Errors in the certification of neonatal death

Objective: To examine the precision of the perinatal death certificate (PDC ) and ascertain the possible sources of error in the certification of neona tal deaths. Methods: The 'Main' and 'Other' causes of death recorded on the PDC were ob tained from the Registry of Births, Deaths and Marriages and compared with those from a clinicopathological summary (CPS) completed after all pending laboratory results and/or autopsy information were available. Results: There were 179 neonatal deaths during the 7 year period under revi ew. The PDC and CPS main causes of death were concordant in 103 of 179 infa nts (58%) and discordant in the remaining 76 infants (42%). The PDC main ca use of death was incorrectly classified in 61 of 76 infants (80%) with disc ordant findings and was incompletely classified in the remaining 15 infants (20%). The following discordancies were recorded for the 61 infants with a n incorrect classification: (i) transposition of the 'Main' and 'Other' cau ses of death, resulting in a sequencing discordancy in 14 infants (23%); (i i) recording a non-pathological condition as the main cause of death in 40 infants (66%); and (iii) recording an incorrect pathological condition as t he main cause of death in seven infants (11%). Eight of the 61 (13%) incorr ect classifications and four of the 15 (27%) incomplete classifications wer e associated with laboratory and/or autopsy data being unavailable when the PDC was completed. Conclusions: The concordancy between the PDC and CPS would have increased f rom 58 to 91% if the 'Main' and 'Other' causes of death had been sequenced correctly, if the main cause of death had been ascribed to a pathological d isease rather than a non-pathological condition and if corrective informati on from pending laboratory tests and/or autopsy examination had been made a vailable to the Registry of Births, Deaths and Marriages.