Is the approval status considered in Cochrane reviews of drug therapies?

Article type
Authors
Potthast R, Kaiser T, Wieseler B, Vervölgyi V
Abstract
Background: In many countries decisions on reimbursement of drugs are based on systematic reviews. Cochrane reviews (CRs) may be used for this purpose. However, in countries reimbursing drugs only within the license defined by regulatory authorities, CRs used for decision making would need to consider this approval status. Objectives: This study examined whether the EMEA and FDA approval status of drugs is taken into account in CRs. Methods: For active ingredients listed on the EMEA website, excluding orphan drugs, drug combinations, cytostatics and vaccines, the respective CRs were identified via full text search. From the identified CRs we selected those which assessed a single active ingredient that was approved by EMEA and FDA under the same trade name. Two reviewers assessed: (1) whether the approval status was mentioned in the CR; (2) whether the inclusion criteria for studies of the CR met the approval status of EMEA and/or FDA. Discrepancies between the reviewers were solved by discussion. For this study, approval status was defined exclusively by the therapeutic indication of the drug. Results: For 239 drugs listed on the EMEA website, 741 CRs were identified of which 214 evaluated specific drugs or drug groups. Of these, 17 assessed one specific drug approved by EMEA and FDA under the same trade name. Three of these described the relationship between included studies and the approval status. Additional six CRs mentioned the licensed indication without considering the approval status in study selection and evaluation. In only two of the 17 CRs study inclusion criteria corresponded to the approval status of EMEA and FDA. Three of the 17 CRs assessed the respective drug in an off label indication. Only one of these three CRs mentioned this issue. Conclusions: Most CRs insufficiently mentioned the approval status. Moreover, study inclusion criteria of nearly all included CRs did not take the approval status of the assessed drugs into account. Hence, CRs are not per se a basis for healthcare decisions. Additional data, including a description of the impact of studies not fulfilling the approval status on the conclusions of CRs, will be presented.