Table 2 Levels and themes contributing to theories

1. Patient

Patients attend EDs based on knowledge of their medical conditions or due to convenience/ preference.

A. Experience and assessment of condition:

Clinical Directors (CDs) reported patients are “correct” in choosing which services to attend (Hospitals 3, 4, 6, 11) and stated that those who attend are sicker than patients who attend community primary care (Hospitals 3, 4, 18). A patient in one hospital reported that she knew when to take her child to the ED because of past experience (hospital 3)

“[We’ve been in] hospital once a month…admitted through the GP or knowing the signs and going to A&E to get him checked out.”

Parent of child seen by GP, Hospital 3 (inside-integrated model).

A. Patients with persistent health problems who have previously accessed EDs (C), or with potentially serious symptoms, have good knowledge and seek emergency care EDs (M) and judge that it is appropriate to attend EDs rather than community primary care services (O).

B. Convenience (location or time) & C. Preferences (including second opinions)

Patient’s access EDs with separate (distinct) primary care services in EDs if they are geographically convenient or at convenient times. Reception staff revealed that services can be accessed all day (Hospitals 6, 10, 13). Clinicians (Hospital 6) stated that patients attend EDs in the morning, despite having afternoon appointments GPs, because they are at the hospital for other appointments or when accompanying relatives e.g., a patient attended for an ultrasound in the hospital and had a foot problem, so attended the ED too.

“[Patient had appointment with GP] but thinks ‘if I go to the hospital now, I’ll be done by lunch time’. And [within] consultations say “excellent, that means I don’t have to go to my GP”.

CD, Hospital 6, (inside-parallel model).

B. Patients with difficulty accessing GP appointments in an acceptable timeframe (C) they may believe that EDs are the best place to attend for urgent care (M) and attend EDs with a distinct primary care service for emergency or primary care services (O).

C. Patients who live /work closer to EDs (especially those with a visible and distinct service) are more likely to attend there with a primary care problem (C) because of convenient local access where people expect to be seen more quickly than in community primary care or for a second opinion (M), generating additional demand for primary care at the ED (O)

2. Local systems

Whether patients can access care and referral pathways or new buildings and publicity.

A. Access to community primary care/hospital services

Some local primary care services were perceived as inaccessible (Hospitals 4, 5, 14) and the 111 service directed patients to EDs, which increased demand for primary care services in the ED.

Patients unable to access timely GP appointments attended EDs either, with the intention of accessing emergency care (Hospitals 13, 16), or to access an inside-parallell (distinct) primary care services at an ED (Hospital 6). These influences increased demand for primary care services and overall, ED service workload.

“I was breathless for, well, days beforehand and couldn’t get an appointment at the doctors and I thought, well, I’ll just go up to the walk-in centre at the hospital.”

Patient seen by ED clinician (Hospital 6 – inside-parallel model)

A. When patinets perceive that they are not able to access local primary care services (C), they may choose to access an ED with a distinct primary care service or they contact 111 service and are advised to seek urgent care at an ED (M) thus generating additional demand in the ED (O).

B. Urgent care referral pathways

Some patients were advised to attend an ED by community primary care services due to their insufficient capacity to see urgent presentations, 111 services also referred primary care patients to an inside- integeated ED when they had conditions that would be more suitably treated within community primary care.

“[Patients say] “we phoned 111 and they said go to A&E” … 111 is not a re-direction service… it’s a misdirection service…”,

CD, Hospital 14 (inside-integrated model)

B. When primary care services have to refer patients to an ED because they have no capacity to see urgent care patients or patients are inappropriately assessed by the 111 service as having a problem that could be seen in the ED (C) patients that could otherwise be seen in community primary care are sent to an ED (M) thus generating additional primary care demand in the ED (O).

C. Service improvements & publicity

Service developments such as new buildings or renovations to add a primary care services to an ED were seen by CDs to potentially influence additional demand (hospital 3 and 6), and was reported to be evident where there is a distinct service that patients can walk in to (hospital 6) Publicity, and increased public knowledge about services also predisposed patients from wider areas to attend (Hospital 6).

“When we opened this building, our attendance rate went up 30%... We started to see whole populations coming to us which never came before.... When you put the service and make it available, it generates work.”

ED consultant, Hospital 6, inside-parallel model)

C. Service developments involving new or renovated buildings (C), media publicity and increased public awareness, may lead to patients from local and further areas choosing to attend the service (M) creating additional demand for emergency and primary care (O).

3. Wider system (regional / national influences)

A. Population characteristics

Populations of patients were sometimes viewed by staff as able to judge which conditions were “appropriate” to present to the ED or considered “stoic” in terms of their health-seeking behaviour. In some areas, populations were characterised as having large numbers of temporary residents, such as visitors, tourists, and transient workers who, due to unfamiliarity with services, choose to attend EDs if unwell.

In diverse cities in which people have recently arrived in the UK, or with large student populations, different cultural perceptions of accessing healthcare, or not being registered with a GP were factors considered to make it more likely to seek primary care at EDs.

“We see minor injury and fractures, and it’s linked with the rural population, in that a high proportion do have fractures or true injuries”.

CD, Hospital 9, (inside-integrated model)

A. In areas where people are not registered with a GP, have different cultural perceptions about how to access health care, or are unfamiliar with local services (C), patients may attend EDs for primary health care needs (M), and generate demand for primary care at the ED (O). Similarly, in areas with less diversity, or rural areas (C), attendances were perceived to be associated with minor injuries and fractures (C), and staff report that people tend not attend EDs with primary care problems (M) ensuring less demand for primary care in EDs (O).

B. Service improvements & unintended consequences:

We noted a regional 111 directive of advising patients to attend EDs with the shortest waiting times.

“Somebody who lived in [elsewhere] was told [by 111] to come here, bypassing the ED at [other hospital] to come to us…all his follow-up will now be under us”

Nurse, Hospital 6, (inside-parallel model)

B. Service improvements that focus on waiting times (C) may lead to patients being referred to other EDs (M) which creates extra demand and workload at more efficient hospitals (O).