9.03- and 12.7-times higher risk of mortality compared to a negative qSOFA and ASA-PS IE/IIE (P < 0.001), respectively. Every unit increase of SOFA, POSSUM and P-POSSUM scores was associated with a 50%, 18% and 17% higher risk of mortality, respectively. SOFA, POSSUM and P-POSSUM AUCROC curves showed good discrimination between survivors and non-survivors (AUC 0.8829, 0.85 and 0.86, respectively).
Conclusions: SOFA score has a higher sensitivity to predict 30-day postoperative mortality as compared to POSSUM and P-POSSUM. However, in absence of a control group of non-COVID-19 patients, actual risk attributable to COVID-19 infection could not be determined.

Keywords: COVID-19, emergency surgery, postoperative morbidity, prognostic risk indices

How to cite this article:
Karna S T, Gouroumourty R, Ahmad Z, Trivedi S, Thaware P, Singh P. Performance of prognostic scores in prediction of 30-day postoperative mortality in COVID-19 patients after emergency surgery: A retrospective cohort study. J Postgrad Med 2022;68:199-206

How to cite this URL:
Karna S T, Gouroumourty R, Ahmad Z, Trivedi S, Thaware P, Singh P. Performance of prognostic scores in prediction of 30-day postoperative mortality in COVID-19 patients after emergency surgery: A retrospective cohort study. J Postgrad Med [serial online] 2022 [cited 2023 Jun 8];68:199-206. Available from:

 :: Introduction Top

Emergency surgery is associated with higher morbidity or mortality, which may further be exacerbated by coronavirus disease 2019 (COVID-19).[1–3] In surgical patients, the American Society of Anesthesiologists physical status classification (ASA-PS), Physiologic and Operative Severity Score for numeration of Mortality and Morbidity (POSSUM) and its modification, Portsmouth POSSUM (P-POSSUM), are used for risk stratification.[4–8] In critically ill patients, Sequential Organ Failure Assessment (SOFA) is routinely used whereas the Quick SOFA (qSOFA) score is used to identify sepsis in non-ICU settings.[9–10]

There is paucity of literature to suggest which risk assessment score is appropriate in COVID-19 patients undergoing emergency surgery. To address this research gap, this study was conducted with the objective to study the 30-day mortality after emergency surgery in COVID-19 patients and assess the performance of these routinely used prognostic risk indices in prediction of postoperative mortality.

 :: Subjects and Methods Top

Study design and setting: This retrospective, observational, cohort study was conducted in a tertiary care institute, in Central India, which serves as a referral centre for COVID-19 cases.

Study population: Patients above 18 years age of either gender with confirmed COVID-19, who underwent emergency surgery from 1 May 2020 to 31 October 2021 in our institute were included. We defined a positive COVID-19 laboratory test as positive via reverse transcription-polymerase chain reaction (RTPCR) test within seven weeks prior to or within one week after emergency surgery. We excluded patients with a negative COVID-19 test, pediatric patients and minor surgical procedures like re-suturing, lumbar puncture, and tracheostomies.

Ethical issues: Data for this study was retrieved from another ongoing study which had received local institutional human ethics committee approval (IHEC LOP/IM03075). All patients who participated in this other study, gave their informed consent for the use of their anonymized data for scientific purposes.

Outcomes: The primary outcome was to find the 30-day mortality rate and the discrimination and calibration of risk assessment scores like ASA-PS, SOFA, qSOFA, POSSUM and P-POSSUM in prediction of postoperative mortality in COVID-19 patients undergoing emergency surgery.

Data collection: Perioperative anesthesia case file and electronic hospital records were used as sources of data. Besides anesthesiologists, the surgical team involved in care of the patient included obstetricians, neurosurgeons, ear, nose and throat surgeons, and general and orthopedic surgeons. Lung protective ventilation was used in all patients as a protocol during COVID-19 pandemic, ensuring driving pressure of <15 cm H2O with plateau pressure of <30 cmH2O. Compliance before start of surgery was recorded on anesthesia charts as ratio of tidal volume to driving pressure (plateau pressure minus positive end expiratory pressure).

The study investigator collected the following preoperative data: demographic characteristics, comorbidities, SARS-CoV-2 infection status, diagnosis and indication of surgery, presence of acute respiratory distress syndrome (ARDS) with severity as per Berlin classification, type of oxygen support and need for intensive care.[11] The prognostic scores were calculated by a separate investigator, who did not participate in collection of postoperative data. The ASA-PS classification, and SOFA and qSOFA scores were calculated as per standard definitions in the immediate preoperative period.[4],[9],[10] The POSSUM and P-POSSUM scores are a combination of 12 weighted physiological (PS) and six operative (OS) variables obtained for individual patients.[6],[8] Both scores were evaluated with inputs regarding operative data from the surgical team. A dedicated investigator entered the patient data into an online calculator (MDCalc) which was used to calculate the PS, OS, predicted morbidity as per POSSUM, and predicted mortality as per POSSUM and P-POSSUM.[12] Furthermore, the PS and OS were added to yield a total score (TS).

We noted the surgery, anesthetic and airway management techniques, need for ICU stay, morbidity over a 30-day post-operative period assessed by Clavien Dindo classification along with cumulative 30-day mortality.[13] The course of hospital stay was followed till hospital discharge or death. The data regarding postoperative complications was collected from the postoperative medical records of the patients with inputs from the surgical team involved in the care of the patient. All data collected into a standardized data collection tool by two dedicated study investigators was checked for internal consistency, completeness, and transcription errors by reviewing the patient's charts again before entry into an electronic database.

Sample size

In view of the exploratory nature of the study, convenience sampling was done with inclusion of all emergency surgical patients satisfying the inclusion and exclusion criteria during the study period.

Statistical analysis

Data entry was done in Microsoft Excel 2010, followed by analysis using statistical programming language R version 4.1.0.[14] Data was prepared for analysis using standard data cleaning and wrangling procedures using base R and the tidyverse, gtsummary packages. The model building, visualizations and diagnostics were run with the performance, ggplot, survival, and survminer packages in R software. Continuous data were summarized as mean (SD) and median (IQR) based on the distribution of data. Association between two categorical variables was determined using Chi-squared test and Fischer's exact as appropriate. Kruskal–Wallis rank sum test was used to determine the association between more than two categorical groups and the numerical determinants.

The primary outcome of the analysis was a 30-day mortality. The 30-day time point was chosen because it has been used in previous evaluations of prognostic scores. The secondary outcome was the area under the curve (AUC) of the receiver operating curve (ROC) of the ASA grade, SOFA, qSOFA, POSSUM, and P-POSSUM scores.

The univariate Cox proportional hazards regression analysis was used to assess hazard ratio for mortality from time of positive COVID-19 status to the final outcome. The ROC was plotted using the “ROCR” to assess the potential of the SOFA, POSSUM, and P-POSSUM scores in predicting mortality among COVID-19 emergency surgical patient; this was compared to their actual outcome of either survival or death within 30 postoperative days. The AUC was used to delineate quality of discrimination between values. For each point of measurement over the individual curve, Youden's index [YI = Sensitivity + Specificity -1] was used to determine the optimal cutoff. P value of < 0.05 was considered significant.

Model performance of POSSUM and P-POSSUM was assessed by measuring their ability to discriminate between the non-survivors and survivors, observed over expected mortality ratios and calibration fit across various risk bands using Hosmer-Lemeshow goodness of fit model. Large χ2HL values suggest poor fit, with calibration considered to be poor if P ≤ 0.05.

 :: Results Top

In the study period, 5,042 COVID-19 patients were admitted to our institute. Sixty-seven patients, satisfying the study criteria, underwent emergency surgery under anesthesia in our institute.

