CHALLENGES ON THE PATH TO REIMBURSEMENT OF ONCOLOGY DRUGS: INSIGHTS FROM THE COST-EFFECTIVENESS DECISIONS OF THE PAN-CANADIAN ONCOLOGY DRUG REVIEW

Purpose:

The pan-Canadian Oncology Drug Review (pCODR) provides recommendations on the addition of cancer medications to provincial formularies. Assessments of cost-effectiveness factoring into recommendation development rely on manufacturer-submitted economic evaluations and re-analyses conducted by pCODR. Our objective is to identify positive and negative submission elements from pCODR’s appraisal of submitted economic evidence.  

Method:

Commentary addressing economic evaluations was examined through a rigorous document review of all final recommendations and Economic Guidance Panel reports (n=116) published by pCODR for all submissions (n=58) received since its inception (2010) to December 1, 2015. Qualitative data isolating key points were extracted into a range of thematic categories that developed inductively. From amongst these categories, those most salient in analyzing economic submissions were selected based on an assessment of existing literature: economic model structure, time horizon, extrapolation methods used, and treatment of post-progression survival modeling. Representative methodological features within these categories were revealed through iterative descriptive content analysis and recorded as having been positively or negatively valued as indicators of submission quality. Enumeration of features’ citations within reports determined their relative weight in pCODR deliberations. Ongoing analysis involves a second iteration of the analytic process to validate results, tracing co-occurrence of citations to explore relationships, and assessing correspondence of positively valued features with economic evaluation guidelines followed in pCODR submission development.

Result:

The most frequently cited features described model type and characteristics, approach taken in modeling effects, assumptions and methods used, selection of treatment comparators, data usage, and extent of extrapolation. In order of declining frequency, commonly criticized features of models were: unjustified assumptions; unreliable data; the use of partitioned-survival analysis; excessive reliance on extrapolation of treatment effects, especially post-progression; a lack of appropriate comparators; overestimation of effects; and discernible bias. In order of declining frequency, features commonly lauded or suggested for submission improvement were: the selection of a time horizon appropriate for the targeted population; a thorough exploration of uncertainty; the inclusion of appropriate treatment comparators; flexibility; and alignment with available data. 

Conclusion:

These findings locate priorities and strategies for future committee deliberations and successful submission development. Additionally, they identify areas for improving the alignment of pCODR procedures with standard economic evaluation guidelines. These applications have potential to clear the path to public coverage of efficacious, cost-effective cancer medications.