To study consistency of data and completeness of reporting in a national vascular registry, NorKar, and a national administrative registry, The Norwegian patient register (NPR).
Comparative registry-based national study supplemented with a comprehensive control of patients registered in one major hospital.
All patients registered with a procedure-code for treatment of AAA in NorKar or NPR during 2001 or 2002, were included.
We compared the reporting of procedure-codes, diagnosis-codes and in-hospital deaths after treatment for abdominal aortic aneurysm (AAA) in the two registries to evaluate completeness. Consistency between procedure-codes and diagnoses were evaluated within both registries. Completeness of reporting to one NorKar Local Registry was investigated in more detail in one of the hospitals.
Compared with the NPR numbers, NorKar contained 69% of the patients treated for AAA in Norway, while completeness for NorKar member hospitals was 84%. The detailed investigation in one of the hospitals showed a completeness of 91% and a false inclusion of 5.3% of all cases treated for AAA. The consistency between procedure-codes and diagnosis-codes was 93% in both registries. We found evidence of substantial underreporting of in-hospital deaths to NorKar in several hospitals. Overall reporting of early deaths to NorKar relative to completeness of reported cases was estimated to 72%.
There is an underreporting of patients with AAA to NorKar according to the NPR numbers and a need for better control of procedure-diagnosis consistency in both registries. There seems to be a substantial underreporting of early deaths to NorKar. Introduction of unique patient-identifiable data could improve the quality of both registries by making matching of data possible.