The coverage of elective revascularisation procedure codes in the National Finnish Hospital Discharge Register

INTRODUCTION: Administrative registers provide an attractive data source for real-life effectiveness studies. The validity of coronary artery disease diagnoses in the Finnish National Hospital Discharge Register (HDR) is high but the validity of revascularization procedure codes (percutaneous coronary intervention; PCI and coronary artery bypass grafting; CABG) are unknown. METHODS: All elective PCIs (n = 1771) and CABGs (n = 1819), performed at the Heart Center, Kuopio University Hospital, Finland between years 2007 and 2014 were identified from the unit's Special Electronic Medical Records (sEMR). Personal identity codes were used to link these data to the HDR. Sensitivity was estimated as proportion of operations recorded in sEMR and HDR. RESULTS: With the strictest assumption of exactly the same intervention day in HDR and sEMR, the sensitivity for procedure codes were 84.6% for PCI and 97.1% for CABG, respectively. When a one-day difference was allowed in the dates, the sensitivity increased to 87.6% for PCI and 98.0% for CABG. Altogether 62.1% of elective PCIs and 69.2% of CABGs were correctly classified as elective procedures in the HDR. CONCLUSION: The potential of the research application of the HDR extends beyond traditionally used diagnostic codes. One feasible application is the assessment of real-life effectiveness of different procedures. KEY MESSAGES The majority of PCIs (87.6%) and CABGs (98.0%), performed in the Heart Center, Kuopio University Hospital responsible for the treatment of approximately one fifth of the Finnish population, were captured by the Finnish National Hospital Discharge Register. However, only 62.1% of elective PCIs and 69.2% of CABGs were correctly classified as elective procedures in the HDR. Electivity data were missing from approximately one third of the procedures. This study produces new information of the potential use of HDR for real-life effectiveness studies to support evidence-based decision making in health care.