Background
There is increasing strain on public hospital emergency departments. Between 2014 and 2015 there was an annual growth of presentations to emergency department (ED) of 3.2%, however this increased to 6.9% in 2020–2021 in Australia [
1]. This trend has also been evidenced in other parts of the world [
2], causing increased pressure on health systems. ED crowding is a common theme across different health systems, even in countries with universal healthcare, though less so in Scandinavian countries [
3]. It is important to understand the reasons for presentation so that models of care can be developed which best meet the needs of patientsA study by Toloo et al. examined GP type presentations to a metropolitan hospital in Brisbane over 3 years found that the number of GP type presentations to the ED varied depending on the method used to measure these presentations [
4]. This varied from 27% using the Australian Health and Welfare Institute metric, verses 6% using the Australasian College of Emergency Medicine (ACEM) measure to 8% using the Spivulis method [
4]. Dinh et al. undertook a retrospective analysis of 11 million presentations over an 11 year period in the greater Sydney area and found that 40% of presentations could be classified as GP presentations and that GP presentations to ED increased across all age groups during the study period [
5]. Regardless of the number of GP presentations to the ED, there is a cohort which could be reviewed in the community rather than the ED.
Discrete choice experiments have been undertaken in the UK [
6] and Singapore [
7] to explore the factors influencing patient decisions to present to the ED. A systematic literature review also examined patient preferences for emergency care but most of the papers were from the US and UK [
8‐
10]. An Australian study examined patient preferences for emergency care but looked specifically at an older cohort of patients over 70 years of age [
11]. An Australian Institute of Health and Welfare (AIHW) report on the use of emergency departments found that the majority of people who visited the ED rather than the GP did so because they were taken to the ED by an ambulance, their condition was serious or they were referred by their GP [
12]. Wong and Hall found that patients with a better GP experience are less likely to visit the ED, suggesting that improving GP quality could reduce ED use [
13]. Similar issues were raised in the rural setting [
14]. Outside of these studies, however, the complex factors influencing patient decisions to present to ED rather than primary care have not been well explored. Morley et al. showed that chronic conditions are a driver for ED presentations but that local factors should also be examined [
15].
This study sought to understand patient preferences and care seeking behaviour for ED care for lower acuity presentations during GP opening hours by undertaking interviews with patients at two tertiary emergency departments. The aim of the study was to understand patient preferences for care so that models of care can be developed which best meet the needs of patients.
Methods
Study site
This was a qualitative study using semi-structured interviews with patients at two tertiary emergency departments in Sydney. Hospitals A and B are both level six tertiary hospitals, major trauma hospitals with ICU and cardiothoracic departments situated in metropolitan areas. Hospital A is in the inner city and Hospital B is in southwestern Sydney. Both sites were chosen as they have different demographic profiles [
16,
17] to ensure a range of views were represented.
Ethics was approved by Sydney local health district (X19–0228 and 2019/ETH10574). Specific applications (SSA) were approved at both sites before the study was carried out.
Study design
The questions in the interviews were informed by reviewing the literature on the topic and VK’s clinical practice and experience as a clinician working in the ED. The interviews began with demographic questions, then moved to patient factors and their time in Australia (if they were from overseas). Patients were asked if they used primary care in the community and their relationship with a GP. Patients were then asked why they chose the ED over the GP that day. They were also asked questions around accessing the ED and their experience of the ED and whether they considered any other options before presenting to the hospital.
The semi- structured interviews were carried out by VK (qualifications in economics, public health and medicine). The interviewer introduced themselves to the participants, explained the purpose of the study, answered any questions the participants had and obtained written consent. Interviews were continued until data saturation was achieved [
18,
19]. The COREQ checklist was used to report findings [
20].
Inclusion and exclusion criteria
All adult patients with a triage category of 3, 4 or 5 were included who were in the waiting room between 9 am and 5 pm on weekdays. Triage 1 is life threatening and needs to be seen immediately, triage 2 should be seen by a doctor within 10 minutes, triage 3 within 30 minutes, triage 4 within 60 minutes and triage 5 within 120 minutes [
21].
If patients did not speak English telephone interpreting services were available. If a telephone interpreter was used, this interview was recorded and then later the English translations transcribed as per the other interviews to ensure the inclusion of culturally and linguistically diverse populations. Patients were excluded if they were: paediatric (less than 18 years old), brought in by ambulance or referred by a GP. They were also excluded if they had suicidal ideation or presented with vaginal bleeding, as these were considered potentially sensitive presentations and it was thought to be inappropriate to approach patients.
Sampling
Participants were selected by convenience sampling with all patients who met the inclusion criteria during the time the researcher was conducting interviews, invited to participate. Interviews were carried out at the two sites in February and March 2022. All the interviews were carried out face to face.
Analysis
These interviews were recorded and then transcribed verbatim by a professional transcribing service. The interviews were checked by VK who listened to all the audio and crosschecked the text transcription. VK also kept field notes during the interview process and reflected on their experience as a clinician undertaking a qualitative study and how that may impact on interactions with patients, in a dynamic different to the one they were used to (as a researcher rather than a doctor in a patient interaction).
The study used an inductive approach [
22] and undertook thematic analysis. Six phases of thematic analysis were undertaken [
23] including familiarising oneself with the data, generating initial codes, searching for themes, reviewing themes, defining and naming themes and producing the report [
23]. Two coders (VK and KY) coded the data in Nvivo 12 Pro and another co-author (TL) was available for discussion if there were any differences that could not be resolved. Initially a subset of interviews was coded and the coders met to discuss their respective codes. They proceeded to code an additional 10 interviews and met again to discuss the coding framework. Once initial themes were captured and the coders had agreed, VK coded the remaining interviews using the developed framework. This was an iterative process and VK read and coded the transcripts multiple times, whilst reflecting on the individual interviews and notes from their field journal. ‘Peer-check in’ was carried out with an external academic who was not a part of this project but has extensive experience in thematic analysis. ‘Member check-in’ was not possible as the patients were from the waiting room in the two respective emergency departments and could not be contacted again.
Discussion
Four main themes were identified concerning why patients present to the ED during GP opening hours for low triage acuity presentations. Firstly, they are often referred by HealthDirect, work, friends or family. The second theme was ED factors, which included having the convenience of investigations in one location with immediate results. Another sub theme within ED factors was severity of their condition which they believed needed hospital rather than community care. The third theme identified from the interviews included GP factors such as not having a GP, not being able to secure an appointment or having a negative experience with a GP previously. The fourth theme identified personal factors linking patients to the respective hospitals.
The reasons for presenting to the ED for low triage acuity presentations during GP hours are multifactorial. ED factors, in particular the convenience of investigations, was a driving factor. Patients generally liked having a diagnosis at the point of care. They were also more likely to present to the ED if they thought that their symptoms needed hospital rather than community care, despite the lower triage. These concerns were often justified as a lower acuity triage may still need hospital treatment, for example, intravenous antibiotics for a spreading cellulitis infection.
Some patients presented because they were seeking reassurance, or sometimes a second opinion. NSW Health has attempted to address this through the ‘keep emergency departments for emergencies’ [
26] campaign. These are posters adopting a traffic light approach and encouraging patients to call HealthDirect in the first instance, go to a pharmacy for minor ailments, GP for more chronic issues and the ED for life threatening emergencies. There is scope to redirect patients to more appropriate care and it is clear from the interviews that HealthDirect is widely used. None of the patients interviewed mentioned that they considered pharmacy as an option prior to coming to the ED. However, several patients mentioned calling HealthDirect before coming to the ED.
Multiple factors drive patients towards the ED. A lack of understanding of the Australian health care system was identified. The patients interviewed represented a culturally and linguistically diverse cohort. This also reflects Australia’s multicultural population with 27.6% of the population being born overseas and 48.2% having a parent who is born overseas [
27]. Several of the patients who had a GP, and who spoke English as a second language, had a GP who spoke their primary language and the medical consultation with the GP was carried out in this primary language. Any policies which seek to link patients into primary care should consider the multicultural nature of Australia’s society and develop policies which are culturally inclusive and appropriate.
