Validation study of uterine rupture registration in the Medical Birth Registry of Norway

Objective. To validate uterine rupture registration in the Medical Birth Registry of Norway (MBRN) between 1999 and 2008, and to identify rupture type and causes of incorrect or missed registration during 1967-2008. Design. Population-based study. Population. The validation sample was 392 958 maternities from 21 maternity units registered in MBRN and local Patient Administration Systems in 1999-2008. In addition we identified type of rupture and causes of incorrect registration among 2 422 934 maternities from 48 units, and 1 449 201 maternities from 21 units during 1967-2008. Methods. Information about uterine rupture in MBRN was compared with information in medical records. Main outcome measures. Sensitivity, specificity, positive and negative predictive value of uterine rupture registration in MBRN. Results. In 1999-2008, there were 10 false positive cases among 151 uterine ruptures registered in MBRN. In addition, 82 ruptures not registered in MBRN, were identified through Patient Administration Systems. The sensitivity, specificity, positive and negative predictive value of the MBRN to detect uterine rupture was 63.2, 99.99, 93.4 and 99.98%, respectively. The incidence of uterine rupture changed after correction from 0.38 to 0.56/1000. During 1967-2008, false positive cases (125) were mainly due to wrong coding by MBRN. Around 60% of 141 false negative cases were due to lack of reporting by maternity units. Complete rupture accounted for 63.9% of ruptures registered in MBRN. Conclusions. The validity of MBRN data on uterine rupture is not optimal, diagnosis being under-reported by around 37%. Ticked boxes may improve the quality of registration.