Potential sponsorship bias in cost-effectiveness analyses of healthcare interventions: A cross-sectional analysis

Objective: To examine the relationship between the funding source of cost-effectiveness analyses of healthcare interventions published in Spain and study conclusions. Design: Descriptive cross-sectional study. Location: Scientific literature databases (until December 2014). Participants (analysis units): Cohort of cost-effectiveness analysis of healthcare interventions published in Spain between 1989-2014 (n = 223) presenting quality-adjusted life years (QALYs) as the outcome measure. Main measurements: The relationship between qualitative conclusions of the studies and the type of funding source were established using Fisher's exact test in contingency tables. Distributions of the incremental cost-effectiveness ratios by source of funding in relation to hypothetical willingness to pay thresholds between (sic)30,000-(sic)50,000 per QALY were explored. Results: A total of 136 (61.0%) studies were funded by industry. The industry-funded studies were less likely to report unfavorable or neutral conclusions than studies non-funded by industry (2.2% vs. 23.0%; P<.0001), largely driven by studies evaluating drugs (0.9% vs. 21.4%; P<.0001). The incremental cost-effectiveness ratios in studies funded by industry were more likely to be below the hypothetical willingness to pay threshold of (sic)30,000 (73.8% vs. 56.3%; P<.0001) and (sic)50,000 (89.4% vs. 68.2%; P<.0001) per QALY. Conclusions: This study reveals a potential sponsorship bias in cost-effectiveness analyses of healthcare interventions. Studies funded by industry could be favoring the efficiency profile of their products. (C) 2016 Elsevier Espana, S.L.U.