Baseline characteristics

The mean age of patients was 46.3 ± 15.6 years. One or more comorbidities were present in 44 patients (66%). Sixty-three patients (94%) were diagnosed preoperatively while 4 patients (6%) tested positive after surgery. 54 patients (81%) presented with headache, altered sensorium, or with respiratory, digestive or cardiac symptoms with a frequency of 32 (48%), 1 (1.49%), 22 (32.8%), 12 (17.9%), 2 (3%) respectively. Bilateral peripheral opacities on chest radiograph were documented in 27 patients (40%) before surgery. Out of the 17 patients who had ARDS before surgery, 16 (94%) presented with dyspnea, with severity of ARDS ranging from mild, moderate, to severe in 9 (13%), 6 (9.0%) and 2 (3.0%) patients, respectively. Proning sessions were given in 13 patients. Before surgery, 12 patients were on low-flow oxygen, non-invasive ventilation was needed in 3 patients while 2 patients required endotracheal intubation and invasive ventilation. Intubation and ventilation before surgery was needed in an obstetric patient with severe ARDS and a neurosurgical patient with altered sensorium and moderate ARDS. Of the 17 patients who had ARDS before surgery, 16 (94%) were found to be diagnosed as COVID-19 positive before surgery.

Intensive care support was required before surgery in 11 patients (16%): 4 patients were antenatal with COVID-19 ARDS (moderate in 3 and severe in 1); 4 patients had abdominal pathologies with associated sepsis, hemodynamic instability and ARDS; 1 had rhino-cerebral mucormycosis with coronary artery disease and cardiac failure; 1 had toe gangrene with moderate ARDS, septic shock, and 1 was a mechanically ventilated neurosurgical patient with intracranial bleed.

Though the median duration from symptom onset to hospital admission was 7 (IQR 2–14) days, surgery was done 14.5 (IQR 5.5–21) days after symptom onset in our patients with COVID-19.

Prognostic scores and predicted outcomes

Only 24% of patients were ASA IIIE and ASA IVE. Quick SOFA score was positive (≥1) in 14 patients (21%). The median SOFA score value was 1 (IQR 0,2). The mean POSSUM predicted morbidity was 49.4% ± 23.9%. The mean physiological and operative scores were 22.7 ± 6.4 and 12.31 ± 2.68 scores for POSSUM, and 22.5 ± 6.6 and 12.3 ± 2.6 for P-POSSUM. The POSSUM and P-POSSUM predicted mortality in median (IQR) was 8.2% (5.2,16.6)% and 2% (1,7)%, respectively.

Surgical treatment and type of anesthesia

The most common emergency surgical procedure performed in COVID-19 patients was sinus surgery for rhino cerebral mucormycosis in 36 patients (54%) followed by obstetric in 20 (30%), gastrointestinal in 6 (9%), orthopedic in 3 (4.5%), and neurosurgery in 2 (3%) patients. Blood product transfusion was needed in 18 patients (27%). General anesthesia with endotracheal intubation was administered in 46 patients (69%). The mean lung compliance noted prior to start of surgery was low (33.2 ± 7 ml/cm H2O). Regional anesthesia was given in 21 patients (31%), mainly in obstetric and orthopedic surgeries.

Postoperative complication

Major complications (>3A Clavien-Dindo class) were observed in 22 patients in the postoperative period. Intensive care was needed after surgery in 21 patients (31.3%), including 8 (22%) ENT, 5 (83.3%) abdominal, 5 (25%) obstetric, 2 (100%) neurosurgery, and 1 (33%) orthopedic surgery. Most frequent complications were ARDS (26.9%), followed by sepsis (25.3%), septic shock (18%), acute kidney injury (AKI) (19.4%), multiorgan dysfunction syndrome (MODS) (12.6%), poor myocardial contractility (5.9%), arrhythmias (4.5%), and thrombotic complications (3%). The severity of complications as per Clavien Dindo classification are mentioned in [Table 1]. Amongst the 22 patients with major postoperative complications, only 9 survived.
Table 1: Postoperative outcomes: morbidity and mortality

