9 General introduction With the aim of gaining an understanding of variation in childbirth interventions, research has focused on increasing rates of interventions and variation between countries (1–5). Beside variation between countries, there also appears to be variation in childbirth interventions within countries. Practice variation in Dutch maternity care was described in 2014 based on data from the national perinatal register (6). Analysis of these data showed that the rates of interventions, such as caesarean section, instrumental birth, and induction of labour varied widely between different hospitals. For example, there was little variation between hospitals in the rates of planned caesarean sections, but there was much variation in the rates of unplanned caesarean sections. Hospitals with a low rate of unplanned caesarean sections had a rate of around 4.8%, while the highest rates were around 21%. A similar variation was found between hospitals for induction of labour. Practice variation has also been described in primary midwifery practices, with variations in intrapartum referral rates from 9.7% to 63.7% (7). Warranted versus unwarranted variation Interventions in childbirth are in certain circumstances useful to prevent perinatal morbidity and mortality (1). Some practice variation in childbirth interventions is to be expected as care is adapted to medical conditions or the woman’s preferences. However, large variations in care of fairly homogenous populations may indicate insufficient quality of care. Underuse of interventions can lead to preventable morbidity and mortality, while overuse of interventions during maternity care can result in medicalisation of physiological pregnancy and childbirth (1,2). The use of unnecessary interventions can harm women - physically and psychologically - and their newborns, and increase health care costs (3). Variations that reflect under- or overuse of interventions may be unwarranted. Several researchers have tried to define warranted and unwarranted variation (8–10). Sutherland and Levesque have designed an analytical framework that identifies elements associated with warranted and unwarranted variation (8). These elements can be used to assess whether the variation is warranted or unwarranted and include the categories: evidence, capacity, and agency. These elements are interrelated and highly sensitive to context, such as patients living in lower socio-economic status areas are often reported to have worse outcomes than patients living in higher socio-economic status areas. This makes assessment difficult and requires nuance and reflexivity. Because the elements cannot be separated from the context, warranted and unwarranted variation cannot be explained by only using quantitative data sets (8). Causes of unwarranted variation are 1) lack of evidence-based care (evidence); 2) differences in the availability of healthcare resources (capacity); and/or 3) care providers offering care based on the beliefs and personal interests (agency) (8). In other words, when variation cannot be explained by medical 1
RkJQdWJsaXNoZXIy MTk4NDMw