Levine C, Ansquer V, Kurosky S. Evaluation of data-collection duration in retrospective noninterventional medical record review of oncology studies conducted in the United States, Canada, United Kingdom, Germany, France and Spain. Poster presented at the ISPOR 21st Annual European Congress; November 13, 2018. Barcelona, Spain.


OBJECTIVES: Health economic and outcomes focused retrospective noninterventional medical record review studies (NIMRRs) are critical in generating real-world evidence to support economic models, HTA submissions, develop product value strategies, and identify areas of unmet treatment needs. The aim of this study was to describe characteristics associated with overall data collection duration (DCD) in NIMRRs conducted in North America and Europe.

METHODS: We analyzed 5 years (2013-2018) of internal DCD data (time from first to last collected chart) for 68 NIMMRs. In these studies, physicians treating patients with cancer were recruited to anonymously abstract data related to patient and disease characteristics, treatments, clinical outcomes (e.g., overall survival, disease progression, treatment response), and health care resource utilization from patient charts. The mean DCD was compared by country, common vs. rare cancer, estimated data collection form (DCF) abstraction length (20-29 vs. 30+ minutes), and total sample size.

RESULTS: The number of studies completed by country ranged from 8 (France) to 15 (UK). Most studies were in common cancers (92.7%) and the mean number of charts collected per study was 188 (min.:52, max.:500). DCD in weeks varied by country (UK: 7.1, Spain: 7.3, US: 8.0, France: 8.4, Germany: 9.0, Canada: 10.2 [P=0.002]), but did not differ between common and rare cancers (8.2 and 8.7 weeks [P=0.624], respectively). Estimated DCF length was not associated with DCD. DCD was longest when targeting <100 charts (9.9 weeks) and shortest when targeting 300+ charts (6.7 weeks).

CONCLUSIONS: DCD varied significantly by country, potentially driven by cultural differences in seasonal availability and physicians’ willingness to participate. We anticipate the relationship between sample size and DCD is confounded by country and disease area as sample size targets are typically determined a priori based on these characteristics. Understanding differences in factors associated with DCD can inform study launch timing, country selection, and overall timelines.

Share on: