Unexplained variation in hospital caesarean section rates

To the Editor: Lee and colleagues assessed the hospital caesarean section (CS) rates among hospitals in New South Wales according to the Robson 10-group classification. They identified the 20th centile CS rate for each group.1 These data could be used in Australian maternity services to reduce CS rates and potentially increase the quality of care.

Monitoring the quality of care in maternity services has been difficult owing to difficulty in defining quality indicators and appropriate thresholds for these indicators.2 The use of system process control mechanisms3 and cumulative sum (CUSUM) charts may allow individual institutions to monitor and detect deviation from targets in order to improve quality of care.4 The feasibility of CUSUM charts for the monitoring of quality indicators in obstetrics has been demonstrated using a range of quality indicators.5 One of the challenges in applying CUSUM charts to health care is determining acceptable and unacceptable rates for quality indicators.4 Identification of the 20th centile CS rates would allow for the development of an achievable control limit for the CUSUM chart that could be used in a dashboard of indicators.

Variance in care leads to differing CS rates. Using CUSUM charts as part of a process control system aimed at reducing this variance may improve the quality of care in Australian maternity services.