Outcomes of open abdomen versus primary closure following emergent laparotomy for suspected secondary peritonitis: A propensity-matched analysis.

BACKGROUND: Optimal management following index laparotomy is poorly defined in secondary peritonitis patients. Although "open abdomen" (OA), or temporary abdominal closure with planned relaparotomy, is used to reassess bowel viability or severity of contamination, recent studies demonstrate comparable morbidity and mortality with primary abdominal closure (PC). This study evaluates differences between OA and PC following emergent laparotomy.
METHODS: Using the Premier database at a quaternary care center (2012-2016), nontrauma patients with secondary peritonitis requiring emergent laparotomy were identified (N = 534). Propensity matching for PC (n = 331; 62%) or OA (n = 203; 38%) was performed using variables: Mannheim Peritonitis Index, lactate, and vasopressor requirement. One hundred eleven closely matched pairs (PC:OA) were compared.
RESULTS: Five hundred thirty-four patients (55.0% female; mean age, 59.6 ± 15.5 years) underwent emergent laparotomy. Of the OA patients, 136 (67.0%) had one relaparotomy, while 67 (33.0%) underwent multiple reoperations. Compared to daytime cases, laparotomies performed overnight (6 pm-6 am) had more temporary closures with OA (42.8% OA vs. 57.2% PC, p = 0.04). When assessing by surgeon type, PC was performed in 78.7% of laparotomies by surgical subspecialties compared to 56.7% (p < 0.0001) of acute care surgeons. After propensity matching, OA patients had increased postoperative complications (71.2% vs. 41.4%, p < 0.0001), mortality (22.5% vs. 11.7%, p = 0.006), and longer median length of stay (13 vs. 9 days, p = 0.0001).
CONCLUSION: Open abdomen was performed in 38.0% of patients, with one-third of those requiring multiple reoperations. Complications, mortality rates, and costs associated with OA were significantly increased when compared to PC. Given these findings, future studies are needed to determine appropriate indications for OA. LEVEL OF EVIDENCE: Therapeutic/care management, level IV.