Identifying Outcome Measures for Coronary Artery Disease Value-Based Contracting Using the Delphi Method

Abstract

Introduction Value-based contracts (VBCs) that link drug payments to disease-related performance metrics aim to increase the value and lower the cost of medications by aligning incentives and sharing risk between payers and pharmaceutical manufacturers. This study sought to identify outcome measures that are meaningful to key stakeholders to inform VBCs for coronary artery disease (CAD) medications. Methods We administered a modified Delphi survey to gather expert opinion from a diverse panel of patients ( n = 9), cardiologists ( n = 4), primary care physicians ( n = 5), payers ( n = 2), pharmacy benefits managers ( n = 3), and pharmaceutical company representatives ( n = 2). A list of 16 CAD-associated clinical indicators was generated from the literature and expert consultation. Delphi participants rated the importance of each outcome on a five-point Likert scale, and selected the three most meaningful outcomes. We defined consensus as ≥ 75% agreement on the importance of an outcome (Likert scores 4 or 5 or selection of an outcome as most meaningful). Results Eleven of 13 outcomes reached consensus for importance on the Likert scale. “Preventing heart attacks” was selected as the most meaningful outcome (80%) while “preventing death” ranked second (76%). Conclusions Our study results verify the utility of a widely used clinical CAD outcome measure, myocardial infarction events, for the purpose of pharmaceutical value-based contracting. Electronic supplementary material The online version of this article (10.1007/s40119-019-0132-7) contains supplementary material, which is available to authorized users.

Quality measurement for cardiovascular diseases and cancer in hospital value-based healthcare: a systematic review of the literature

“…In our review, it was found that administrative data were the least frequently used source for outcome measurement, and they were limited to measures such as survival and mortality [ 40 , 41 , 43 , 45 , 50 , 52 , 56 – 61 , 64 – 66 , 69 – 72 , 75 , 76 ]. Data reported by clinicians and patients were more frequently used in the studies reviewed.…”

“…Interestingly, patients were involved in the development of the standardized outcome sets to ensure that their perspectives were incorporated into measures development [ 39 , 42 , 45 , 55 – 58 , 60 , 64 , 68 – 71 , 74 , 75 ]. Relevant PROMs can orient the patients about the potential outcomes of their medical condition and, at the same time, solicit feedback on their needs and priorities regarding the delivered treatment modalities.…”

Tailoring lipid management interventions to reduce inequalities in cardiovascular disease risk management in primary care for deprived communities in Northern England: a mixed-methods intervention development protocol

“…This may be measured according to several characteristics including income, employment, education, disability, crime, housing and services and living environment deprivation. Because there is no universal recommendation for core outcome sets in studies on CVD prevention,19–21 studies will be eligible for inclusion regardless of outcomes measured or reported for health outcomes, this may include vascular-related outcomes, cognitive and functional outcomes, lifestyle, medical risk factors, cardioprotective medications and patient-reported outcome measures. Any measures of professionals’, patients’ and/or families’ knowledge, attitudes or satisfaction will also be included.…”

Identifying Outcome Measures for Migraine Value‐Based Contracting Using the Delphi Method

“…In Round 2, the panel's collective Round 1 rating of each outcome was displayed alongside the individual's original Round 1 rating, following Delphi study methodology. [23][24][25]40,41,42 For example, if in Round 1 a participant rated a given outcome as "slightly important" (2) and the average panel rating for that outcome was "3.50," both descriptors were displayed alongside each outcome in the Round 2 survey. Stakeholders were given the opportunity to adjust their Round 2 outcome ratings based on the collective panel information.…”

“…Stakeholders who completed the Round 1 survey were eligible to complete the Round 2 survey. In Round 2, the panel’s collective Round 1 rating of each outcome was displayed alongside the individual’s original Round 1 rating, following Delphi study methodology 23‐25,40,41,42 . For example, if in Round 1 a participant rated a given outcome as “slightly important” (2) and the average panel rating for that outcome was “3.50,” both descriptors were displayed alongside each outcome in the Round 2 survey.…”