M. Drescher1,2, V. Chang1,2, R. M. Yau2, B. A. Blanco2, S. Vedachalam2,3, S. Besser2,4, P. Kuo2,3, A. N. Kothari2 1Loyola University Chicago Stritch School Of Medicine,Maywood, IL, USA 2Loyola University Medical Center Department Of Surgery,1:MAP Surgical Analytics Research Group,Maywood, IL, USA 3Midwestern University College Of Osteopathic Medicine,Downers Grove, IL, USA 4DePaul University College Of Computing And Digital Media,Chicago, IL, USA
Introduction:
Laparoscopic cholecystectomy performed at freestanding ambulatory surgery centers (ASC) can be safe and cost-effective with appropriate patient selection. Still, a significant proportion of laparoscopic cholecystectomies are performed in hospitals, not ASCs. The objective of this study was to identify the number of patients who underwent hospital-based laparoscopic cholecystectomy that could have safely undergone surgery in an ASC.
Methods:
The Healthcare Cost and Utilization Project State Inpatient Database (HCUP-SID) and State Ambulatory Surgery and Services Database (HCUP-SASD) were used to identify patients undergoing elective laparoscopic cholecystectomy in California hospitals from 2006-2011 and California ASCs in 2011. Inpatients with a postoperative stay greater than 48 hours were excluded from analysis. Patients who were converted to laparotomy, died, or were readmitted within 30 postoperative days were considered unsuccessful. Preoperative variables including comorbidities, diagnosis, CCI, and age were used to generate uplift predictive models to estimate individual success of undergoing surgery at an ASC compared to a hospital. Uplift modeling measures the quantitative effect (lift) of a treatment on the probability of a successful outcome. Patients with positive lift were predicted to benefit from having surgery in an ASC. Prospective cost savings was estimated using normalized charge data.
Results:
18,096 patients undergoing laparoscopic cholecystectomy in hospitals and 37,477 in ASCs met our inclusion criteria. Hospital patients were older (52.1 vs. 46.8 yrs, p<0.0001) and had higher CCI (0.41, 0.28, p<0.0001) than ASC patients. Readmission rates were higher for hospital than ASC patients (3.5%, 2.7%, p<0.0001), as were rates of conversion (0.07%, 0.02%, p=0.009) and mortality (0.03%, 0.01%, p=0.028). Using an uplift model (lift threshold >0.05%), we predicted 13,988 hospital patients could have successfully underwent surgery at an ASC with a mean lift of 0.008. The mean cost per encounter was $7,857 for these patients, compared to reported mean cost of $6,028 for laparoscopic cholecystectomy at ASCs. Therefore, if the patients who underwent hospital-based laparoscopic cholecystectomy were instead cared for at ASCs, the resultant cost savings would be an estimated $5.1M annually.
Conclusion:
The use of ASCs for performing laparoscopic cholecystectomy is increasing, however patients with favorable preoperative characteristics still undergo surgery in the hospital. Using uplift modeling to identify patients who would be expected to have successful outcomes after surgery at both ASCs and hospitals, we demonstrate that ASCs remain underutilized at significant financial cost.