Sponsorship-Related Outcome Selection Bias in Published Economic Studies of Triptans

Abstract

Background. Economic studies funded by the pharmaceutical industry are more likely to report favorable results and recommendations for the sponsor’s product than are studies funded by nonindustry establishments. Purpose. To determine whether clinical outcome data obtained from the same meta-analyses are used differently in various economic studies of oral triptans and whether there is an association between the study sponsorship and the choice of clinical outcome measure. Data Sources. Economic studies of triptans were identified by updating a previously published systematic review. Study Selection. Twelve studies that used the same meta-analyses as the source of clinical outcome data were identified. Data Extraction. Two independent reviewers extracted the essential data from the identified studies. Data Synthesis. In the 12 appraised studies, 9 alternative measures of effectiveness were derived from the same meta-analyses. Eleven studies were industry-related, and in these the selected clinical outcome consistently favored the sponsor’s product. Also the reported results suggested that the sponsor’s product was more cost-effective than the competitors’ products. Limitations. The cost-effectiveness of triptans is dependent on both the definition of clinical effectiveness and the treatment-related costs. Only bias related to the selection of the clinical outcome measure has been taken into account in this review. Conclusions. The results of published economic studies of triptans are conflicting and biased. There is a tendency to select clinical outcome measures that support the sponsor’s product. This leads to concern about the possible poor applicability of these results in decision making.