Click here to view

ARDS after surgery was observed in 2 (100%) neuro, 5 (83.3%) abdominal, 1 (33.3%) orthopedic, 4 (25%) obstetric, and 6 (16.6%) sinus (ENT) surgeries. Mild, moderate, and severe ARDS was seen in 10, 4, and 4 patients, respectively. Postoperatively, invasive and non-invasive ventilation was needed in 15 (22%) and 1 (1.5%) patients, respectively. The critical care and duration of hospital stay in survivors ranged from 1–13 and 8–44 days, respectively. No significant difference was seen in the postoperative complications between those who tested COVID-19 positive before or after surgery (P = 0.6)

Postoperative 30-day mortality

The 30-day mortality rate was 19% with death in 13 patients. Mortality was highest in laparotomies (83%) followed by neurosurgical (50%), orthopedic (33.3%), sinus (rhino-cerebral mucormycosis) surgeries (13.9%) and lower segment cesarean sections (5%) [Figure 1].
Figure 1: Kaplan–Meir curve showing mortality among patients who underwent emergency surgery procedures. red – GIT/general surgery, blue – neurosurgery and orthopedic surgery, green – obstetrics, purple – ENT. Dip in every curve corresponds to 1 death that had occurred and + indicates censored data

Click here to view

Preoperative ARDS suggestive of COVID-19 was seen in 10 non-survivors (mild in 5, moderate in 4 and severe in 1). In the postoperative period, ARDS was noted in 12 non-survivors (mild, moderate, and severe in 5, 4, and 3 patients, respectively). New onset of mild and severe ARDS was noted in one patient each of mucormycosis and pancreatitis, along with increase in severity in 8 patients. Postoperative invasive ventilation was needed in 10/13 patients (77%) among non-survivors in contrast to only 5/53 (9.3%) of the survivors.

The immediate cause of postoperative death was identified as septic shock with COVID-19 ARDS respiratory failure in 9 patients (69.5%), respiratory failure with COVID-19 ARDS in 2 patients (15.3%), post-COVID-19 mucormycosis with MODS in 1 patient, and septic shock with MODS in 1 patient.

After surgery, moderate-severe ARDS was seen in 12/13 non-survivors (92.3%) compared to 1/54 survivors (1.8%). Amongst those administered general anesthesia, the non-survivors were observed to have a statistically significant higher driving pressure (21.4 ± 3.9 vs 18 ± 3.4 mmHg, P = 0.005) and lower compliance (29.8 ± 5.6 vs 34.4 ± 7.8, P = 0.049) as compared to survivors.

No statistically significant difference was seen in the 30-day mortality in those who tested COVID-19 positive before or after surgery (P > 0.9).

Observed morbidity and accuracy of prognostic scores

For correlation of POSSUM predicted morbidity with observed morbidity as per different classes of Clavien-Dindo scale, a vertical box plot was plotted [Figure 2]. The POSSUM predicted morbidity risk was found to be significantly associated with the observed morbidity (P = 0.001).
Figure 2: Box plot of Clavien-Dindo scale and POSSUM predicted morbidity score. The X-axis shows the different classes of Clavien-Dindo scale and Y-axis represents the POSSUM predicted morbidity score. The black dots represent the outliers, and the vertical line represents the minimum and the maximum value of the score for each class. The borders of the box represent the 25th and the 75th centile value of the POSSUM predicted morbidity score. Here the class 4A has only a single data; hence it is underrepresented as compared to the other classes

