gms | German Medical Science

Background: Obstetrical interventions can be important for preserving maternal and neonatal health. Nevertheless, they are not always medically indicated and do not necessarily improve the health of mothers and their children. As an example, higher rates of augmentation of labour or caesarean sections are related to maternal postpartum haemorrhage [1], [2]. Therefore, a critical appraisal of aspects influencing intervention rates is necessary. Mainly, system immanent factors such as the number of cases and the time of day are of interest, as they provide a chance to decrease intervention rates by modifying system structures. It is the aim of this study to inform the debate by analyzing intervention rates in different-sized hospitals with regard to the time of day.

Materials and methods: Retrospective cohort study. Data: hospital deliveries in the German federal state Hesse from 2005 to 2015 (n = 305,980). Source of data: perinatal database of Hesse. Inclusion criteria: Primi-/multiparae, singleton in cephalic presentation, 37+0 – 41+6 weeks gestation. Exclusion criteria: induction of labour, stillbirth, HELLP syndrome; (pre) eclampsia; placenta praevia, and planned caesarean section. Data analysis: descriptive statistics, Chi-square tests, regression analysis; SPSS Statistics 23.0. Dependent variables: augmentation of labour, episiotomy, operative assisted delivery and unplanned c section. Independent variables: day- and night-time (8.01am to 8.00pm/8.01pm to 8.00am), hospital size (category 1: up to 500 births per year, category 2: 501-1,000 births per year, category 3: 1,001-1,500 births per year, category 4: above 1,500 births per year).

Results: In hospitals with more than 1,500 births annually there are significantly less augmentations of labour (OR 0.52 CI 95% [0.50-0.53], p<0.01), episiotomies (OR 0.58 CI95% [0.56-0.60], p<0.01) or unplanned caesarean sections (OR 0.84, CI95% [0.81-0.87], p<0.01) in comparison to hospitals with up to 500 births per year. In all four hospital categories, there are significantly more episiotomies, augmentations of labour, operative assisted deliveries and unplanned caesarean sections in the daytime compared to night shifts. Differences in augmentation-, episiotomy- and caesarean section rates between day and night are particularly obvious in hospitals with up to 500 births per year compared with all other hospital categories.

Conclusion: Considering the sample of low risk women, it was expected that there are no daytime variations in intervention rates during labour. However, our data show some significant variations according to the time of day. It is assumed that system immanent factors exist which cause these variations. Furthermore, our data seem to support current centralisation efforts and the associated close-down of small obstetrical units. However, due to the chosen study design a causal relationship cannot be established. Further research about the relation between intervention rates and system immanent factors is needed.

Ethical criteria and conflict of interests: This research was submitted to an ethics committee. It is self-financed and there is no conflict of interest.