Abstract Background Implementation of the ‘Sepsis Six’ clinical care bundle within an hour of recognition of sepsis is recommended as an approach to reduce mortality in patients with sepsis, but achieving reliable delivery of the bundle has proved challenging. There remains little understanding of the barriers to reliable implementation of bundle components. We examined frontline clinical practice in implementing the Sepsis Six. Methods We conducted an ethnographic study in six hospitals participating in the Scottish Patient Safety Programme Sepsis collaborative. We conducted around 300 h of non-participant observation in emergency departments, acute medical receiving units and medical and surgical wards. We interviewed a purposive sample of 43 members of hospital staff. Data were analysed using a constant comparative approach. Results Implementation strategies to promote reliable use of the Sepsis Six primarily focused on education, engaging and motivating staff, and providing prompts for behaviour, along with efforts to ensure that equipment required was readily available. Although these strategies were successful in raising staff awareness of sepsis and engagement with implementation, our study identified that completing the bundle within an hour was not straightforward. Our emergent theory suggested that rather than being an apparently simple sequence of six steps, the Sepsis Six actually involved a complex trajectory comprising multiple interdependent tasks that required prioritisation and scheduling, and which was prone to problems of coordination and operational failures. Interventions that involved allocating specific roles and responsibilities for completing the Sepsis Six in ways that reduced the need for coordination and task switching, and the use of process mapping to identify system failures along the trajectory, could help mitigate against some of these problems. Conclusions Implementation efforts that focus on individual behaviour change to improve uptake of the Sepsis Six should be supplemented by an understanding of the bundle as a complex trajectory of work in which improving reliability requires attention to coordination of workflow, as well as addressing the mundane problems of interruptions and operational failures that obstruct task completion.
