Predictors of negative intraoperative findings at emergent laparotomy in patients with cirrhosis.

Background: Emergent surgery in the setting of decompensated cirrhosis is highly morbid. We sought to determine the clinical factors associated with negative intraoperative findings at emergent laparotomy.

Methods: We performed a retrospective cohort study of consecutive inpatients with a diagnosis of cirrhosis (ICD-9 571) admitted to the Beth Israel Deaconess Medical Center (Boston, MA) who underwent emergent, nonhepatic, abdominal surgery between May 6, 2005 and September 3, 2012.

Results: Eighty-six patients with cirrhosis were included with a mean model for end-stage liver disease score of 21.3 ± 7.95 and a 90-day mortality rate of 39.5%. Twelve (16.2%) patients had negative laparotomies. Negative intraoperative findings were independently associated with (1) paracentesis prior to a preoperative diagnosis of perforated viscus (P = 0.006), (2) development of an indication for emergent surgery after 24 h into hospital admission for another reason (P = 0.020), and (3) a preoperative diagnosis of bowel ischemia (P = 0.005), with odds ratios of 10.1 (CI 1.92-66.83), 5.80 (CI 1.32-33.39), and 11.1 (CI 2.08-77.4), respectively. Free air on computed tomography (CT) imaging was found in 64.3% (9/14) of patients who had a paracentesis within the preceding 48 h compared to 10.1% (7/72) among patients who did not undergo a paracentesis (P < 0.001). Only 45% of patients with free air following a paracentesis had positive findings at laparotomy compared to 100% in those without a preceding paracentesis (P = 0.038). Negative laparotomy was independently predictive of in-hospital mortality (OR 4.7; P = 0.034).

Conclusion: The possibility of a negative laparotomy is suggested by preoperative clinical factors. In particular, free air following a paracentesis does not necessarily indicate that operative intervention is required. Consideration of close observation before laparotomy in these patients is reasonable.