Canada's Common Drug Review (CDR) is a single process for reviewing the cost effectiveness of new drugs for participating publicly funded drug plans. For each submission, the CDR conducts a systematic review of the clinical evidence, as well as a review of other clinical issues of relevance to its stakeholders, and a critical assessment of an economic evaluation submitted by the manufacturer.

Purpose: To assess the quality of the clinical data used in economic evaluations submitted to the CDR.

Methods: The authors selected economic evaluations submitted to the CDR in 2004. The clinical data used in the economic evaluations were reviewed and compared with CDR systematic review. The limitations identified by CDR reviewers with the clinical data were noted.

Results: Of the 19 submissions reviewed, the data used to support the clinical claims in economic evaluations were from the following sources: single head-to-head RCT (7), multiple RCTs (5), non-RCTs (3), no clinical data but an assumed benefit (1), and assumed clinical equivalence based on a similar class of drug (3). The key limitations with the data used were: poor quality of clinical evidence – lack of control groups, short term studies, unblinded (6), use of un-validated surrogate outcomes (5), poor methodology for meta analyses and indirect comparisons (5), extrapolation of clinical data from different patient populations (1). In 11 of the economic evaluations, the conclusions drawn on the clinical data differed from those ascertained by the CDR systematic review. All of these submissions (11) received negative listing recommendations, as there was no evidence to support that the drug is cost-effective – insufficient clinical evidence to support the price of the drug.

Conclusions: In over half of the submissions, the results of the clinical data used in the manufacturer's economic evaluations do not agree with the CDR systematic review. This may be related to the use of poor quality clinical data in economic evaluations. The use of inappropriate data can render economic evaluations less effective at informing recommendations.