Cost-effectiveness of open versus laparoscopic appendectomy: a multilevel approach with propensity score matching

To compare postoperative complications and cost of treatment of laparoscopic (LA) versus open appendectomy (OA) and to identify the most cost-effective treatment method. Patients treated for appendectomy in US veterans health administration (VHA) hospitals in 2005 were included into our study. Direct medical cost and postoperative complications during hospitalization were used as outcomes. Propensity score matching was employed to adjust for baseline imbalances between treatment groups. It was adjusted for the severity of appendicitis, comorbidities according to Charlson Comorbidity Index, and demographic variables. 1:1 optimal matching with replacement was performed. Based on the matched samples, we estimated generalized linear mixed regression models for costs (gamma model) and postoperative complications (logit model). Besides patients' covariates, predictors of hospital resource use and quality of care at the hospital level were considered as explanatory variables. The total study population comprised of 1,128 patients (370 LA, 758 OA) from 95 VHA hospitals. Type of appendectomy had a significant influence on total costs (P = 0.005), with predicted costs for LA being 17.1% lower in comparison to OA (OA: 10,851 US$ [95%CI: 9,707 US$; 12,131 US$] vs. LA: 8,995 US$ [95%CI: 8,073 US$; 10,022 US$]). Differences in the predicted overall postoperative complication were not significant between LA and OA (P = 0.6311). Severity of appendicitis had a significant impact on costs and postoperative complications. Predicted costs for LA were 1,856 US$ lower than for OA while the postoperative complication rate did not differ significantly. Thus, LA is the treatment of choice from a provider's perspective.