Background Evidence from evaluations of emergency departments (EDs) with co-located primary care services suggests that they influence additional demand for non-urgent care (provider-induced demand). In a realist review of the literature on the effects of primary care services in EDs we proposed a theory that when primary care services are distinct at an ED they may encourage additional primary care demand and when primary care clinicians work indistinctly in the ED there is no additional demand. We aimed to explore evidence for this theory and explain contexts, mechanisms and outcomes that influence such demand.

Methods We used realist evaluation methodology and carried out observations of key processes. We interviewed 23 patients, 21 ED clinical directors, 26 other ED staff members and 26 GPs at 13 EDs (England & Wales). Field notes and audio-recorded interviews were transcribed and analysed by creating context, mechanism and outcome configurations to refine and develop theories relating to provider induced demand.

Results EDs with distinct primary care services were perceived to attract more demand for primary care than EDs where primary care clinicians worked indistinctly because the primary care service was visible, widely known about, enabled direct access, and received NHS 111 referrals. Other influences on demand were patients’ experiences of accessing primary care, the capacity for urgent care in the community, location of the ED and public transport links, service design and developments (new buildings, renovations) and population characteristics (unfamiliarity with local healthcare services, not registered with a GP or different cultural perceptions of seeking health care).

Conclusions A range of patient, local-system and wider-system factors contribute to additional demand at an ED with co-located primary care services. Our findings can inform providers and policymakers in developing strategies to limit the effect of these influences on additional demand.