Assessing the Validity of Netherlands Linked Routine Healthcare Resource Utilisation Data in the Investigation of Single-Inhaler Triple Therapy Effectiveness (INTREPID Trial)
Numbere B1, Smits E2, Holthuis E2, Lu Y1, Fry M3, Compton C4, Ismaila A5, Rothnie KJ4
1Value Evidence and Outcomes, R&D Global Medical, GSK, London, UK, 2PHARMO Institute for Drugs Outcomes Research, Utrecht, Netherlands, 3GlaxoSmithKline, Stevenage, UK, 4Value Evidence and Outcomes, R&D Global Medical, GSK, Brentford, Middlesex, UK, 5GlaxosmithKline, Collegeville, PA, USA
OBJECTIVES: INTREPID was a pragmatic clinical trial, which assessed the effectiveness of single-inhaler triple therapy (fluticasone furoate/umeclidinium/vilanterol) in the treatment of COPD. An exploratory objective of the study was to use linked routine Netherlands healthcare data to assess the validity and practicality of future pragmatic trials using this external database as a method to ascertain healthcare resource utilisation (HCRU) outcomes.
METHODS: We obtained linked secondary care medical record data from Netherlands INTREPID participants who consented to linkage to the Hospital Database of the PHARMO Database Network. We summarised HCRU during the 12 months prior to index (baseline), and up to 6 months post-index (follow-up) to assess face validity of the data.
RESULTS: 203 patients were successfully linked to PHARMO’s Hospital Database (mean age 66 years; 53% female) with 67% of patients having 6 months of follow-up from index date. During the baseline period, 45 patients (22%) had COPD-related hospitalisations with a mean frequency of hospitalisation of 1 (SD: ±0). Overall, 26 patients (13%), had unplanned hospitalisations with a mean frequency of hospitalisation of 1 (SD: ±0), and a mean length of stay of 7 days (SD: ±4); 2 patients (1%) had a mean length of stay in ICU of 15 days (SD: ±4). During follow-up, 3 patients (1%) had COPD-related hospitalisations with a mean frequency of hospitalisation of 2 (SD: ±2). 2 patients (1%) had unplanned hospitalisations with a mean frequency of hospitalisation of 3 (SD: ±1), and a mean length of stay of 14 days (SD: ±14).
CONCLUSIONS: Linked HCRU data were available for INTREPID patients with greater granularity than data generally available from prospectively collected patient recall of HCRU in pragmatic studies. Future studies should consider using linked routine hospital data alongside prospectively collected data in pragmatic trials in routine practice in the Netherlands.Funding: GSK (study 206854).
Conference/Value in Health Info
Value in Health, Volume 25, Issue 12S (December 2022)
Real World Data & Information Systems, Study Approaches
Electronic Medical & Health Records, Health & Insurance Records Systems