Prediction of mortality and morbidity in patients with secondary peritonitis using POMPP scoring

All patients fulfilling inclusion criteria were included with written informed consent. The parameters under evaluation for each patient under the POMPP scoring were Age [> 65 years], serum albumin [< 1.5gm/dl] and blood urea nitrogen [> 45 mg/dl], with each parameter having a score of 1. Risk of mortality and complications associated with POMPP score were analysed using statistical tests including chi square test, student t test and Fisher exact test. A p value < 0.05 was considered statistically significant.


Introduction
Gastrointestinal perforations are a major cause of mortality & morbidity. Peritonitis due to perforation is one of the commonest surgical emergencies attended by a general surgeon. Perforation peritonitis has a varied clinical profile and several scoring systems have been used to predict prognosis of patients with peritonitis. "Practical scoring system of mortality in patients with perforated peptic ulcer[PPOMPP] is a simple, clinically applicable scoring system that may allow surgeons to rapidly predict mortality rates in patients with perforated peptic ulcers [PPU]" [1].  Table 2. POMPP score assigned to subjects Table 3. POMPP score with Type of Perforation Table 4. POMPP Score with Morbidities Table 5. POMPP score with Mortality in Comparison with Non Peptic and Peptic This study was aimed at using the POMPP scoring in secondary peritonitis patients for rapid assessment, to predict the mortality and morbidity and to validate POMPP scoring.

