Clinical Review Abstract
Trials in Progress Abstract
Abstracts selected for publication but not presentation at the Annual Meeting
Abstracts granted an exception in accordance with ASCO's Conflict of Interest Policy
Validation of observed savings from an oncology clinical pathways program.
Health Services Research
2013 ASCO Annual Meeting
J Clin Oncol 31, 2013 (suppl; abstr 6553)
Author(s): Jeffrey A. Scott, Scott Milligan, Winston Wong, Daniel Winn, Joseph Cooper, Neil Schneider, Sheamus Parkes, Bruce A. Feinberg; Cardinal Health Specialty Solutions, Dublin, OH; CareFirst BlueCross BlueShield, Baltimore, MD; CareFirst BlueCross Blue Shield, Baltimore, MD; Milliman, Inc, Indianapolis, IN
Background: Oncology clinical pathways have been suggested as a way to decrease cancer treatment variation and costs. CareFirst BlueCross BlueShield (CFBCBS) partnered with Cardinal Health Specialty Solutions to launch the first cancer clinical pathway in the US in Aug 2008. Savings from that program were reported by Scott et al, ASCO 2010. The purpose of this study was to obtain third-party validation of the observed savings of this pathways program. Methods: We used CFBCBS claims data from Jan 2007 to Dec 2010 to identify patients (pts) with breast, colon, or lung cancer who were treated by physicians participating in the pathways program. We used Truven Health’s MarketScan database to retrospectively identify a control group treated by non-institutional physicians in a similar geographic region outside the CFBCBS network. We further balanced the groups using propensity score weighting to align primary diagnosis and demographics. The primary outcome was the sum of allowed cancer costs for 270 days after a patient’s first chemotherapy treatment. A secondary outcome was the probability of an inpatient (inpt) admission over the same time period. Many generalized linear models were fit for sensitivity testing. Boosted decision tree models were also used to fully capture all nonlinearities and interactions. Both types of models use the propensity score weights. All savings estimates were based on comparing trends between cohorts. Results: A total of 2424 CFBCBS pts were included in the analysis. The aligned control group consisted of 1490 pts. The treatment coefficient from the linear model for the primary outcome was -0.16 with a z-value of -3, which translates to a savings estimate of 15% for the program. The treatment coefficient from the logistic model for the secondary outcome of inpt admission reduction was -0.29 with a z-value of -2.5, which translated to a 7% reduction (from 50% to 43%) in hospital admissions. The boosted decision tree models confirmed results of a more moderate magnitude. Conclusions: We conclude that the CFBCBS pathways program saved upwards of 15% on cancer-related claims costs with a 7% reduction in the probability of an inpt admission. These findings are consistent with those previously presented and peer reviewed.
Other Abstracts in this Sub-Category: