Background. In our population-based study of pregnancy outcome in women with rheumatic disease we based our assessment on the Medical Birth Registry of Norway (MBRN). We evaluated the MBRN as a source of data for such epidemiologic research by assessing the validity of a diagnosis of rheumatic disease in the MBRN against a gold standard. The validity may also be interpreted as a quality indicator, reflecting an obstetrician's attention to rheumatic diseases in pregnancy.
Methods. Using the mother's national identification number the MBRN was linked with local hospital databases (gold standard), which contained data of mothers with rheumatic disease. The sensitivity of the MBRN was calculated as the proportion of all cases registered locally with a diagnosis of rheumatic disease notified to the MBRN. The correctness of type differentiation was calculated as the proportion of all cases notified to the MBRN that was correct with respect to the type of rheumatic disease.
Results. Among 169 mothers, 149 had a diagnosis in the MBRN, representing a sensitivity of 88.2%. Altogether, 97.3% of the diagnoses (145/149) were correct with respect to the type of rheumatic disease.
Conclusion. Taken in to consideration the limitations of the study, namely the small numbers studied, we assume a rather high validity of rheumatic disease diagnoses in the MBRN, probably reflecting a high level of attention in the obstetric care for these patients.