What influences recommendations issued by the agency for health technology assessment in Poland? A glimpse into decision makers' preferences

Abstract

Objective: This study aimed to evaluate the factors that are associated with positive (supporting public funding) and negative recommendations of the Agency for Health Technology Assessment in Poland. Methods: Two independent analysts reviewed all the recommendations publicly available online before October 7, 2011. For each recommendation, predefined decision rationales, that is, clinical efficacy, safety, cost-effectiveness, and formal aspects, were sought, either advocating or discouraging the public financing. In the analysis, we used descriptive statistics and a logistic regression model so as to identify the association between predefined criteria and the recommendation being positive. Results: We identified 344 recommendations-218 positive (62.8%) and 126 negative (37.2%). Negative recommendations were better justified and also the comments were less ambiguous in accordance with the recommendation (except for clinical efficacy). In general, the specified criteria supported the decision (either positive or negative) in 209 (60.8%), 107 (31.1%), 124 (36.0%), 96 (27.9%), and 61 (17.7%) recommendations, respectively, and ran contrary to the actual decision in the remaining ones. Threshold values for either cost-effectiveness or budget impact distinguishing positive from negative recommendations could not be specified. The following parameters reached statistical significance in logistic regression: clinical efficacy (both explicitly positive and explicitly negative evaluations impacted in opposite directions), lack of impact on hard end points, unfavorable safety profile, cost-effectiveness results, and formal shortcomings (all reduced the probability of a positive recommendation). Conclusions: Decision making of the Agency for Health Technology Assessment in Poland is multicriterial, and its results cannot be easily decomposed into simple associations or easily predicted. Still, efficacy and safety seem to contribute most to final recommendations. © 2013 International Society for Pharmacoeconomics and Outcomes Research (ISPOR).