Abstract
Introduction:
Hollow viscus perforation is a significant cause of surgical mortality. Various attempts have been made to identify high-risk patients preoperatively and optimize and manage such patients more aggressively. This study aimed to evaluate the predictors of outcome in patients undergoing emergency laparotomy for perforation peritonitis.
Methods:
This retrospective cross-sectional study was conducted on perforation peritonitis cases admitted to the Department of General Surgery, All India Institute of Medical Sciences, Rishikesh, India. The association between preoperative patient variables with postoperative complications, anastomotic leaks, need for intensive care unit (ICU) admission, and 30-day mortality were evaluated.
Results:
Tachycardia at the time of admission (t = 2.443, p = 0.020), hypotension (χ2 = 18.214, p = <0.001), lower haemoglobin (t = -4.134, p = <0.001), higher blood urea nitrogen levels (W = 1967.000, p = 0.012), International Normalised Ratio (INR) ≥ 1.5 (χ2 = 17.340, p = <0.001), the mean albumin level 2.89 ± 0.77 g/dL (t = -2.348, p = 0.027), and delay in surgery (χ2 = 28.423, p = 0.008) were significant associate factors of mortality. The association between need for ICU admission and higher pulse rate on admission (W = 2782.500, p = 0.011), lower systolic blood pressure (W = 1627.500, p = 0.029), higher blood urea nitrogen (W = 2299.000, p = 0.030) and serum creatinine levels (W = 2192.500, p = 0.045), preoperative coagulopathy (χ2 = 6.773, p = 0.017), hypoalbuminemia (t = -2.515, p = 0.016), and delay in surgery (χ2 = 17.780, p = 0.016) was significant.
Conclusion:
Based on the results of this study, hypotension, azotaemia, coagulopathy, and delay in surgery, increase the risk of postoperative mortality of patients undergoing emergency laparotomy for perforation peritonitis. Tachycardia, hypotension, azotaemia, hypoalbuminemia, and pre-operative coagulopathy were good predictors of need for ICU admission. Shock at presentation, deranged renal function and coagulopathy were associated with an increased risk of postoperative complications.
Key Words: Emergencies, intestinal perforation, mortality, peritonitis
1. Introduction:
Gastrointestinal tract perforation is one of the most common surgical emergencies worldwide. Peritonitis and the resultant sepsis and systemic complications due to the perforation are still responsible for significant mortality despite the advent of newer antibiotics, safer operative and anaesthetic techniques, and an improved understanding of pre- and postoperative management (1). Rapid source control through surgical exploration and prudent antimicrobial therapy is fundamental for treating intra-abdominal sepsis due to perforation (2).
Billing et al. proposed early prognostic assessment of patients with perforation peritonitis to allow triaging of patients for a more aggressive therapeutic approach (3). Several scoring systems have since been developed to enable general and prognostic evaluation of patients with perforation peritonitis (2, 4). Bohen et al. did an anatomical classification of intra-abdominal infections into three groups (group I- appendicitis and perforated duodenal ulcer; group II- peritonitis from all other intra-abdominal organs, not following surgery; and group III- postoperative peritonitis) and showed a difference in outcomes between them (4). The Acute Physiology and Chronic Health Evaluation (APACHE) system, on the other hand, is a non-specific physiologic scoring system that has been validated for risk stratification and has also been used in several studies for intra-abdominal infections (5). Meakins and associates proposed an approach for the study and clinical management of intra-abdominal infections that combined functional and anatomical components (6). Singh et al. did a prospective analysis of 84 patients with perforation peritonitis and identified laboratory indices, delay in presentation, and surgery as good predictors of postoperative mortality (7). Most of these scoring systems are exhaustive and challenging to use in emergency departments.
This study aimed to evaluate the predictive factors of postoperative outcome in patients undergoing emergency laparotomy for perforation peritonitis.
2. Methods:
2.1. Study design and settings
A retrospective cross-sectional study was conducted in the Department of General Surgery at the All India Institute of Medical Sciences, Rishikesh, a government-run medical university and tertiary-care hospital in Northern India. The study period was from 01st July 2017 to 01st July 2020. The association between preoperative patient variables with postoperative complications, anastomotic leaks, need for intensive care unit (ICU) admission, and mortality were studied. Approval was obtained from the Institute’s Ethics Committee before the study (AIIMS/IEC/20/741). The transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) statement was adhered to while reporting this study (8).
2.2. Participants
All adult patients admitted to General Surgery department with the diagnosis of peritonitis due to perforation of the gastrointestinal (GI) tract were included on the basis of clinical findings, pneumoperitoneum on chest X-ray, or Abdominal computed tomography (CT). All cases of primary peritonitis, perforations due to corrosive intake, trauma, postoperative peritonitis due to anastomosis leakage, pregnant patients, and patients whose records were not available were excluded.
2.3. Data collection
Patient data were retrospectively collected from the electronic health records (EHR) of the hospital database. Patient details such as demographic information (age, gender, co-morbidities/addictions), symptoms at the time of presentation (pain abdomen, vomiting, fever, ileus), vital signs at the time of presentation (heart rate, blood pressure), and preoperative blood parameters (haemoglobin, total leucocyte count, serum creatinine, blood urea, International Normalised Ratio (INR), serum albumin) were collected. The type of management (operative/non-operative), delay in surgery, and the anatomical site of perforation were also recorded.
2.4. Outcomes
Postoperatively, data regarding complications (using the Clavein-Dindo classification)(9), anastomotic leaks, need for ICU admission, and 30-day postoperative mortality were collected.
