Common Drug Review recommendations: an evidence base for expectations?

backgroundThe Common Drug Review (CDR) was created to provide a single process to review the comparative clinical efficacy and cost effectiveness of new drugs, and then to make formulary listing recommendations to Canadian publicly funded drug benefit plans.

objectiveThe objective was to conduct an in-depth analysis of Canadian Expert Drug Advisory Committee (CEDAC) recommendations to date, to explore predictors and possible explanatory factors associated with negative recommendations.

methodsFinal recommendations were identified from inception (September 2003) to 31 December 2009. Using only publicly available information, recommendations were analysed under the following categories: submission specifics, drug characteristics, clinical factors and economic factors. Descriptive analyses were conducted, followed by statistical analyses, to determine which factors independently predicted a 'do not list' (DNL) recommendation.

resultsThe database consisted of 138 unique final recommendations. The overall DNL rate was 48%. Significant differences in DNL rates were observed between therapeutic areas, ranging from 0% for HIV antivirals up to 88% for analgesic drugs. In the univariate analysis, several factors were significantly associated with a DNL recommendation, including first-in-class drugs and use of clinical scales as an outcome. In the multivariate regression, four factors were significantly predictive of a DNL recommendation: clinical uncertainty (odds ratio [OR] 14), price higher than comparators (OR 9), request for reconsideration (OR 10) and price as the only economic evidence used (OR 18). Incremental cost-effectiveness thresholds were not predictive of recommendations. The hypothesis that economic factors did not impact recommendations when clinical factors were included first was supported by the analysis.

conclusionsThis analysis documented an evidence-driven process that simultaneously weighted multiple factors. Clinical uncertainty and price considerations, but not economic results, had a strong impact on the recommendations. Insufficiency of clinical evidence may have resulted from the gap in evidence available at the time of product launch and the absence of demonstrated benefits to support innovative drugs.