For those patients who are not confident navigating the Australian healthcare system, there is scope to provide public health messaging to improve health literacy. For example, this could include more messaging about the importance of primary care and of having a GP, and explaining the Australian healthcare system and tailoring the messages to people who speak English as a second language. The messaging about the Australian health care system could be tailored for international students and in immigration as people are entering the country.
Likewise, the pathway to linking patients to GPs could be more specific. HealthDirect offers several services including an online ‘symptom checker’, a telephone hotline with a triage nurse, and also a ‘GP finder’ page on the website. The ‘GP finder’ includes a postcode locator and some other variables such as whether a practice is bulk billing or has wheelchair access. Other variables could be included such as whether the GP is currently accepting new patients, or if the GP speaks other languages and can offer a consultation in another language, or whether they have a special interest in mental health or paediatrics. This would help patients find a GP which would ‘match’ their needs, and more easily link into community care.
However, to offer more GP services, medical workforce shortages need to be addressed. The supply of a medical workforce has been on the agenda for some time in Australia. There is a chronic shortage of general practitioners and this is not a new problem [
28]. Australia continues to rely on overseas trained medical doctors to fill deficits in the Australian health system [
29]. Apart from posing ethical issues around recruiting Australia’s medical workforce from neighbouring countries it highlights that workforce planning needs greater policy attention. In particular, GP shortages are particularly problematic for those living in rural areas [
30] but even in the inner-city area this was identified as an issue in the interviews.
These pressures on general practitioners results in time constraints and shorter consultations, which appears to drive patients to the ED. As one patient explained, she can wait an hour for the GP for a 5-minute consultation but prefers to wait 2 or 3 hours in the ED, be reviewed by a doctor for a longer time and have all her investigations at the point of care. GP workforce pressures appear to have a flow on effect in the ED, especially shorter consultations with the GP. This supports work by Wong and Hall that a negative experience with the GP can increase ED care seeking [
13].
Waiting time in the ED did not seem to be a deterrent for these patients seeking ED care. This is likely because patients expect to wait and had made the conscious decision to attend regardless of the expected wait time. Out of pocket costs did not seem to influence preference for the ED or care seeking behaviour either. All the patients in hospital B had bulk billing GPs and did not have out of pocket costs. The lower socio-economic area of hospital B may likely be driving billing practices.
Investing in the health workforce is one solution but has a long lead time. Other shorter-term measures could include focusing on policy and remunerating General Practitioners in line with their hospital colleagues. Another approach would be to hold a national ‘health summit’ to further explore how Medicare could be strengthened. In the shorter term, the Government has also committed to increasing the number of urgent care centres [
31], however it remains to be seen how these will be staffed. As articulated by Wilson et al., “a whole of system approach” is needed to realign policy priorities in this setting [
32].
It was interesting that patients felt such community ownership of both hospitals. They had overwhelming trust in the health system, and this was evident in the quotes which came from the interviews. Although it was not a direct question in the interviews, it was clear that the patients felt both pride and trust in both hospitals. This could be leveraged to provide more efficient services in the ED or branding new clinics to see low acuity patients as part of the hospital.
Strengths and limitations
A major strength of the research was the diversity of the group of patients that were interviewed in terms of gender, cultural background, language and educational qualifications. The patients were from two local health districts and represented a diverse group which added to the richness of the analysis. Another strength was the number of interviews and the robustness of the coding. Having two coders, from different professional backgrounds, offered a richer interpretation of the data.
A weakness of the study was that the interviews were relatively short in duration, with the meantime of the 44 interviews being 7 minutes. This was likely due to the straightforward nature of questions. We were able to achieve thematic saturation, increasing confidence that we managed to elicit the key factors driving the decisions of these patients. There were five patients who, during the course of the interview, stated that they were advised to attend the ED becuase they were referred by the GP, however this was not in the triage note. The interviewer made every effort to reassure patients that they did the right thing in presenting to the ED but the question itself may have made some people feel uncomfortable and some may have said that they were referred by the GP when they were not. Any patient who stated that they were referred by the GP was subsequently removed from the analysis. Also, although interpreting services were available, only one patient was interviewed with an interpreter. Despite this, the sample includes a culturally diverse cohort.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.