OBJECTIVES: To review ISPOR Presentations Database (IPD) to assess the feasibility of extracting clinical, humanistic, and economic data for Centers for Medicare & Medicaid Services (CMS) public call for evidence on Medicare Drug Price Negotiation (DPN).

METHODS: IPD, a searchable archive of more than 60,000 citable research abstracts of podium and poster presentations from ISPOR conferences, was used to extract evidence on apixaban and rivaroxaban, top two cardiovascular medications selected for CMS DPN in 2023. Abstracts were published in ISPOR's flagship journal, Value in Health, and session presentations from ISPOR's conferences were also included. Apixaban and rivaroxaban were searched exclusively as keywords. Outcomes of interest were study methods, QALY reporting, population, and CMS-specified considerations. Non-English abstracts were excluded. Two independent reviewers conducted extraction and cross-verification.

RESULTS: Of 298 total IPD entries, 43% had either a poster or presentation content attached. In the database, 80% of apixaban and 74% of rivaroxaban assessed comparative effectiveness. The most captured study methods were cost-effectiveness analyses (44% for apixaban; 42% for rivaroxaban). 44% of apixaban and 45% of rivaroxaban reported QALYs. 24% for apixaban and 23% for rivaroxaban were in US perspective. CMS's interests such as unmet needs, special populations, risks/harms, and unique scenarios related to benefit, safety, patient experience, and health equity were inadequately captured, partly due to ambiguity in CMS criteria. 12% of apixaban IPD entries and 7% for rivaroxaban specified Medicare or elderly population.

CONCLUSIONS: The ISPOR Presentations Database provides grey literature that can add HEOR evidence to CMS’ decision-making process. Its contents may have reduced publication bias compared to published papers. Further clarification of CMS’ definition of special populations, health equity, and unique considerations related to drug efficacy and safety is needed. Limitations such as potential duplications and preliminary, non-peer-reviewed posters should be considered when leveraging this database.