Click here to view

Observed mortality and accuracy of prognostic scores

ASA-PS IIIE or IVE had 12.7 times higher risk of mortality compared to ASA-PS IE and IIE (P < 0.001). For every unit increase of SOFA, POSSUM, and P-POSSUM score, there was a 50%, 18%, and 17% higher risk of mortality. A positive qSOFA was associated with 9.03 times higher risk of mortality as compared to negative (P < 0.001) [Table 2].
Table 2: Predictors of mortality with respect to prognostic scores among COVID-19 positive patients who underwent emergency surgery

Click here to view

On analysis of ROC, the AUC for the POSSUM, P-POSSUM, and SOFA was 0.85, 0.86 and 0.8829, respectively [Figure 3]. The Youden's index (YI) value was 0.6, 0.6, and 0.89 for POSSUM, P-POSSUM, and SOFA score, respectively [Figure 4]a,[Figure 4]b,[Figure 4]c.
Figure 3: Receiver operating curve of POSSUM, P-POSSUM, and SOFA models combined in ROCR package. The area under the curve (AUC) for the POSSUM, P-POSSUM, and SOFA: Receiver operator curve was 0.85 with 95% CI (0.70–0.93) shown in “red”, for P-POSSUM was 0.86 with 95% CI (0.71–0.94) as “blue”, and for SOFA 0.883 with 95% CI (0.747–0.959) as “green”

Click here to view
Figure 4: (a) Youden's cutoff point of the POSSUM ROC Curve. The Youden's index (YI) value was 0.6 with a cutoff threshold of 29.55 for POSSUM with sensitivity of 61.5% and specificity of 98.1%. (b) Youden's cutoff point of the P-POSSUM ROC Curve. The Youden's index (YI) value was 0.6 for P-POSSUM model with a cutoff threshold of 16.22 for P-POSSUM with sensitivity of 61.5% and specificity of 98.1%. (c) Youden's cut off point of the SOFA ROC Curve. The Youden's index (YI) value was 0.89 for SOFA score with corresponding cut off threshold value of 1.5, with sensitivity of 84.6% and specificity of 75.9%

Click here to view

The Hosmer-Lemeshow (HL) goodness of fit test revealed both POSSUM and P-POSSUM models to be a good fit for predicting mortality at 30-days, with small χ2 HL values (POSSUM: χ2 HL statistic = 3.94; df = 8; P = 0.26 and P-POSSUM: χ2 HL statistic = 3.26; df = 8; P = 0.19) [Table 3]. Smaller χ2 HL values suggest good fit, with the data considered to be fitting the model if P > 0.05. The observed mortality was higher than that predicted for the first three of the 20% risk bands, that is, up to 41%–60% risk band.
Table 3: The Hosmer-Lemeshow (HL) goodness of fit test for predicting mortality at 30 days, with small χ2 HL values

Click here to view

 :: Discussion Top

The COVID-19 pandemic was associated with a significant health care burden in India, with more than 34.5 million infected cases reported till date.[15] We observed a postoperative 30-day mortality rate of 19% in COVID-19 emergency surgical patients. ASA-PS IIIE/IVE and a positive qSOFA were significantly associated with mortality. Further, SOFA, P-POSSUM, and POSSUM scores had a good discriminant value to differentiate between survivors and non-survivors in our COVID-19 emergency surgical cohort.

Carrier et al. reported 50% of symptomatic patients with 18% in need of mechanical ventilation and a 30-day mortality rate of 15.9% in 44 COVID-19 emergency surgical patients.[16] The COVIDSurg Collaborative observed a higher mortality of 26% after emergency surgery.[1] In our patient cohort, the 30-day mortality rate was 19% with higher mortality in general/gastrointestinal surgery (83%) followed by neurosurgical (50%), orthopedic (33.3%), sinus (rhino-cerebral mucormycosis) surgeries (13.9%), and lower segment cesarean sections (5%). However, comparisons between different studies are difficult because of different individual patient-related demographics, clinical presentation and pathologies, along with variation in surgical approaches, resources, and overall perioperative care.

The clinical presentation of our emergency surgical COVID-19 patients was varied with presence of respiratory, digestive, cardiac, and neurological symptoms. SARS-CoV-2 is known to affect multiple organ systems with a vast variety of symptoms.[17] The strong association between COVID-19-induced severe inflammatory response and ACE-2 receptors expressed in nasal mucosa, bronchus, lung, heart, esophagus, stomach, ileum, kidney, and bladder is reported in literature.[18] However, it is difficult to establish causality with COVID-19 of any symptom other than respiratory.

ARDS can result from direct lung damage as in pneumonia, aspiration, or indirect causes like sepsis and pancreatitis.[19] We observed the presence of preoperative ARDS in 17 patients. ARDS was likely due to COVID-19 in the obstetric, mucormycosis, and toe gangrene patients. However, the ARDS present in the abdominal pathologies may be attributed to either COVID-19 or bacteremic sepsis, which in abdominal infection is known to be associated with ARDS.[20] In patients with abdominal pathologies, common inflammatory pathways may cause endothelial damage, inhibit alveolar immunological responses, reduce lung compliance with onset of hypoxemia and ARDS.[20]

In our cohort, 94% of the patients tested positive for COVID-19 before surgery. No difference in morbidity or mortality was noted in patients who had tested COVID-19 positive prior to or after surgery. These patients who tested positive within a week after surgery were likely already in the incubation period with ongoing pathophysiological changes, and hence, behaved similarly to patients who had tested positive preoperatively. Patients testing positive for COVID-19 could be susceptible to poor perioperative outcomes due to synergistic immunological dysregulation, a hyperinflammatory response to surgery, and need for mechanical ventilation.[1],[21]

Most postoperative complications developed in patients with preoperative ARDS. Postoperative mortality was often secondary to septic shock, ARDS and multiorgan dysfunction in our patients, which could be multi-factorial due to COVID-19 as well as sepsis. Though sepsis may be present as a post-surgical complication, mortality due to postoperative ARDS is rare in only abdominal pathologies, making contribution of COVID-19 likely.[20]

The intensive care beds were overwhelmed and few during the COVID-19 pandemic. Thus, ICU discharge was expedited in relatively stable patients to care for sicker patients, though the patients remained hospitalized for duration ranging from 8 to 44 days.

Previous studies on COVID-19 surgical patients did not study the performance of prognostic risk indices in prediction of postoperative mortality.[2] The COVIDSurg study observed that ASA-PS grade III–V was an independent risk factor for mortality.[1] ASA grade IIIE and IVE were seen in 77% of non-survivors with hazard ratio of 12.7 for mortality as compared to ASA IE and IIE in our study.

SOFA score had a good discriminant function between non-survivors and survivors (AUC 0.89), with 1.5 times higher odds of mortality with every unit increase. Since patients with COVID-19 have an ongoing cytokine storm and systemic inflammatory response, the SOFA score which evaluates six major organ functions was observed to have a better discriminant function.[9] The quick SOFA score (qSOFA) needs just three parameters: respiratory rate (RR) of ≥22 breaths per minute, altered mentation (Glasgow Coma Scale (GCS) <15), and systolic blood pressure (SBP) of <100 mmHg for rapid identification of sepsis.[22] The COVIDSurg study reported that qSOFA score was ≥1 in 62.8% of non-survivors at 30 days after surgery.[1] Increased risk of postoperative ARDS and sepsis was observed after vascular surgery in COVID-19 patients with a positive qSOFA score.[23] We observed that a positive quick SOFA score had a 9 times higher risk of mortality as compared to negative.

