Validation of local review for the identification of contributory factors and potentially avoidable perinatal deaths

Background Reporting on perinatal mortality commenced 2006 in New Zealand through the Perinatal and Maternal Mortality Review Committee (PMMRC). Following review of international models, a process was developed for use in local review to identify contributory factors and potentially avoidable perinatal deaths. Local review of 720 perinatal deaths in 2009 found contributory factors in 24% of deaths and 14% to be potentially avoidable. Aims To validate the process of local review for identification of contributory factors and potentially avoidable perinatal deaths. Material and Methods Records of 48 perinatal deaths were reviewed by an independent multidisciplinary panel using the same methodology as local review to determine agreement between local and independent review for identification of contributory factors and potentially avoidable perinatal death. Results Independent review found contributory factors in 54% of deaths compared to 40% by local review. Independent review identified eight deaths and local review identified one death with contributory factors not identified by the other review. Kappa statistic for agreement for identifying contributory factors was substantial [0.63 (0.42, 0.84)]. Independent review found 42% of deaths potentially avoidable compared to 23% by local review. Independent review identified 10 deaths and local review identified one death not identified by the other review. Kappa statistic for agreement for identifying potentially avoidable deaths was moderate [0.50 (0.26, 0.73)]. Conclusions This study provides validation of local review for identification of contributory factors in perinatal death. The higher proportion of potentially avoidable perinatal deaths identified by independent review compared to local review requires further exploration.