How Long Is Long Enough? Data Collection Duration for Medicines in Managed Access in England

OBJECTIVES: To review the duration of prospective data collection in managed access agreements (MAAs – a form of managed entry agreement) in England. MAAs are funded by NHS England through the Cancer Drugs Fund (CDF) and Innovative Medicines Fund (IMF). The duration of data collection specified in the data collection arrangement (DCA) for an MAA is the designated minimum period, up to a maximum five years, to resolve key uncertainties and optimise the chance of routine commissioning at re-evaluation.

METHODS: Start and end dates of prospective data collection were extracted for all treatments that entered MAAs in England from January 2016 to December 2023. Exclusion criteria were applied, and data sources characterised for analysis.

RESULTS: At December 2023, 67 treatments have or have had an MAA. Nine associated DCAs were excluded over publication date, guidance termination, or retrospective data collection. For the remaining 58 DCAs, median data collection duration was 2.16 years, range less than 1 year to more than 5 years. Most DCAs (48/58, 83%) specified both trial and real-world data (RWD), median duration 1.99 years. A further 12% (7/58) specified RWD only, median duration 3.85 years. The remaining 5% (3/58) specified Phase III trial data only, median duration 0.57 years. Three quarters of DCAs (38/51, 75%) specifying trial data with or without RWD, included later cut-off data from ongoing Phase III trials, median duration 1.78 years. The remaining quarter (13/51, 25%) specified Phase I or II trial data with RWD, median duration 3.16 years.

CONCLUSIONS: There appears to be a correlation between later phase trials with or without RWD and shorter duration of data collection. Treatments where earlier phase trials with RWD or RWD alone were specified to resolve key uncertainties required longer data collection. Earlier licencing by regulators may affect future data source and data collection duration for managed access.