2.5. Statistical analysis
The sample size was based on a study by Jhobta et al., who reported 10% mortality in patients with perforation peritonitis (10). It was calculated according to the formula by Lemeshow et al., (11). With a precision (δ) of 0.05 (5%), and type I error (α) at 0.05 (5%), z was taken as 1.96. Based on the above formula, the required sample size was calculated as, N = [1.96² x 0.10 x (1-0.10)] / 0.05² = 138.29 ≈ 139. Thus, with a 95% confidence interval, the minimum sample size required for the study was 139.
Statistical analysis was done using the SPSS statistics package v23 (IBM Corp., USA)(12). We tried to explore the association between the preoperative patient variables with the postoperative outcomes, as mentioned above. Group comparisons for continuously distributed data were made using the independent sample t-test. For non-normally distributed data, an appropriate non-parametric test, such as the Wilcoxon test, was used. Chi-squared test was used for group comparisons of categorical data. In case the expected frequency in the contingency tables was found to be <5 for >25% of the cells, Fisher's exact test was used instead. Linear correlation between the variables was explored using Pearson's and Spearman's correlation for normally and non-normally distributed data, respectively. Statistical significance was kept at p<0.05.
3. Results:
3.1. Baseline characteristics of studied cases
One hundred eighty-three consecutive cases of perforation peritonitis with the mean age of 42.61 ± 15.99 (Range: 18-85) years presenting during our study period were included in the study (80.5% male). 13% of patients had some comorbidity such as diabetes mellitus, hypertension, tuberculosis, etc. Most of the patients also had some form of addiction, with smoking (52.4%), and alcohol intake (20.5%) being common.
3.2. Associated factors of outcomes
Postoperative anastomotic leaks
Participants in the age group 41-50 years had the highest proportion of leaks (χ2 = 16.846, p = 0.026). Among the presenting symptoms, ileus was significantly associated with anastomotic leaks (χ2 = 4.941, p = 0.043). The site of perforation was also associated with postoperative leaks (χ2 = 41.051, p = 0.045), with duodenum, caecum or ascending colon perforations contributing to the majority of leaks. Table 1 shows the associated factors of postoperative anastomotic leak of studied cases.
Table 1.
Associated factors of postoperative anastomotic leak in the studied patients
| Parameters | Anastomotic Leak | P | |
|---|---|---|---|
| Present (n = 7) | Absent (n = 178) | ||
| Age (Years) | |||
| Mean ± SD | 50.14 ± 6.31 | 42.31 ± 16.19 | 0.133 |
| 41-50 Years | 5 (14.7) | 29 (85.3) | 0.026 |
| 51-60 Years | 2 (7.1) | 26 (92.9) | |
| Gender | |||
| Male | 7 (4.7) | 142 (95.3) | 0.349 |
| Female | 0 (0.0) | 36 (100.0) | |
| Co-Morbidities | |||
| Addiction | 6 (4.9) | 116 (95.1) | 0.426 |
| Symptoms | |||
| Pain | 7 (3.8) | 176 (96.2) | 1.000 |
| Vomiting | 3 (3.4) | 85 (96.6) | 1.000 |
| Fever | 4 (8.5) | 43 (91.5) | 0.073 |
| Ileus | 7 (6.4) | 102 (93.6) | 0.043 |
| Duration of symptoms (days) | |||
| Pain | 4.43 ± 2.82 | 7.35 ± 18.24 | 0.591 |
| Vomiting | 4.00 ± 2.00 | 3.66 ± 3.69 | 0.329 |
| Fever | 3.75 ± 2.06 | 9.84 ± 17.45 | 0.589 |
| Ileus | 2.71 ± 1.60 | 2.74 ± 1.85 | 0.924 |
| Vital signs | |||
| Systolic BP (mmHg) | 113.57 ± 21.63 | 109.76 ± 18.31 | 0.635 |
| Pulse Rate (bpm) | 111.57 ± 19.23 | 104.74 ± 17.30 | 0.342 |
| Investigations | |||
| Haemoglobin (g/dL) | 11.38 ± 2.23 | 12.66 ± 3.17 | 0.403 |
| TLC (/cu.mm) | 9663.3 ± 3447.4 | 12408.4 ± 9243.02 | 0.727 |
| Platelet Count (/cu.mm) | 187.67 ± 56.89 | 1471.42 ± 9188.72 | 0.384 |
| Blood Urea (mg/dL) | 80.60 ± 64.35 | 59.33 ± 45.46 | 0.269 |
| Serum Creatinine (mg/dL) | 1.63 ± 1.82 | 1.29 ± 0.90 | 0.604 |
| INR | 1.21 ± 0.11 | 1.41 ± 0.55 | 0.492 |
| Serum Albumin (g/dL) | 3.00 ± 0.43 | 3.29 ± 0.84 | 0.334 |
| Delay in surgery | |||
| Yes | 6 (5.7) | 99 (94.3) | 0.251 |
| Site of perforation | |||
| Gastric (Type I) | 1 (20.0) | 4 (80.0) | 0.045 |
| Gastric (Type III) | 4 (5.9) | 64 (94.1) | |
| Duodenum* | 1 (100.0) | 0 (0.0) | |
| Jejunum | 1 (33.3) | 2 (66.7) | |
| Management | |||
| Operative | 7 (4.7) | 141 (95.3) | 1.000 |
| Non-operative | 0 (0.0) | 18 (100.0) | |
Data are presented as mean ± standard deviation (SD) or frequency (%). BP: Blood Pressure; INR: International Normalised Ratio; TLC: Total Leucocyte Count. *: duodenum, caecum, and ascending colon.
Postoperative complications
Vomiting as a presenting complaint was a predictor of postoperative complications (p = 0.005). The duration of ileus at presentation also predicted delayed complications (p = 0.027). Shock (systolic blood pressure (SBP) < 100 mmHg) on admission correlated with poor prognosis (p = 0.013). Among the blood parameters, raised serum creatinine (p = 0.043) and coagulopathy (INR > 1.5) (p = 0.017) predicted postoperative complications. Table 2 shows the association between different grades of Clavien-Dindo postoperative complications (9) and preoperative variables.
Table 2.
Association between different grades of Clavien-Dindo postoperative complications and preoperative variables
| Parameters | Post-operative complications based on Clavien-Dindo Grade | P | |||||||
|---|---|---|---|---|---|---|---|---|---|
| I (n =1) | II (n = 21) | IIIa (n=15) | IIIb (n=7) | IVa (n=23) | IVb (n = 4) | V (n = 15) | |||
| Age (Years) | |||||||||
| Mean ± SD | 42.00 ± 0 | 35.81 ± 10.97 | 39.93 ± 15.81 | 43.57 ± 14.66 | 44.04 ± 16.83 | 51.75 ± 11.00 | 50.07 ± 18.80 | 0.199 | |
| Gender | |||||||||
| Male | 1 (1.4) | 14 (20.3) | 12 (17.4) | 7 (10.1) | 20 (29.0) | 4 (5.8) | 11 (15.9) | 0.448 | |
| Female | 0 (0.0) | 7 (41.2) | 3 (17.6) | 0 (0.0) | 3 (17.6) | 0 (0.0) | 4 (23.5) | ||
| Co-Morbidity | |||||||||
| Addiction | 1 (1.9) | 14 (25.9) | 8 (14.8) | 2 (3.7) | 15 (27.8) | 2 (3.7) | 12 (22.2) | 0.307 | |
| Symptom | |||||||||
| Pain | 1 (1.2) | 21 (24.7) | 15 (17.6) | 6 (7.1) | 23 (27.1) | 4 (4.7) | 15 (17.6) | 1.000 | |
| Vomiting | 1 (2.6) | 16 (42.1) | 6 (15.8) | 1 (2.6) | 7 (18.4) | 3 (7.9) | 4 (10.5) | 0.005 | |
| Fever | 1 (3.6) | 6 (21.4) | 8 (28.6) | 3 (10.7) | 4 (14.3) | 1 (3.6) | 5 (17.9) | 0.177 | |
| Ileus | 1 (1.8) | 11 (19.3) | 8 (14.0) | 5 (8.8) | 18 (31.6) | 3 (5.3) | 11 (19.3) | 0.422 | |
| Duration of symptoms (Days) | |||||||||
| Pain | 2.00 ± 0 | 3.38 ± 1.75 | 16.40 ± 37.16 | 3.83 ± 3.25 | 4.22 ± 2.68 | 4.50 ± 1.29 | 13.87 ± 30.92 | 0.054 | |
| Vomiting | 2.00 ± 0 | 2.69 ± 1.40 | 5.67 ± 3.50 | 1.00 ± 0 | 4.43 ± 3.87 | 4.33 ± 1.53 | 4.50 ± 3.79 | 0.186 | |
| Fever | 2.00 ± 0 | 7.33 ± 6.25 | 5.62 ± 4.96 | 4.00 ± 1.73 | 4.25 ± 3.86 | 6.00 ± 0 | 10.00 ± 9.90 | 0.950 | |
| Ileus | 2.00 ± 0 | 2.27 ± 1.42 | 4.50 ± 2.98 | 1.20 ± 0.45 | 3.28 ± 1.96 | 3.67 ± 2.08 | 2.36 ± 1.29 | 0.027 | |
| Vital signs | |||||||||
| Pulse Rate (BPM) | |||||||||
| Mean ± SD | 128.0 ± 0 | 107.6± 19.12 | 106.00 ± 11.50 | 115.5 ± 25.7 | 106.3 ± 16.1 | 110.2 ± 16.8 | 112.07 ± 17.1 | 0.758 | |
| Systolic BP (mmHg) | |||||||||
| Mean ± SD | 128.0 ±0 | 109.50 ± 16.4 | 114.20 ± 20.8 | 118.4 ± 11.7 | 114.78 ± 10.43 | 110.5 ± 30.44 | 100.5 ± 23.06 | 0.167 | |
| <100 | 0 (0.0) | 5 (29.4) | 2 (11.8) | 0 (0.0) | 1 (5.9) | 2 (11.8) | 7 (41.2) | 0.013 | |
| ≥100 | 1 (1.5) | 15 (22.1) | 13 (19.1) | 7 (10.3) | 22 (32.4) | 2 (2.9) | 8 (11.8) | ||
| Diastolic BP (mmHg) | |||||||||
| Mean ± SD | 70.00 ± 0 | 75.95 ± 11.27 | 76.27 ± 23.89 | 78.43 ± 8.83 | 69.30 ± 9.45 | 70.00 ± 14.14 | 61.87 ± 13.45 | 0.051 | |
| Investigations | |||||||||
| Haemoglobin (g/dL) | - | 12.64 ± 3.15 | 11.65 ± 2.76 | 11.55 ± 3.66 | 12.36 ± 3.17 | 12.95 ± 2.52 | 9.34 ± 2.83 | 0.059 | |
| TLC (/cu.mm) | - | 12395.1±7774.4 | 18355.6±11001.3 | 8580.0±6869.9 | 10847.4±7975.3 | 7480.6±5304.0 | 13102.5±15258.4 | 0.089 | |
| Platelet Count (/cu.mm) | - | 282.05 ± 155.89 | 263.85 ± 244.21 | 280.29 ± 232.78 | 231.32 ± 105.79 | 202.67 ± 160.75 | 5508.55 ± 17409.69 | 0.597 | |
| Blood Urea (mg/dL) | - | 51.38 ± 27.00 | 61.41 ± 51.76 | 48.57 ± 20.75 | 80.47 ± 62.90 | 170.82 ± 97.71 | 76.00 ± 47.52 | 0.103 | |
| Cr (mg/dL) | |||||||||
| Mean ± SD | - | 1.21 ± 0.65 | 1.10 ± 0.53 | 0.78 ± 0.29 | 1.54 ± 1.37 | 3.09 ± 1.39 | 1.66 ± 1.20 | 0.058 | |
| ≤2 mg/dL | 0 (0.0) | 16 (26.2) | 14 (23.0) | 7 (11.5) | 15 (24.6) | 1 (1.6) | 8 (13.1) | 0.043 | |
