Transparency vs. closed-door policy: Do process characteristics have an impact on the outcomes of coverage decisions? A statistical analysis

• Published in: Health policy (Amsterdam) Vol. 112; no. 3; pp. 187 - 196
• Main Authors: Fischer, Katharina E., Rogowski, Wolf H., Leidl, Reiner, Stollenwerk, Björn
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ireland Ltd 01.10.2013; Elsevier
• Subjects: Associations; Biological and medical sciences; Cost-Benefit Analysis; Coverage; Coverage decision-making; Decision Making; Diseases; Drug Prescriptions - economics; Drug reimbursement; Empirical Research; European Union; Evidence based; Evidence-Based Practice; Fourth hurdle; Health Policy; Humans; Industrialized countries; Insurance Coverage - economics; Internal Medicine; Internet; Logistic regression; Market access; Medical sciences; Medicine; Miscellaneous; Models, Economic; Models, Statistical; Outcome Assessment (Health Care); Participation; Policy making; Prescribed; Public drug formularies; Public health. Hygiene; Public health. Hygiene-occupational medicine; Statistical analysis; Survey data; Surveys; Surveys and Questionnaires; Technology; Technology Assessment, Biomedical - economics; Voting; Logistic regression; Market access; Participation; Industrialized countries; Coverage decision-making; Public drug formularies; Fourth hurdle; Drug reimbursement; Drug; Transparency; Statistical analysis; Prognosis; Accessibility; Decision making; Characteristic; Markets; Decision; Health policy; Characteristics; Evolution; Industrialized country; Comparative study; Public health
• ISSN: 0168-8510; 1872-6054; 1872-6054
• DOI: 10.1016/j.healthpol.2013.04.011

The aim of this study was to analyze influences of process- and technology-related characteristics on the outcomes of coverage decisions. Using survey data on 77 decisions from 13 countries, we examined whether outcomes differ by 14 variables that describe components of decision-making processes and the technology. We analyzed the likelihood of committees covering a technology, i.e. positive (including partial coverage) vs. negative coverage decisions. We performed non-parametric univariate tests and binomial logistic regression with a stepwise variable selection procedure. We identified a negative association between a positive decision and whether the technology is a prescribed medicine (p=0.0097). Other significant influences on a positive decision outcome included one disease area (p=0.0311) and whether a technology was judged to be (cost-)effective (p<0.0001). The first estimation of the logistic regression yielded a quasi-complete separation for technologies that were clearly judged (cost-)effective. In uncertain decisions, a higher number of stakeholders involved in voting (odds ratio=2.52; p=0.03) increased the likelihood of a positive outcome. The results suggest that decisions followed the lines of evidence-based decision-making. Despite claims for transparent and participative decision-making, the phase of evidence generation seemed most critical as decision-makers usually adopted the assessment recommendations. We identified little impact of process configurations.