Objective To explore implementation of the modified early obstetric warning system (MEOWS) in practice to further understanding about the influence of contextual factors.

Methods An ethnographic study using observations (>120 h), semi-structured interviews (n=45) and documentary review was performed in the maternity services in two UK hospitals over a 7-month period. Doctors, midwives and managers participated in the study and data were analysed thematically.

Results For women admitted to hospital in the antenatal and postnatal period with an established risk of morbidity, the MEOWS enabled communication about vital signs from junior to senior midwives and obstetricians. The trigger prompts helped shape shared understandings of maternal complications. However, midwifery and obstetric staff questioned the added value of an extra chart in the postnatal period given the low incidence of maternal complications and the resulting increase in workload. In an effort to prioritise workload demands and respond to the immediate needs of both women and their babies, midwives exercised professional discretion regarding its use. However, discretionary use of MEOWS meant the loss of a potential universal safety net for detection of deterioration.

Conclusions Despite a decade of use in acute settings, research into the effectiveness of early warning systems still yields conflicting results. Widespread policy support for the MEOWS is based on its intuitive appeal and no validated system for use in the maternity population currently exists. Our findings suggest that, while the MEOWS has value in structuring the surveillance of hospitalised women with an established risk of morbidity, the complexities of managing risk and safety within the maternity pathway, the associated opportunity costs of MEOWS and variation in implementation currently call into question its role for routine use.