Intrapartum adverse events in Scotland: ascertainment, assessment and feedback

Abstract

This thesis describes three studies contributing to the goal of reducing perinatal mortality and morbidity. First, I established a national system of adverse outcome audit for intrapartum-related perinatal mortality and morbidity. Second, I measured agreement among different panels assessing quality of care in these cases. Third, I used the technique of critical fiction to develop fictionalised vignettes for use in feedback of audit findings and assessed the impact of these vignettes on maternity clinicians.

In 2005, 81 infants who died or suffered morbidity due to intrapartum events were identified in Scotland’s 18 consultant-led maternity units [Rate 1.48/1000 births (95% CI, 1.18-1.84)]. Case assessment by local risk management panels using a structured proforma identified major sub-optimal care in 30% of cases; misinterpretation of cardiotocographs in labour was the major contributory factor.

I assessed the reproducibility of panel assessments on a random sample of eight cases. Two external panels assessed each case for comparison with the local assessments. When responses of good care / incidental sub-optimal care and of minor sub-optimal care / major suboptimal care were combined, crude agreement among the three panels ranged from 71% to 90% for the various components of care examined.

A group of five maternity clinicians discussed five cases in depth. Discussions were recorded and transcribed and themes elicited. Two fictionalised vignettes were created using the technique of critical fiction. To assess the impact of fictionalised vignettes, 232 maternity clinicians were surveyed (response rate, 70%). Questionnaires contained four vignettes: two fictionalised and two describing real individual cases. When no infoimation was provided on the source of the vignette, 53% of respondents chose a fictionalised vignette as the most believable. When informed which vignettes were ‘real’ and which were fictionalised, Fewer respondents (27%) chose a fictionalised vignette as the most ‘believable’ (p= 0.0004).

It proved feasible to establish a national mechanism for ascertainment and local assessment of cases of intrapartum-related mortality / morbidity. Use of a structured case assessment tool with pre-defined standards resulted in reasonable reproducibility of assessments. Maternity clinicians find fictionalised vignettes ‘believable’ and this approach may be useful in feedback of the findings of adverse outcome audits.