Estimating the optimal rate of adjuvant chemotherapy utilization in stage III colon cancer.

Abstract

6591 Background: Identifying optimal chemotherapy utilization rates can drive improvements in quality of care. We report a benchmarking approach to estimate the optimal rate of adjuvant chemotherapy (ACT) for stage III colon cancer. Methods: The Ontario Cancer Registry was linked to electronic chemotherapy records to identify ACT utilization among a random 25% sample of patients with stage III colon cancer diagnosed during 2002-2008 in Ontario, Canada. We explored whether hospital factors (teaching status, regional cancer centre, medical oncologist on-site) were associated with ACT rates. The benchmark population included hospitals with the highest ACT rates that accounted for 10% of the patient population. Hospital ACT rates were adjusted for case mix in a multi-level model accounting for random variation at the hospital level. A Monte Carlo simulation was used to estimate the proportion of observed ACT rate variation that could be due to chance alone. Results: The study population included 2,801patients with stage III colon cancer; ACT was delivered to 66% (1861/2801) of patients. There was no difference in hospital ACT rate by teaching status (64% academic vs 67% non-academic, p = 0.107), comprehensive cancer centre status (65% cancer centre vs 67% non-cancer centre, p = 0.362), or having medical oncology on site (67% on site vs 66% not on site, p = 0.840). After excluding hospitals that had case volumes less than 10 (N = 150), unadjusted ACT rates varied across hospitals (range 44% to 91%, p = 0.017). The unadjusted benchmark ACT rate was 81% (95%CI 76%-86%); utilization rate in non-benchmark hospitals was 65% (95%CI 63%-66%). When using adjusted ACT rates in a multi-level model significant variation remained across hospitals (p < 0.001). The adjusted benchmark ACT rate was 74% (95%CI 63%-83%); non-benchmark hospital ACT rate was 65% (95%CI 53%-75%). The simulation analysis suggested that the non-random component of ACT rate variation across hospitals was 1.5%. Conclusions: There is significant variation in ACT rates across hospitals in routine practice. The estimated benchmark ACT rate is 74%. However, simulation analyses suggest that most of the variation in ACT utilization across hospitals may be due to chance alone.