Both POSSUM and P-POSSUM scores were significantly higher in non-survivors compared to survivors, with good discriminative ability (AUC 0.85 and 0.86, respectively). Further, we found that POSSUM underpredicted postoperative mortality in COVID-19 patients in the lower pentiles of risk. Previous studies in pre-COVID surgical patients concentrating on different specialties have also observed significant variations from the predicted models at lower bands of predicted risk.[24],[25]

Strengths of our study are that it is the first to assess the performance of prognostic risk indices for assessment of postoperative 30-day mortality in COVID-19 patients undergoing emergency surgery from the Indian subcontinent. Second, we included only the COVID-19 positive cohort reported to have higher risk of mortality as per previous literature. Third, we were able to complete outcome analysis in all patients and draw a perspective regarding the goodness of fit scores like POSSUM and P-POSSUM in prediction of 30-day mortality in COVID-19 emergency surgical patients. Fourth, we included objective scores like SOFA, POSSUM, and P-POSSUM, calculated using easily available online calculators with prospective data collection, thereby decreasing any error in recording of the data sets.

Our study has a few limitations. First, the small sample size stresses upon severe restriction in surgeries, with only the emergent ones done in this pandemic. Second, the comparison with a cohort of non-COVID-19 patients was not done in this study, which limits the understanding of the actual risk attributable to the presence of COVID-19. Third, the clinical presentation, severity of COVID-19, and surgical and anesthetic characteristics of each patient were different. This study may give a risk estimate regarding mortality but ultimately, prognostication needs to be individualized in each patient depending on multiple clinical and surgical factors as well as subtle differences in presentation of similar comorbidities. Fourth, this is a single-center study. However, the data analyzed spans a period of 1.5 years and we were able to find a good discriminant activity of POSSUM, P-POSSUM, qSOFA, and SOFA in predicting mortality. Fifth, the possibility of false positive COVID-19 RT-PCR tests cannot be ruled out.

In conclusion, though only few COVID-19 emergency surgical patients were included in our study, SOFA score was observed to have a better risk prediction potential, with good AUC and a higher sensitivity as compared to P-POSSUM and POSSUM. A positive qSOFA and ASA IIIE and IVE were also significantly associated with increased mortality. However, further multicentric studies with inclusion of higher number of patients and a control group of non-COVID-19 patients are needed for better validation of these prognostic scores in COVID-19.


We acknowledge the tireless work of all faculty, resident doctors, nursing staff and other health care professionals in treating COVID-19 patients throughout the pandemic.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