| >2 mg/dL | 0 (0.0) | 3 (20.0) | 1 (6.7) | 0 (0.0) | 4 (26.7) | 3 (20.0) | 4 (26.7) | ||
| INR | |||||||||
| Mean ± SD | - | 1.40 ± 0.35 | 1.38 ± 0.20 | 1.23 ± 0.09 | 1.55 ± 0.81 | 1.25 ± 0.22 | 1.80 ± 0.50 | 0.154 | |
| ≤1.5 | 0 (0.0) | 10 (25.0) | 10 (25.0) | 5 (12.5) | 11 (27.5) | 2 (5.0) | 2 (5.0) | 0.017 | |
| >1.5 | 0 (0.0) | 6 (28.6) | 3 (14.3) | 0 (0.0) | 4 (19.0) | 0 (0.0) | 8 (38.1) | ||
| Serum Albumin (g/dL) | - | 3.38 ± 0.89 | 3.18 ± 0.73 | 2.61 ± 0.38 | 2.95 ± 0.83 | 3.10 ± 0.46 | 2.86 ± 0.84 | 0.081 | |
| Imaging | 0.761 | ||||||||
| Delay in Surgery | 1 (1.8) | 15 (27.3) | 9 (16.4) | 3 (5.5) | 14 (25.5) | 3 (5.5) | 10 (18.2) | 0.879 | |
| Site of Perforation | 0.367 | ||||||||
| Management | |||||||||
| Operative | 1 (1.2) | 21 (25.6) | 13 (15.9) | 7 (8.5) | 21 (25.6) | 4 (4.9) | 15 (18.3) | 0.682 | |
| Non-operative | 0 (0.0) | 0 (0.0) | 1 (50.0) | 0 (0.0) | 1 (50.0) | 0 (0.0) | 0 (0.0) | ||
Data are presented as mean ± standard deviation (SD) or frequency (%). BP: Blood Pressure; Cr: Creatinine; INR: International Normalised Ratio; TLC: Total Leucocyte Count; BPM: Beat Per Minute.
ICU admission
Table 3 summarizes the association between need for ICU admission and preoperative parameters. Patients who required ICU admission had a higher pulse rate on admission (W = 2782.500, p = 0.011). The median (interquartile range; IQR) of systolic BP in the ICU admission group was 103 (90-120) mmHg. There was a significant difference in systolic BP (W = 1627.500, p = 0.029) between groups, with the median systolic BP being highest in the group that did not require ICU admission. Subgroup analysis revealed a significant difference between the groups, SBP < 100 and SBP ≥ 100 (χ2 = 12.194, p = <0.001).
Table 3.
Association between need for ICU admission and preoperative parameters
| Parameters | Need for ICU admission | P | |
|---|---|---|---|
| Yes (n = 32) | No (n = 138) | ||
| Age (Years) | |||
| Mean ± SD | 46.59 ± 17.33 | 41.62 ± 15.51 | 0.141 |
| Gender | 0.397 | ||
| Male | 28 (20.0) | 112 (80.0) | |
| Female | 4 (13.3) | 26 (86.7) | |
| Co-Morbidity | |||
| Addiction | 23 (20.4) | 90 (79.6) | 0.472 |
| Symptom | |||
| Pain | 32 (18.9) | 137 (81.1) | 1.000 |
| Vomiting | 12 (15.4) | 66 (84.6) | 0.275 |
| Fever | 9 (20.5) | 35 (79.5) | 0.765 |
| Ileus | 21 (21.0) | 79 (79.0) | 0.409 |
| Duration of symptoms (days) | |||
| Pain | 9.81 ± 24.18 | 6.91 ± 17.08 | 0.575 |
| Vomiting | 2.83 ± 1.40 | 3.70 ± 3.85 | 0.723 |
| Fever | 5.00 ± 6.00 | 10.86 ± 19.03 | 0.317 |
| Ileus | 2.67 ± 1.32 | 2.77 ± 2.00 | 0.693 |
| Pulse Rate (BPM) | |||
| Mean ± SD | 111.41 ± 15.69 | 102.87 ± 16.72 | 0.011 |
| <100 | 6 (9.8) | 55 (90.2) | 0.020 |
| ≥100 | 26 (24.5) | 80 (75.5) | |
| Systolic BP (mmHg) | |||
| Mean ± SD | 105.09 ± 22.42 | 112.01 ± 16.00 | 0.029 |
| <100 | 12 (42.9) | 16 (57.1) | <0.001 |
| ≥100 | 20 (14.4) | 119 (85.6) | |
| Diastolic BP (mmHg) | |||
| Mean ± SD | 64.53 ± 13.75 | 72.67 ± 12.23 | 0.001 |
| Laboratory | |||
| Haemoglobin (g/dL) | 12.26 ± 2.96 | 12.69 ± 3.27 | 0.505 |
| TLC (/cu.mm) | 11794.44 ± 9330.32 | 12709.10 ± 9307.87 | 0.406 |
| Platelet Count (/cu.mm) | 2390.15 ± 11114.30 | 968.27 ± 8028.70 | 0.224 |
| Blood Urea (mg/dL) | 77.87 ± 61.75 | 54.17 ± 40.04 | 0.030 |
| Serum Creatinine (mg/dL) | |||
