Mortality in adult congenital heart disease: Are national registries reliable for cause of death?

BACKGROUND: Statistics on cause-specific mortality are important for prognostic research. The aim of this study was to assess the utility of the national mortality registry in research on causes of death in adult patients with congenital heart disease (CHD). METHODS: The CONCOR registry of over 10,000 adults with CHD was used to verify the causes of death provided by the WHO guidelines based national mortality registry, by linkage. RESULTS: Of 7277 patients linked to the national mortality registry, 196 (2.4%) were recorded deceased, versus 228 deceased patients (3.1%) recorded in the CONCOR registry, during a follow-up of 25,900 patient years. Median age at death was 48.9years. Of all deaths in the CONCOR registry, 77% had a cardiovascular origin; nearly 50% were due to progressive heart failure and arrhythmias. The national mortality registry recorded death due to progressive heart failure and arrhythmias in only 8.5%. Moreover, this registry recorded death with an 'unspecified' cause in approximately 30%, primarily containing patients who died due to progressive heart failure and arrhythmias according to their medical records. CONCLUSION: WHO guidelines based national mortality registries lack the specificity and completeness needed for accurate research on causes of death in adult patients with CHD.