Prognosis and treatment of peritonitis. Do we need new scoring systems?

OBJECTIVE: To assess the clinical significance of present scoring systems for prognosis and treatment in patients with secondary bacterial peritonitis and to define risk factors for patient survival and outcome not included in the scores. A secondary objective was to review our therapeutic regimens and the need for reoperation with regard to outcome.

DESIGN: Prospective observational study.

SETTING: University hospital, secondary referral center.

PATIENTS: From 1992 to 1995, 92 patients with secondary peritonitis were examined at the University Surgical Clinic, Vienna, Austria. the populations as a whole consisted of 56 men and 36 women with an average age of 56 +/- 19 years. Forty-four percent of patients had postoperative peritonitis.

OUTCOME MEASURES: Mortality, multiple organ system failure (MOSF), relaparotomy.

RESULTS: The mortality rate in patients with an APACHE II (Adult Physiology and Chronic Health Evaluation) score of less than 15 was 4.8%, while mortality rose to 46.7% in those with a score of 15 or higher (P = .001). The average total mortality rate was 18.5%. The prognosis for patients without organ failure or with failure of one organ system was excellent (mortality rate, 0%); quadruple organ failure, however, had a mortality rate of 90%. Initial thrombocytopenia ( < 60 x 10(9)/L), four-quadrant peritonitis, and diabetes mellitus were associated with significantly higher mortality. Leukopenia (white blood cells, < 6 x 10(9)/L) and inappropriate antibiotic therapy as determined by the antibiogram were mildly significant for higher mortality. The need for relaparotomy resulted in substantially higher mortality (P < .001). The impossibility of definitive operative resolution of the intra-abdominal pathologic findings at initial operation had no significant effect on mortality, possibly because planned reoperations were always carried out in those cases. For patients with definitive resolution at initial operation, it was possible to reduce the traditionally high mortality rate associated with relaparotomy on demand by making the decision for reexploration promptly, within the first 48 hours. Nevertheless, the 52.4% mortality rate observed in those cases was still much higher than the 33% found in patients who were not free of disease after the initial operation.

CONCLUSION: The prognosis in peritonitis is decisively influenced by the health status of the patient at the beginning of treatment and by any concomitant risk factors. As a result, a fairly accurate prediction of the outcome of the disease can initially be made on the basis of the APACHE II score and the MOSF score according to Goris. However, the certainty that severely ill patients with high scores often die has little clinical relevance, since it does not provide any therapeutic alternatives to the attending physician. The decision to perform a relaparotomy must be made as soon as possible, at least before MOSF emerges. Already existing MOSF will lead to the "point of no return."