Reliability in perinatal audit. A qualitative study--a preliminary investigation.

The aim of this study was to elucidate some aspects of reliability in perinatal audit, both by describing aspects of the ways in which consensus is reached between experts in perinatal audit and by examining the extent of agreement between two separate panels assessing peri- and neonatal care in 17 cases of neonatal deaths. The consensus process was elucidated by a questionnaire answered by 16 perinatal experts who had participated in a perinatal audit and by analyses of audiotapes. In 70-95% of the cases, most of the experts had reached the same conclusions during their work at home as the panel did after the joint discussions. The audiotapes revealed that the experts adjust their individual interpretations during discussions with other experts. Agreement in evaluation between two panels was elucidated by agreement in evaluation of the quality of antenatal care, intrapartum care, and neonatal care in the 17 cases of neonatal deaths. The two panels agreed in evaluation of antenatal care in 15 cases, in intrapartum care in 14 cases, and in neonatal care in 15 cases. Although the material is small, our results seem to indicate that the conclusions drawn in perinatal audit are reliable when the experts have detailed knowledge of the structure of perinatal and neonatal care, when there are agreed standards of satisfactory care, and when they have the possibility of adjusting their individual interpretations during discussion with other experts.