 :: References Top

COVIDSurg Collaborative. Mortality and pulmonary complications in patients undergoing surgery with perioperative SARS-CoV-2 infection: An international cohort study [published correction appears in Lancet. 2020 Jun 9]. Lancet 2020;396:27-38.  Back to cited text no. 1
COVIDSurg Collaborative; GlobalSurg Collaborative. Timing of surgery following SARS-CoV-2 infection: An international prospective cohort study. Anaesthesia 2021;76:748-58.  Back to cited text no. 2
Madrazo Z, Osorio J, Otero A, Biondo S, Videla S. COVID-CIR Collaborative Group. Postoperative complications and mortality following emergency digestive surgery during the COVID-19 pandemic: A multicenter collaborative retrospective cohort study protocol (COVID-CIR). Medicine (Baltimore) 2021;100:e24409.  Back to cited text no. 3
Mayhew D, Mendonca V, Murthy BV. A review of ASA physical status-historical perspectives and modern developments. Anaesthesia 2019;74:373-9.  Back to cited text no. 4
Hall JC, Hall JL. ASA status and age predict adverse events after abdominal surgery. J Qual Clin Pract 1996;16:103-8.  Back to cited text no. 5
Copeland GP, Jones D, Walters M. POSSUM: A scoring system for surgical audit. Br J Surg 1991;78:355-60.  Back to cited text no. 6
Whiteley MS, Prytherch DR, Higgins B, Weaver PC, Prout WG. An evaluation of the POSSUM surgical scoring system. Br J Surg 1996;83:812-5.  Back to cited text no. 7
Prytherch DR, Whiteley MS, Higgins B, Weaver PC, Prout WG, Powell SJ. POSSUM and portsmouth POSSUM for predicting mortality. Physiological and operative severity score for the enUmeration of mortality and morbidity. Br J Surg 1998;85:1217-20.  Back to cited text no. 8
Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996;22:707-10.  Back to cited text no. 9
Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801-10.  Back to cited text no. 10
ARDS Definition Task Force, Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, et al. Acute respiratory distress syndrome: The Berlin Definition. JAMA 2012;307:2526-33.  Back to cited text no. 11
Online calculator for calculating POSSUM and P-POSSUM scores and risk of morbidity and mortality. Available from:  Back to cited text no. 12
Dindo D, Demartines N, Clavien PA. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205-13.  Back to cited text no. 13
R: A language and environment for statistical computing. Available from: [Last accessed on 2021 Oct 09].  Back to cited text no. 14
Worldometer. Coronavirus cases. Available from: [Last accessed on 2021 Dec 04].  Back to cited text no. 15
Carrier FM, Amzallag É, Lecluyse V, Côté G, Couture ÉJ, D'Aragon F, et al. Postoperative outcomes in surgical COVID-19 patients: A multicenter cohort study. BMC Anesthesiol 2021;21:15.  Back to cited text no. 16
Mir T, Almas T, Kaur J, Faisaluddin M, Song D, Ullah W, et al. Coronavirus disease 2019 (COVID-19): Multisystem review of pathophysiology. Ann Med Surg (Lond) 2021;69:102745.  Back to cited text no. 17
Jin Y, Yang H, Ji W, Wu W, Chen S, Zhang W, et al. Virology, epidemiology, pathogenesis, and control of COVID-19. Viruses 2020;12:372.  Back to cited text no. 18
Wheeler AP, Bernard GR. Acute lung injury and the acute respiratory distress syndrome: A clinical review. Lancet 2007;369:1553–64.  Back to cited text no. 19
Eggimann P, Harbarth S, Ricou B, Hugonnet S, Ferriere K, Suter P, et al. Acute respiratory distress syndrome after bacteremic sepsis does not increase mortality. Am J Respir Crit Care Med 2003;167:1210-4.  Back to cited text no. 20
Osorio J, Madrazo Z, Videla S, Sainz B, Rodríguez-González A, Campos A, et al. Analysis of outcomes of emergency general and gastrointestinal surgery during the COVID-19 pandemic. Br J Surg 2021;108:1438-47.  Back to cited text no. 21
Seymour CW, Liu VX, Iwashyna TJ, Brunkhorst FM, Rea TD, Scherag A, et al. Assessment of clinical criteria for sepsis: For the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [published correction appears in JAMA. 2016 May 24-31;315 (20):2237]. JAMA 2016;315:762-74.  Back to cited text no. 22
San Norberto EM, De Haro J, Peña R, Riera L, Fernández-Caballero D, Sesma A, et al. Outcomes after vascular surgery procedures in patients with COVID-19 infection: A National Multicenter Cohort Study (COVID-VAS). Ann Vasc Surg 2021;73:86-96.  Back to cited text no. 23
Brooks MJ, Sutton R, Sarin S. Comparison of surgical risk score, POSSUM and p-POSSUM in higher-risk surgical patients. Br J Surg 2005;92:1288-92.  Back to cited text no. 24
Merad F, Baron G, Pasquet B, Hennet H, Kohlmann G, Warlin F, et al. Prospective evaluation of in-hospital mortality with the P-POSSUM scoring system in patients undergoing major digestive surgery. World J Surg 2012;36:2320-7.  Back to cited text no. 25


  [Figure 1], [Figure 2], [Figure 3], [Figure 4]

  [Table 1], [Table 2], [Table 3]


Print this article  Email this article
Online since 12th February '04
2004 - Journal of Postgraduate Medicine
Official Publication of the Staff Society of the Seth GS Medical College and KEM Hospital, Mumbai, India
Published by Wolters Kluwer - Medknow