Interfacility Transfer for Nonelective Cholecystectomy in High MELD Patients: An ACS-NSQIP Analysis

Interfacility transfer to a referral center is often considered for patients with liver disease undergoing nonelective cholecystectomy given management complexities and perioperative risk. We sought to determine the association between the Model for End Stage Liver Disease (MELD) score, transfer frequency, and outcomes in those patients using a national database.

The ACS-NSQIP participant use files were queried for nonelective open or laparoscopic cholecystectomy from 2016 to 2018. Patients were grouped according to low (6-11), intermediate (12-18), or high (>18) MELD. In the high MELD group, patient characteristics and outcomes were compared between transferred and nontransferred patients and multivariate regression was performed to evaluate independent predictors of outcomes. Outcomes included in-hospital mortality, complications, length-of-stay (LOS), and 30-d reoperation and readmission.

30,171 subjects were included. Transfer was more likely as MELD increased (19.5% high versus 12.1% low, P < 0.001). High MELD patients had increased LOS, reoperation, readmission, and mortality rates compared to low MELD. In high MELD patients (n = 1016), those transferred were more likely older, white, obese, and septic. Transferred patients had increased mortality (7.6% versus 4.2%, P = 0.044), LOS, reoperation, and complications. After controlling for differences between transferred and nontransferred patients, transfer status was not independently associated with mortality (OR = 1.593, P = 0.177), postoperative complications or LOS, but was associated with increased risk for reoperation. Sepsis and laparoscopic surgery were independently associated with higher and lower mortality, respectively.

Transfer status is not independently associated with mortality, postoperative complications, or prolonged LOS, suggesting patients with advanced liver disease undergoing acute cholecystectomy may not benefit from interfacility transfer.