| Mean ± SD | 1.70 ± 1.19 | 1.18 ± 0.84 | 0.045 |
| ≤2 | 20 (15.2) | 112 (84.8) | 0.032 |
| >2 | 8 (36.4) | 14 (63.6) | |
| INR | |||
| Mean ± SD | 1.69 ± 0.75 | 1.30 ± 0.31 | 0.018 |
| ≤1.5 | 10 (10.9) | 82 (89.1) | 0.017 |
| >1.5 | 9 (31.0) | 20 (69.0) | |
| Serum Albumin (g/dL) | |||
| Mean ± SD | 2.94 ± 0.79 | 3.37 ± 0.84 | 0.016 |
| <2.5 g/dL | 7 (30.4) | 16 (69.6) | 0.149 |
| ≥2.5 g/dL | 20 (16.8) | 99 (83.2) | |
| Delay in Surgery | |||
| Yes | 18 (17.1) | 87 (82.9) | 0.572 |
| Hours | 14.78 ± 12.12 | 12.67 ± 22.15 | 0.191 |
| Reason for the delay | |||
| Unavailability of OT slot | 14 (14.7) | 81 (85.3) | 0.016 |
| Unavailability of ICU/ventilator | 0 (0.0) | 1 (100.0) | |
| Delay in diagnosis | 0 (0.0) | 3 (100.0) | |
| Initial Resuscitation | 3 (100.0) | 0 (0.0) | |
| Left against medical advice | 0 (0.0) | 0 (0.0) | |
| Non-operative management | 1 (50.0) | 1 (50.0) | |
| Delay in CT scan | 0 (0.0) | 1 (100.0) | |
| Impending Perforation | 0 (0.0) | 1 (100.0) | |
| Management | |||
| Operative | 29 (19.6) | 119 (80.4) | 0.531 |
| Non-operative | 2 (11.1) | 16 (88.9) | |
Data are presented as mean ± standard deviation (SD) or frequency (%). BP: Blood Pressure; INR: International Normalised Ratio; TLC: Total Leukocyte Count; ICU: Intensive Care Unit; CT: Computed Tomography; OT: Operation Theatre; BPM: Beat Per Minute.
Deranged renal function was significantly associated with ICU admission, with both blood urea (W = 2299.000, p = 0.030) and serum creatinine levels significantly elevated (W = 2192.500, p = 0.045). Preoperative coagulopathy also predicted ICU admission (χ2 = 6.773, p = 0.017). Hypoalbuminemia was also a strong predictor of ICU admission (t = -2.515, p = 0.016), with 2.17 times higher chance of admission in those with albumin <2.5g/dL (95% CI= 0.79-5.94). The reason for delay in surgery was also a significant predictor of ICU admission (χ2 = 17.780, p = 0.016).
Postoperative mortality
Table 4 summarizes the association between 30-day mortality and preoperative parameters. Tachycardia at the time of admission was associated with higher postoperative mortality (t = 2.443, p = 0.020). However, on subgroup analysis, no difference was observed between the groups, pulse rate (PR) < 100 and PR ≥ 100 (χ2 = 3.722, p = 0.054). Hypotension was also associated with increased postoperative mortality (χ2 = 18.214, p = <0.001) with 4.55 times higher risk of mortality in the group with systolic BP ≤ 100 mmHg (95% CI = 2.19 - 9.22). Diastolic BP was also significantly associated with postoperative mortality (W = 988.500, p = <0.001). The haemoglobin (Hb) in the postoperative mortality group was significantly lower (t = -4.134, p = <0.001). However, no difference in the group Hb ≤ 8 g/dL and Hb > 8g/dL (χ2 = 4.925, p = 0.061) was evident on subgroup analysis. On the other hand, blood urea levels influenced postoperative mortality (W = 1967.000, p = 0.012). INR ≥ 1.5 was also associated with higher mortality (χ2 = 17.340, p = <0.001). The mean (standard deviation; SD) of albumin level in mortality group was 2.89 (0.77) g/dL and was significantly associated with postoperative mortality (t = -2.348, p = 0.027). No difference was evident on subgroup analysis between the groups, albumin ≤ 2.5 g/dL and > 2.5 g/dL (χ2 = 3.685, p = 0.089).
Table 4.
