Validation of perinatal causes of death in death certificates

OBJECTIVE: The objective of this study was to assess the validity of perinatal causes of death in death certificates and the influence of some variables related to perinatal mortality. PATIENTS AND METHODS: All perinatal deaths having occurred in Barcelona in 1994 and reported in death certificates were included. Additional sources of information were certificates of abortions and also other documents such as autopsy reports and medical records, that served as references to determine the most reliable or standard cause of death (SBC) The SBC was determined by a group of experts. The validity was assessed by comparing the cause of death as stated in the death certificate (BC) and the SBC, calculating the general agreement, the detection rate (DR) or sensitivity and the confirmation rate (CR) or positive predictive value. The chi-square test was used to compare proportions. RESULTS: In 166 (91.2%) of the 182 cases, the BC and SBC were assigned to the same group of the International Classification of Diseases. For congenital defects, the DR was 85.7 (95% CI: 75.1-96.3) and the CR was 81.8 (95% CI: 70.4-93.2). For problems in the perinatal period the DR was 95.6 (95% CI: 92.1-99.1) and the CR was 94.9 (95% CI: 91.2-98.6). Both were considered reliable causes. These rates were similar for the different categories of other variables (birth weight, gestational age, vital status, sex, hospital of birth, autopsy and mother's age). Nevertheless, in the study of detailed causes (3 digits of the International Classification of Diseases) the general agreement was much lower (40.1%). Except for the congenital defects of the nervous system, the remaining causes of death had a low validity. CONCLUSIONS: This study shows the low validity of detailed causes of perinatal death reported in death certificates. It is necessary to improve the completion and coding of death certificates following the WHO recommendations.