Background
The increasing number of older, complex patients who require emergency admission to hospital has prompted calls for better models of medical generalist care, especially for smaller hospitals, whose size constrains resources and staffing.

Objective
To investigate the strengths and weaknesses of the current models of medical generalism used in smaller hospitals from patient, professional and service perspectives.

Methods
The design was a mixed-methods study. Phase 1 was a scoping and mapping exercise to create a typology of models of care, which was then explored further through 11 case studies. Phase 2 created a classification using the Hospital Episode Statistics of acute medical ‘generalist’ and ‘specialist’ work and described differences in workload and explored the links between case mix, typology and length of stay and between case mix and skill mix. Phase 3 analysed the relationships between models of care and patient-level costs. Phase 4 examined the strengths and weaknesses of the models of care through focus groups, a discrete choice experiment and an exploration of the impact of typology on other outcomes.

Results
In total, 50 models of care were explored through 48 interviews. A typology was constructed around generalist versus specialist patterns of consultant working. Twenty-five models were deployed by 48 hospitals, and no more than four hospitals used any one model of care. From the patient perspective, analysis of Hospital Episode Statistics data of 1.9 million care episodes found that the differences in case mix between hospitals were relatively small, with 65–70% of episodes accounted for by 20 case types. The skill mix of hospital staff varied widely; there were no relationships with case mix. Patients exhibited a preference for specialist care in the discrete choice experiment but indicated in focus groups that overall hospital quality was more important. From a service perspective, qualitative work found that models of care were contingent on complex constellations of factors, including staffing, the local hospital environment and policy imperatives. Neither the model of care nor the case mix accounted for variability in the length of stay (no associations were significant at p < 0.05). No significant differences were found in the costs of the models. Professionally, the preferences of doctors for specialist versus generalist work depended on their experiences of providing care and were associated with a healthy organisational culture and a co-operative approach to managing emergency work. Concepts of medical generalism were found to be complex and difficult to define, with theoretical models differing markedly from models in action.

Limitations
Smaller hospitals in multisite trusts were excluded, potentially leading to sample bias. The rapidly changing nature of the models limited the analysis of typology against outcomes.

Conclusions
The case mix of smaller hospitals was dominated by patients with presentations amenable to generalist approaches to care; however, there was no evidence to support any particular pattern of consultant working. Matching hospital staff to better meet local need and the creation of more collaborative working environments appear more likely to improve care in smaller hospitals than changing models.

Future work
The exploration of the relationships between workforce, measures of hospital culture, models of care, costs and outcomes in both smaller and larger hospitals is urgently required to underpin service reforms.

Study registration
This study is registered as Integrated Research Application System project ID 191393.

Funding
This project was funded by the National Institute for Health Research (NIHR) Health Services and Delivery Research programme and will be published in full in Health Services and Delivery Research; Vol. 9, No. 4. See the NIHR Journals Library website for further project information.