Association between postoperative 30-day mortality and preoperative parameters
| Parameters | Mortality | P | |
|---|---|---|---|
| Present (n = 23) | Absent (n=162) | ||
| Age (Years) | |||
| Mean ± SD | 49.00 ± 18.51 | 41.70 ± 15.45 | 0.072 |
| Gender | |||
| Male | 17 (11.4) | 132 (88.6) | 0.403 |
| Female | 6 (16.7) | 30 (83.3) | |
| Co-Morbidity | |||
| Addiction | 14 (11.5) | 108 (88.5) | 0.583 |
| Symptom | |||
| Pain | 23 (12.6) | 160 (87.4) | 1.000 |
| Vomiting | 9 (10.2) | 79 (89.8) | 0.358 |
| Fever | 7 (14.9) | 40 (85.1) | 0.577 |
| Ileus | 15 (13.8) | 94 (86.2) | 0.555 |
| Duration of symptoms (Days) | |||
| Pain | 14.43 ± 29.99 | 6.21 ± 15.29 | 0.534 |
| Vomiting | 3.44 ± 2.65 | 3.70 ± 3.75 | 0.657 |
| Fever | 7.86 ± 8.88 | 9.57 ± 17.87 | 0.844 |
| Ileus | 2.40 ± 1.12 | 2.79 ± 1.91 | 0.706 |
| Pulse Rate (BPM) | |||
| Mean ± SD | 112.43 ± 15.33 | 103.92 ± 17.42 | 0.020 |
| <100 | 4 (6.2) | 60 (93.8) | 0.054 |
| ≥100 | 19 (16.2) | 98 (83.8) | |
| Systolic BP (mmHg) | |||
| Mean ± SD | 98.26 ± 22.49 | 111.60 ± 17.15 | 0.002 |
| <100 | 12 (34.3) | 23 (65.7) | <0.001 |
| ≥100 | 11 (7.5) | 135 (92.5) | |
| Diastolic BP (mmHg) | |||
| Mean ± SD | 61.70 ± 12.04 | 72.08 ± 12.96 | <0.001 |
| Laboratory data | |||
| Haemoglobin (g/dL) | |||
| Mean ± SD | 10.17 ± 2.62 | 12.94 ± 3.09 | <0.001 |
| <8 g/dL | 3 (33.3) | 6 (66.7) | 0.061 |
| ≥8 g/dL | 15 (9.6) | 141 (90.4) | |
| TLC (/cu.mm) | 14282.7± 14684.3 | 12135.6 ± 8356.3 | 0.648 |
| Platelet Count (/cu.mm) | 3464.1 ± 13610.6 | 1197.4 ± 8415.4 | 0.165 |
| Blood Urea nitrogen (mg/dL) | 91.61 ± 68.97 | 55.63 ± 40.37 | 0.012 |
| Serum Creatinine (mg/dL) | |||
| Mean ± SD | 1.78 ± 1.22 | 1.24 ± 0.87 | 0.063 |
| ≤2 mg/dL | 14 (10.1) | 125 (89.9) | 0.094 |
| >2 mg/dL | 6 (23.1) | 20 (76.9) | |
| INR | |||
| Mean ± SD | 1.91 ± 0.80 | 1.35 ± 0.47 | 0.002 |
| ≤1.5 | 4 (4.2) | 92 (95.8) | <0.001 |
| >1.5 | 10 (30.3) | 23 (69.7) | |
| Serum Albumin (g/dL) | 2.89 ± 0.77 | 3.34 ± 0.83 | 0.027 |
| Mean ± SD | 0.089 | ||
| <2.5 g/dL | 6 (24.0) | 19 (76.0) | |
| ≥2.5 g/dL | 13 (10.2) | 115 (89.8) | |
| Reason for the delay | |||
| Unavailability of OT slot | 9 (9.5) | 86 (90.5) | 0.008 |
| Unavailability of ICU/ventilator | 0 (0.0) | 12 (100.0) | |
| Delay in diagnosis | 1 (33.3) | 2 (66.7) | |
| Initial Resuscitation | 3 (100.0) | 0 (0.0) | |
| Left against medical advice | 0 (0.0) | 2 (100.0) | |
| Non-operative management | 0 (0.0) | 2 (100.0) | |
| Delay in investigations (CT) | 0 (0.0) | 1 (100.0) | |
| Impending Perforation | 0 (0.0) | 1 (100.0) | |
| Management | |||
| Operative | 20 (13.5) | 128 (86.5) | 1.000 |
| Non-operative | 2 (11.1) | 16 (88.9) | |
Data are presented as mean ± standard deviation (SD) or frequency (%). BP: Blood Pressure; INR: International Normalised Ratio; TLC: Total Leukocyte Count; ICU: Intensive Care Unit; CT: Computed Tomography; OT: Operation Theatre; BPM: Beat Per Minute.
There was a significant difference in mortality due to surgery delay for multiple reasons (χ2 = 28.423, p = 0.008). Delay due to initial resuscitation led to the highest rate of mortality.
4. Discussion:
Based on the results of this study, hypotension, azotaemia, coagulopathy, and delay in surgery increase the risk of postoperative mortality of patients undergoing emergency laparotomy for perforation peritonitis. Tachycardia, hypotension, azotemia, hypoalbuminemia, and preoperative coagulopathy were good predictors of ICU admission. Shock at presentation, deranged renal function and coagulopathy were associated with an increased risk of postoperative complications.
Generalised peritonitis is a common surgical emergency. It is one of the leading causes of death in non-trauma surgical patients, with a mortality as high as 20% (2, 3). Even with the advancement in diagnostic and therapeutic aspects over the years, a significant number of lives are being lost to this illness.
Several modifiable and non-modifiable factors can influence the clinical outcome in patients with perforation peritonitis. Attempts must be made to identify and optimize the high-risk patient preoperatively, while simultaneously preparing for emergency surgery. Multiple studies have tried to identify the factors that can influence the clinical outcome in these patients. Certain factors and lab parameters can be used to predict the outcome, and several scoring systems have been devised using them, such as the Acute Physiology and Chronic Health Evaluation (APACHE) score, the Simplified Acute Physiology Score (SAPS), the Boey Score, the Multi-Organ Failure (MOF) Score, and the Mannheim Peritonitis Index (MPI) (2, 7). These scores are not simple to use and are time-consuming. Preoperative functional status is also being used for predicting the postoperative outcome (13). It is, thus, more relevant to identify simple patient parameters that can predict postoperative complications and mortality.
Anastomotic leak is one of the major complications following bowel repair or anastomosis. The UK Surgical Infection Study Group defined an enteric leak as "leakage of luminal contents from a surgical join between two hollow viscera" (14). Several factors have been linked to anastomotic leaks such as, malnutrition, steroids, tobacco use, leukocytosis, cardiovascular disease, alcohol use, lower GI anastomoses, suboptimal anastomotic blood supply, operative time of more than 2 hours, bowel obstruction, perioperative blood transfusion, and intra-operative septic conditions not conducive to a primary anastomosis (15, 16). We report the highest rate of anastomotic leaks between 41-50 years of age (χ2 = 16.846, p = 0.026). Mcdermott et al. found that the mean age group was 60 years and that age did not correlate with postoperative leakage (17). We found preoperative ileus to be significantly associated with anastomotic leaks (χ2 = 4.941, p = 0.043). Peter et al. observed similar findings in patients undergoing colorectal resection (15). Multiple studies have shown that those with lower GI anastomoses are more prone to leaks than those having anastomoses in the upper GI tract, especially after emergency surgery (16, 18). We found the site of perforation to be associated with anastomotic leaks (χ2 = 41.051, p = 0.045). However, most leaks in our patients occurred following duodenal, caecal, and colonic perforations, in that order. In another study, Gupta et al. observed that the size of the duodenal perforation determines the risk of postoperative leak (19).
