Background: Hepatic resection has become common in the United States for both primary and secondary hepatic tumors.
Hypothesis: Variation in outcomes after hepatic resection is related to patient characteristics, the indication for operation, and hospital procedural volume.
Design: Observational study using a nationally representative database.
Patients: All patients in the Nationwide Inpatient Sample for 1996 and 1997 with a primary procedure code for hepatic resection (N = 2097).
Main outcome measures: Outcomes included in-hospital mortality and length of stay. Risk-adjusted analyses were performed using hierarchical multivariate models.
Results: Overall mortality for the 2097 patients was 5.8%. The most common indications for hepatic resection were secondary metastases (52%), primary hepatic malignancy (16%), biliary tract malignancy (10%), and benign hepatic tumor (5%). High-volume hospitals had a mortality rate of 3.9% vs 7.6% at low-volume hospitals (P<.001). In the multivariate analysis adjusting for patient case-mix, high-volume hospitals had a 40% lower risk of in-hospital mortality compared with low-volume hospitals (odds ratio, 0.60; 95% confidence interval, 0.39-0.92; P =.02). Other predictors of mortality in the multivariate analysis included age older than 65 years, hepatic lobectomy (vs wedge resection), primary hepatic malignancy (vs metastases), and the severity of underlying liver disease.
Conclusions: Hospital procedural volume is an important predictor of mortality after hepatic resection. Patients who require resection of primary and secondary liver tumors should be offered referral to a high-volume center.