Exploring 'risk' in the context of alternatives to unscheduled hospital admission: a realist review conducted as part of the ENHANCE (EvaluatioN of Hospital Admission avoidaNCE services) project.

Problem

Urgent and emergency care services are under unprecedented, sustained pressure. The NHS Long Term Plan and wider health policy advocates for innovative healthcare models that ease pressure on emergency services and minimise adverse events associated with hospital admission. A range of Hospital Admission Avoidance Interventions (HAAIs) exist across settings with significant inequity of provision. Many HAAIs do not endure, or struggle with effective implementation in different settings. This combines to make commissioning, researching and evaluating HAAIs challenging.

Approach

Using a realist evaluation approach, the ENHANCE (EvaluatioN of Hospital Admission avoidaNCE Services) study aims to develop a theory-rooted contextual framework exploring why certain HAAIs work and endure, by combining evidence synthesis, case studies, and an outcome evaluation. Realist methods are appropriate for investigating complex interventions in contextually dynamic settings, as reflected in the current landscape of HAAIs. A rapid realist review was conducted as the first stage of the ENHANCE study to elucidate the causal association between contexts, mechanisms and outcomes (CMO) in HAAIs to generate initial programme theories (IPTs).Selected policy and strategy documents were reviewed, alongside input from public contributors, to refine the scope and develop a framework for evidence synthesis. Database and grey literature searches were conducted to identify empirical evidence, with no limitations on study type. A reverse chronology quota approach for screening prioritised the most contemporaneous (assumed contextually relevant) evidence. Three rounds of data screening, extraction, analysis, and synthesis, based on constructing a long-list of CMO configurations, were carried out until no new CMOs were identified. CMOs were discussed within the wider research team and synthesised under two IPTs, which were further refined by expert and public contributors. The protocol is registered on PROSPERO: CRD42023468852.

Findings

3161 records were identified from database searches and 48 selected for inclusion; a further 13 articles were included from alternative sources . HAAIs identified were broadly categorised as providing either resources for decision-making or resources for responding to and managing complexity outside of hospital. The majority of HAAIs were in emergency departments, people’s usual places of residence, or virtual. Drawing on aspects of middle-range theory (health beliefs, decision-making under uncertainty, and subjective risk perception), two overarching IPTs were generated: for healthcare professionals, decision-making and managing complexity for hospital admission avoidance is dependent on finding an acceptable subjective risk threshold; for patients, acceptability of hospital admission avoidance is dependent on being able to navigate uncertainty. Three intervention contexts were identified which influenced the mechanistic pathway to outcomes within different HAAIs: risk perception, risk appetite and perceived control.

Consequences

Hospital admission avoidance has inherent risks which are often unmeasurable, introducing uncertainty. Intervention strategies that act on risk perception, risk appetite or perceived control of risk may improve acceptability of HAAIs and their ability to endure.