Quality assurance of the Western Denmark Heart Registry, a population-based healthcare register

INTRODUCTION: During the past decade, the mandatory population-based healthcare database, the Western Denmark Heart Registry (WDHR), has provided the data for several research projects. As in most clinical registries, the data quality has not been validated thoroughly. This study was undertaken to evaluate the quality of registrations in the WDHR.

METHODS: The audit supervised procedures from involved departments that were performed in 2013. An experienced research nurse completed data collection, and an experienced consultant evaluated the agreement between the WDHR and patient records. Indistinct data from patient records were determined after consulting a specialist from the department in question. Patient files were double-checked in case of disagreements between the involved systems.

RESULTS: The total proportion of errors in the referral date was 16.4% in surgery, 9.8% in percutaneous invasive procedures (PCI), 16.1% in coronary angiography (CAG) and 19.5% in computed tomography (CT)-CAG, while the errors in inhospital dates were slightly lower. In the cardiac surgery registries, the proportion of errors was 3.3% in the history and EuroSCORE module, 1.0% in the procedure module and 2.8% in the discharge module. For PCI procedures, the errors were 3.8% in the history module, 2.2% in the procedure module and 1.6% in the discharge module. CAG and CT-CAG had slightly more errors.

CONCLUSIONS: The quality control of the WDHR revealed that overall data errors were lower than 3% and for procedure-specific registrations including indications and complications, the error rate was below 1.5%. The WDHR is valid and may be used in contemporary epidemiological studies.

FUNDING: none.

TRIAL REGISTRATION: not relevant.