Impact of the EURO-PERISTAT Reports on obstetric management: a difference-in-regression-discontinuity analysis.

Abstract

Population health monitoring, such as perinatal mortality and morbidity rankings published by the European Perinatal Health (EURO-PERISTAT) reports may influence obstetric care providers' decision-making and professional behaviour. We investigated short-term changes in the obstetric management of singleton term deliveries in the Netherlands following publication of the EURO-PERISTAT reports in 2003, 2008 and 2013. We used a quasi-experimental difference-in-regression-discontinuity approach. National perinatal registry data (2001-15) was used to compare obstetric management at delivery in four time windows (1, 2, 3 and 5 months) surrounding publication of each EURO-PERISTAT report. The 2003 EURO-PERISTAT report was associated with higher relative risks (RRs) for an assisted vaginal delivery across all time windows [RR (95% CI): 1 month: 1.23 (1.05-1.45), 2 months: 1.15 (1.02-1.30), 3 months: 1.21 (1.09-1.33) and 5 months: 1.21 (1.11-1.31)]. The 2008 report was associated with lower RRs for an assisted vaginal delivery at the 3- and 5-month time windows [0.86 (0.77-0.96) and 0.88 (0.81-0.96)]. Publication of the 2013 report was associated with higher RRs for a planned caesarean section across all time windows [1 month: 1.23 (1.00-1.52), 2 months: 1.26 (1.09-1.45), 3 months: 1.26 (1.12-1.42) and 5 months: 1.19(1.09-1.31)] and lower RRs for an assisted vaginal delivery at the 2-, 3- and 5-month time windows [0.85 (0.73-0.98), 0.83 (0.74-0.94) and 0.88 (0.80-0.97)]. This study showed that quasi-experimental study designs, such as the difference-in-regression-discontinuity approach, are useful to unravel the impact of population health monitoring on decision-making and professional behaviour of healthcare providers. A better understanding of the contribution of health monitoring to the behaviour of healthcare providers can help guide improvements within the (perinatal) healthcare chain.