Team interventions in acute hospital contexts: protocol for the evaluation of an initial programme theory using realist methods

Abstract

Introduction : Literature on multi-disciplinary healthcare team interventions to improve quality and safety of care in acute hospital contexts tends to focus on evaluating the success of the intervention by assessing patient outcomes. In contrast, there is little focus on the team who delivered the intervention, how the team worked to deliver the intervention or the context in which it was delivered. In practice, there is therefore a poor understanding of why some interventions work and are sustained and why others fail. There is little emphasis in the literature on how the team delivering the intervention might impact success or failure. Given that team is the vehicle through which these interventions are introduced, it is important to understand interventions from their perspectives. This research seeks to deepen understanding of enablers and barriers for effective team interventions. Using two case studies, we will evaluate previously developed initial programme theories to understand, what worked for whom, in what conditions, why, to what extent and how? Methods and analysis : A realist evaluation approach will be employed to test the previously formed set of initial programme theories. Two multi-disciplinary acute hospital team interventions in two different geographical and organisational contexts will be identified. In case study 1, a theory based approach to interviewing will be used. In case study 2, interview transcripts obtained using a semi- structured approach for primary research purposes will undergo secondary analysis. This will enable a more sensitive look at patterns and variations in patterns of multi-disciplinary team interventions. Researchers will first iteratively interrogate each respective dataset to identify the characteristics or resources present within the specific context that influenced how the team intervention worked to produce particular outcomes. Data will then be synthesised across contexts in order to produce middle range theories and thereby more generalisable insights.