Drug Registries and Approval of Drugs: Promises, Placebo, or a Real Success?

Carla J. Jonker*, Marcel S.G. Kwa, H. Marijke van den Berg, Arno W. Hoes, Peter G.M. Mol

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

2 Downloads (Pure)

Abstract

Purpose: As part of the approval process, regulatory authorities often require postauthorization studies that involve patient registries; it is unknown, however, whether such registry studies are adequately completed. We investigated whether registry studies for new drugs were performed as agreed at time of approval. Methods: This study reviewed protocols and follow-up reports for 73 registry studies that were proposed for 43 drugs approved by the Committee for Medicinal Products for Human Use in Europe in the period 2007 to 2010. Results: The data lock point of January 1, 2016, was taken to allow a 5-year follow-up period for each drug after approval. At that time, 2 studies (3%) in registries had been finalized, 19 registries (26%) had not enrolled any patients, and 52 studies (71%) were ongoing. The median enrollment was 31% (interquartile range [IQR], 6–104) of the required number of patients for 41 registry studies that had a predefined sample size, 30% (IQR, 2–101) for nonimposed registries, and 61% (IQR, 18–144) for imposed registries. Implications: Enrollment of patients into postapproval registries is poor, although the results for imposed registries seem better. Currently, registries only have a limited impact on resolving gaps in the knowledge of a drug's benefits and risks at time of marketing authorization.

Original languageEnglish
Pages (from-to)768-773
Number of pages6
JournalClinical Therapeutics
Volume40
Issue number5
DOIs
Publication statusPublished - May 2018

Keywords

  • new drugs
  • patient enrollment
  • postapproval data
  • registries

Fingerprint

Dive into the research topics of 'Drug Registries and Approval of Drugs: Promises, Placebo, or a Real Success?'. Together they form a unique fingerprint.

Cite this