We report hypotension at the time of admission as an important predictor of postoperative mortality (χ2 = 18.214, p = 0.001) with 4.55 times higher risk in those with systolic BP ≤ 100 mmHg (95% CI = 2.19 - 9.22). Diastolic BP was also significantly associated with postoperative mortality (W = 988.500, p =0.001). Singh et al. also found that shock could predict poor postoperative outcomes, which is in line with our findings (7). In a study by Wesselink et al., the authors observed that intraoperative mean arterial pressure (MAP) less than 60-65mm Hg was associated with poor surgical outcomes (20).
We conclude that deranged renal function and hypoalbuminemia are important predictors of postoperative complications. This is in concordance with studies conducted previously (3, 21). Presence of coagulopathy (INR >5) was also related to postoperative mortality (t = -2.348, p = 0.027). This could be a result of sepsis-induced disseminated intravascular coagulopathy (DIC). In a single-centre analysis, Nakamura et al. found that preoperative DIC score is a prognostic factor for colonic perforation associated with peritonitis (22). Moreover, patients with deranged kidney function, hypoalbuminemia, and deranged INR were more likely to require ICU admission post-surgery. However, the other important causes of raised creatinine in these patients, such as urinary tract obstruction (stones, neoplasms, prostatic hyperplasia), diabetes, and nephrotoxic drug intake, must also be kept in mind.
It seems that preoperative variables such as tachycardia, hypotension, deranged renal function, coagulopathy, and hypoalbuminemia are strong predictors of poor prognosis in patients with perforation peritonitis. Identifying one or more of these high-risk predictors calls for a more aggressive resuscitation with rapid source control for a favourable patient outcome.
5. Limitations
This study was a retrospective one, and data collection was record-based. A larger prospective study is, thus, required to generate more substantial evidence. This study was conducted at a tertiary-care referral centre, thus receiving the sickest patients from the state and outside. Moreover, there was a significant delay in the presentation of patients due to the arduous Himalayan terrain. All of these could potentially cause a systematic error in favour of the most critical patients, which may not be the case at other centres, and thus, result in an overestimation in our findings. Age, and pre-existing systemic illnesses, are specific confounders that must also be individually matched for to generate more decisive evidence. The role of inflammatory markers such as C-reactive protein and procalcitonin in severity assessment in patients with perforation peritonitis and abdominal sepsis is well known (23, 24). However, due to the high costs of these tests and non-affordability by the majority of our patients, they could not be included in the present study.
6. Conclusion
Based on the results of this study, hypotension, azotaemia, coagulopathy, and delay in surgery increase the risk of postoperative mortality of patients undergoing emergency laparotomy for perforation peritonitis. Tachycardia, hypotension, azotaemia, hypoalbuminemia, and pre-operative coagulopathy were good predictors of ICU admission. Shock at presentation, deranged renal function, and coagulopathy were associated with an increased risk of postoperative complications.
7. Declarations:
7.1. Acknowledgements
We acknowledge the support of faculty and residents of the Department of General Surgery, All India Institute of Medical Sciences, Rishikesh (India).
7.2. Funding and support
Nil.
7.3. Conflict of interest
The authors declare that they have no conflict of interest.
7.4. Authors’ contribution
Dr. Somprakas Basu, Dr. Farhanul Huda, Dr. Praveen Kumar conceptualised the study; Dr. Ankit Rai, Dr. Lena Elizabath David, Dr. Chezhian S did the data collection; Dr. Ankit Rai, and Dr. Lena Elizabath David did the data analysis; Dr. Ankit Rai, Dr. Lena Elizabath David, Dr. Chezhian S prepared the first manuscript; Dr. Somprakas Basu, Dr. Farhanul Huda, Dr. Sudhir Singh, Dr. Praveen Kumar, Dr. Ankit Rai, Dr. Lena Elizabath David, Dr. Chezhian S reviewed the manuscript; Dr. Somprakas Basu, Dr. Farhanul Huda, Dr. Praveen Kumar supervised the study at all stages.
7.5. Data availability
The data used and/or analyzed in the study are available with the corresponding author and can be provided on request.
7.6. Ethical considerations
This study was approved by the Institute Ethics Committee of the All India Institute of Medical Sciences, Rishikesh (AIIMS/IEC/20/741). The committee waived patient consent for this study as medical records were used for data collection and a statement on patient data confidentiality and compliance with the Declaration of Helsinki was provided.
