OBJECTIVES :

The discrepancy between effects of a health technology in routine clinical practice and effects demonstrated in randomized controlled clinical trials is a source of structural uncertainty affecting the value of health technology. Our objective was to conduct a systematic literature review (SLR) on economic evaluations of anticancer drugs based on claims databases to identify methodological issues to address for the improvement of economic evaluation of anticancer drugs.

METHODS :

We performed a SLR on PUBMED and WEB of SCIENCE from January 2008 to January 2019. The search strategy focused on cost-effectiveness/utility analyses (CEAs or CUAs) of anticancer drugs entirely based on claims databases. Selected articles were independently analysed. The quality of economic evaluations was assessed using the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) and risk of bias assessment checklists.

RESULTS :

Of the 306 identified records on PUBMED, 12 articles were included in our review and 1 additional was identified through WEB of SCIENCE. Ten of 13 were CEA, 3 CUA. Most of them were carried out in North America: USA (n = 9), Canada (n = 2). The used economic metric was the cost per life-year gained (n = 10) or cost per QALY (n = 3). All the studies included direct costs. Time horizons varied from 1 year to lifetime. Reporting of the target analysis population and compared strategies were totally in agreement with CHEERS. Structural assumptions underpinning the economic model was the poorest reported item. Representativeness bias (n = 11) and censored medical costs issue (n = 8) were the most frequent analysed risks.

CONCLUSIONS :

Our SLR identified few CEA/CUA but this should rise with the increasing widespread access to claims databases. Our findings highlighted the overall potential and the specific issues of this type of data in economic evaluation of anticancer drugs.