October 4, 2019: Ascertaining Death and Hospitalization Endpoints: The TRANSFORM-HF Experience (Eric Eisenstein, DBA, Kevin Anstrom, PhD)


Eric L. Eisenstein, DBA
Associate Professor in Medicine
Duke University School of Medicine

Kevin J. Anstrom, PhD
Professor of Biostatistics and Bioinformatics
Director of Biostatistics, Duke Clinical Research Institute
Duke University School of Medicine


Ascertaining Death and Hospitalization Endpoints: The TRANSFORM-HF Experience


Clinical endpoints: Ascertaining death; Hospitalization; TRANSFORM-HF; National Death Index

Key Points

  • When patient deaths occur outside the care setting, the cause of death may not be reliably documented. For researchers, the challenges of measuring deaths include the lack of a national death data source and incomplete or hard-to-access sources.
  • The death identification and adjudication process differs for explanatory versus pragmatic trials, and has implications for how death endpoints are acquired and measured.
  • The TRANSFORM-HF pragmatic trial is comparing the effects of treatment strategies on long-term outcomes for hospitalized patients with heart failure. The primary study endpoint is all-cause mortality, which is ascertained and verified using a hybrid approach at the clinical site and call center, and includes searching the National Death Index data.

Discussion Themes

What are the tradeoffs in making endpoint ascertainment more simple?

If using a hybrid death data collection strategy, how are discrepancies adjudicated?

Use of call centers that coordinate follow-up patient contact and data collection is a valid approach that ensures a single point of contact for patients or proxies and care providers. This approach should also be supplemented with redundant data sources.

Read more in the Living Textbook about Using Death as an Endpoint and Inpatient Endpoints in Pragmatic Clinical Trials.

#pctGR, @Collaboratory1, @DCRINews