References
- 1.Meshram S, Lal M. Clinico-bacteriological profile of non-traumatic perforation peritonitis cases attending a tertiary care hospital of central India region. Int Surg J. 2018;5(6):2185. [Google Scholar]
- 2.Nyström PO, Bax R, Dellinger EP, Dominioni L, Knaus WA, Meakins JL, et al. Proposed definitions for diagnosis, severity scoring, stratification, and outcome for trials on intraabdominal infection. Joint Working Party of SIS North America and Europe. World J Surg. 1990;14(2):148–58. doi: 10.1007/BF01664867. [DOI] [PubMed] [Google Scholar]
- 3.Billing A, Fröhlich D, Schildberg FW. Prediction of outcome using the Mannheim peritonitis index in 2003 patients. Peritonitis Study Group. Br J Surg. 1994;81(2):209–13. doi: 10.1002/bjs.1800810217. [DOI] [PubMed] [Google Scholar]
- 4.Bohnen J, Boulanger M, Meakins JL, McLean AP. Prognosis in generalized peritonitis Relation to cause and risk factors. Arch Surg. 1983;118(3):285–90. doi: 10.1001/archsurg.1983.01390030017003. [DOI] [PubMed] [Google Scholar]
- 5.Knaus WA, Le Gall JR, Wagner DP, Draper EA, Loirat P, Campos RA, et al. A comparison of intensive care in the U. S.A. and France. Lancet. 1982;2(8299):642–6. doi: 10.1016/s0140-6736(82)92748-9. [DOI] [PubMed] [Google Scholar]
- 6.Meakins JL, Solomkin JS, Allo MD, Dellinger EP, Howard RJ, Simmons RL. A proposed classification of intra-abdominal infections Stratification of etiology and risk for future therapeutic trials. Arch Surg. 1984;119(12):1372–8. doi: 10.1001/archsurg.1984.01390240010002. [DOI] [PubMed] [Google Scholar]
- 7.Singh R, Kumar N, Bhattacharya A, Vajifdar H. Preoperative predictors of mortality in adult patients with perforation peritonitis. Indian J Crit Care Med. 2011;15(3):157–63. doi: 10.4103/0972-5229.84897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Moons KG, Altman DG, Reitsma JB, Ioannidis JP, Macaskill P, Steyerberg EW, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162(1):W1–73. doi: 10.7326/M14-0698. [DOI] [PubMed] [Google Scholar]
- 9.Clavien PA, Strasberg SM. Severity grading of surgical complications. Ann Surg. 2009;250(2):197–8. doi: 10.1097/SLA.0b013e3181b6dcab. [DOI] [PubMed] [Google Scholar]
- 10.Jhobta RS, Attri AK, Kaushik R, Sharma R, Jhobta A. Spectrum of perforation peritonitis in India--review of 504 consecutive cases. World J Emerg Surg. 2006;1:26. doi: 10.1186/1749-7922-1-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lemeshow S W HD, Klar J, Lwanga SK. Adequacy of Sample Size in Health Studies: World Health Organization. 1990. 4 pp. [Google Scholar]
- 12.IBM. IBM SPSS Statistics for Windows. 23 ed. Armonk, New York: IBM Corp: 2015. [Google Scholar]
- 13.Leung JM, Dzankic S. Relative importance of preoperative health status versus intraoperative factors in predicting postoperative adverse outcomes in geriatric surgical patients. J Am Geriatr Soc. 2001;49(8):1080–5. doi: 10.1046/j.1532-5415.2001.49212.x. [DOI] [PubMed] [Google Scholar]
- 14.Langell JT, Mulvihill SJ. Gastrointestinal perforation and the acute abdomen. Med Clin North Am. 2008;92(3):599–625, viii-ix. doi: 10.1016/j.mcna.2007.12.004. [DOI] [PubMed] [Google Scholar]
- 15.Peters EG, Dekkers M, van Leeuwen-Hilbers FW, Daams F, Hulsewé KWE, de Jonge WJ, et al. Relation between postoperative ileus and anastomotic leakage after colorectal resection: a post hoc analysis of a prospective randomized controlled trial. Colorectal Dis. 2017;19(7):667–74. doi: 10.1111/codi.13582. [DOI] [PubMed] [Google Scholar]
- 16.Calin MD, Bălălău C, Popa F, Voiculescu S, Scăunașu RV. Colic anastomotic leakage risk factors. J Med Life. 2013;6(4):420–3. [PMC free article] [PubMed] [Google Scholar]
- 17.McDermott FD, Heeney A, Kelly ME, Steele RJ, Carlson GL, Winter DC. Systematic review of preoperative, intraoperative and postoperative risk factors for colorectal anastomotic leaks. Br J Surg. 2015;102(5):462–79. doi: 10.1002/bjs.9697. [DOI] [PubMed] [Google Scholar]
- 18.Oheneh-Yeboah M. Postoperative complications after surgery for typhoid ileal perforation in adults in Kumasi. West Afr J Med. 2007;26(1):32–6. doi: 10.4314/wajm.v26i1.28300. [DOI] [PubMed] [Google Scholar]
- 19.Gupta S, Kaushik R, Sharma R, Attri A. The management of large perforations of duodenal ulcers. BMC Surgery. 2005;5(1):15. doi: 10.1186/1471-2482-5-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wesselink EM, Kappen TH, Torn HM, Slooter AJC, van Klei WA. Intraoperative hypotension and the risk of postoperative adverse outcomes: a systematic review. Br J Anaesth. 2018;121(4):706–21. doi: 10.1016/j.bja.2018.04.036. [DOI] [PubMed] [Google Scholar]
- 21.Doklestić SK, Bajec DD, Djukić RV, Bumbaširević V, Detanac AD, Detanac SD, et al. Secondary peritonitis - evaluation of 204 cases and literature review. J Med Life. 2014;7(2):132–8. [PMC free article] [PubMed] [Google Scholar]
- 22.Nakamura F, Yui R, Muratsu A, Sakuramoto K, Muroya T, Ikegawa H, et al. Study of the prognostic factor of the colon perforation case with the pan-peritonitis that needed emergency surgery: a single-center observational study. Acute Med Surg. 2019;6(4):379–84. doi: 10.1002/ams2.429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mulari K, Leppäniemi A. Severe secondary peritonitis following gastrointestinal tract perforation. Scand J Surg. 2004;93(3):204–8. doi: 10.1177/145749690409300306. [DOI] [PubMed] [Google Scholar]
- 24.Reith HB, Mittelkötter U, Wagner R, Thiede A. Procalcitonin (PCT) in patients with abdominal sepsis. Intensive Care Med. 2000;26(Suppl 2):S165–9. doi: 10.1007/BF02900731. [DOI] [PubMed] [Google Scholar]