Using perinatal audit to promote change: a review.

Close to half of all infant deaths world-wide now occur in the first week of life, almost all in developing countries, and the perinatal mortality rate (PNMR) is used as an indicator of the quality of health service delivery. Clinical audit aims to improve quality of care through the systematic assessment of practice against a defined standard, with a view to recommending and implementing measures to address specific deficiencies in care. Perinatal outcome audit evaluates crude or cause-specific PNMRs, reviewing secular trends over several years or comparing rates between similar institutions. However, the PNMR may not be a valid, reliable and sensitive indicator of quality of care at the institutional level in developing countries because of variations in the presenting case-mix, various confounding non-health service factors and the small number of deaths which occur. Process audit compares actual practice with standard (best) practice, based on the evidence of research or expert consensus. Databases reviewing the management of reproductive health problems in developing countries are currently being prepared so as to provide clinicians and health service managers with up-to-date information to support the provision of evidence-based care. Standard practice should be adapted and defined in explicit management guidelines, taking into account local resources and circumstances. Forms of process audit include the review of care procedures in cases which have resulted in a pre-defined adverse outcome, know as 'sentinel event audit'; and the review of all cases where a particular care activity was received or indicated, known as 'topic audit'. These are complementary and each depends on the quality of recorded data. The forum for comparing observed practice with the standard may be external, utilising an 'expert committee', or internal, in which care providers audit their own activities. Local internal audit is more likely to result in improvements in care if it is conducted in a structured and culturally sensitive way, and if all levels of staff are involved in reviewing activities and in formulating recommendations. However, further research is needed to identify the factors which determine the effectiveness and sustainability of perinatal audit in different developing country settings.