Materials and methods
The observational study was conducted on patients with suspected perforation peritonitis at the Himalayan Institute of Medical Sciences, Dehradun over a period of 12 months from January 2017 to December 2017. This study was conducted with the ethical approval from the internal ethical committee.
All clinically diagnosed cases of perforation peritonitis with investigatory support and absence of evidence of primary or tertiary aetiology were included in the study with written informed consent.
All patients who fulfilled the inclusion criteria underwent the following Ÿ All the relevant clinical details of history and physical examinations and investigations were recorded as per the pre-set format.
Ÿ Patients were investigated accordingly and POMPP variables were taken and POMPP score was calculated Ÿ Operative procedure and site of perforation was recorded. Ÿ All mortalities occurring among the study subjects, irrespective of cause, were noted. Ÿ The patients were followed up by the principle investigator to assess morbidities at 24 hour intervals in the post-operative period till the patient was either discharged or till mortality during the course of the admission. The specific morbidities were assessed using classifications given by Dindo   There were 8 patients aged over 65 years who had worse prognosis. [3]. In present study, 42.42 +/-16.9 years was observed as mean age of patients which was less when compared to mean ages of patients as noted in previous studies. Increased age [especially > 65 years] has previously been shown to be associated with significantly higher mortality after perforation [3, 4, 5, 7, and 8].
On evaluation of the other parameters of POMPP score, 10.9% patients had albumin less than 1.5 gm /dl, while 16.7 % patients had BUN level more than 45 mg/dl. The above findings noted in our study were similar to a previous study conducted by Kemparaj [9].
A total of 42.9 % of subjects in our study suffered some form of morbidity, of which non-peptic group had 40.7% morbidity and peptic group had 45.6 % morbidity. Emergency surgery for perforated ulcers was noted to be associated with a higher rate of postoperative complications [between 21% and 43%].
In a review of literature available on the different scoring systems of GI perforations, there is around 20-50 % morbidity reported amongst cases of peptic perforations whereas nearly 30-60% morbidity has been noted amongst patients of intestinal i.e. non-peptic perforations [7,10,11,12]. Complications were influenced by association with emergency surgery, inadequate post-operative care, etc. The study showed similar rates of morbidity in among total study subjects as well as in both subsections hence validating the POMPP scoring as per historic evidence.
Various complications were studied according the score allotted. Local complications that included, surgical site infection, wound dehiscence, burst abdomen were the most common comprising 27.5% which corroborates with findings from past studies by Kempraj and Bali [9,15].
The number of cases suffering from systemic and respiratory complications increased with increasing score hence indicating the validity of POMPP score in predicting severity of disease. Cardiovascular and renal complications were also higher at a Score of 1 when compared to 0 thus indicating severity. However no cases were recorded to have a score of 2 or 3. There was a notably increasing trend in morbidity with increased score but there were no noted complications at score 3, as there was a only single case with mortality.
which was one of the parameters. There was an overwhelming preponderance of males in the study with 89.9%. The second parameter of POMPP score assessed patients with serum albumin level less than 1.5 gm/dl, which were found to be 10 Mortality outcomes were calculated separately for peptic and non-peptic groups [ Table 5]. There was 13.6 % risk of mortality in non-peptic group while there was 7.0 % risk in peptic group. An overall of risk of 10.9 % mortality was noted in all gastrointestinal perforations studied which was significant based on Fisher's exact test [p<0.01]. [ Table 6] In our study, there was a total of 10.8% mortality noted amongst subjects which was statistically significantly associated with POMPP score. This observation was fairly comparable to past study by Shin et al had who noted 12.8% mortality in 117 patients who underwent surgery for an intestinal perforation [13]. Additionally, mortality in study by Kemparaj et al [14%] was higher than our overall mortality whereas that noted in Bali et al was lower [7%]. There was increased risk of mortality with increasing POMPP score like Boey & Pulp [16,17]. Adequate preoperative resuscitation [with fluids, etc.], correction of electrolyte imbalances followed by an early surgical intervention, to remove the source of infection and stop further contamination, is responsible for good outcomes minimising morbidity and mortality [15,18] The mean duration of hospital stay in study was 9.57 days with standard deviation of 5.83 days. Longest duration of stay was seen with POMPP score 2 group with mean of 11.67 days. Qureshi et al had similar post-operative stay of 9.6 days [19] and Chalya et al had mean stay of 8.93± 6.63 days [20]. Patients of non-peptic perforations had a longer stay of 10.51 ± 7.05 days. Patients in the non-peptic group also showed strong positive correlation in mean stay with increasing score [p=0.019].
On logistic regression analysis, albumin was found to have a significant correlation with mortality with Odds ratio 0.065 and was an independent factor influencing mortality like in pervious study by Mekense et al [1] while BUN level was not significant in our study with Odds ratio 1.007 and p value 0.58 while it was affecting mortality in study by Mekense et al [1] and Moller et al [17] Age independently did not have significant effect on the prediction of mortality like in other studies [19].
POMPP scoring system involving only two investigations which is albumin and BUN level and one demographic profile that was of age, is simple tool which can satisfactorily predict mortality and morbidity in the both the group for peptic and non-peptic groups. The study was done on a small sample size which may lead to some degree of selection bias. Another possible limitation of the study may be that the sample population was taken from one part of one city, and thus it may not be representative of the population in the region or the country. Since there was one patient with POMPP score 3 which was an important statistical limitation.
Therefore, to further establish the validity of this scoring system and to make this common practice in the assessment and management of GI perforations, a large multicentre study of all geographical areas of the country is required. Another limitation is that a no standard protocol and grading followed for assessment of specific morbidity so need for further morbidity specific studies should be done.

Conclusion
Morbidity and mortality associated with gastrointestinal perforations have been shown to have a strong association with advanced age, blood urea nitrogen and serum albumin levels. Analysis of these simple clinical and biochemical parameters in the study has helped in establishment of the POMPP scoring system for risk stratification of cases of secondary peritonitis. It has been observed to be a simple, clinically applicable scoring system that may allow surgeons to rapidly predict morbidity and mortality rates in all types of gastrointestinal perforations.
Our institute along with others which are similarly placed high volume tertiary care centres, as well as smaller individual practices could benefit with risk stratification and pre-operative assessment of morbidity and mortality using a simplified scoring system like POMPP score for management of patients with secondary peritonitis.