Understanding patients' health-seeking behaviour for non-emergency conditions: a qualitative study.

INTRODUCTION Overcrowding at Accident and Emergency (A&E) is a global public health issue.[1-3] A&E overcrowding has been associated with negative patient outcomes like increased mortality,[4] reduced quality of care[5] and increased medication errors.[6] Studies have highlighted the use of A&E by non-urgent patients as one of the contributing factors to A&E overcrowding.[7-9] Singapore’s A&E visits have grown about 4% annually between 2006 and 2015.[10] This increase is higher than the population growth rate, which is about 1%–2% annually.[11] In 2013, non-urgent visits contributed to >50% of A&E cases at four public hospitals.[12] Hence, there is an urgency to address the issue of non-urgent visits at Singapore’s A&E. Non-urgent patients are individuals whose conditions have low urgency and can be handled by other healthcare services such as general practitioners (GPs).[13] Previously, Singapore had implemented interventions to reduce non-urgent A&E visits with varying success.[14] Past education campaigns resulted in reductions, which were not sustained in the longer term. The redirection of patients to other healthcare services often resulted in disputes between A&E staff and patients. Increasing the A&E fees only translated into <10% drop in non-urgent A&E visits.[14] The limited success could be due to the lack of understanding of the health-seeking behaviour of patients with non-urgent conditions. International studies reported that non-urgent patients’ choice of healthcare providers was influenced by a few factors, including (a) perception of one’s condition’s criticality and the need to be reassured,[15-18] (b) preference and greater trust in A&E over other healthcare services,[15,19,20] (c) easier access to A&E relative to other services like GPs[21-23] and (d) financial considerations such as cost and insurance coverage.[15,24] Locally, studies targeting throughput factors such as queuing, scheduling, resource allocation or patient profiling for non-urgent A&E visits are far more pronounced,[25-27] and little work has been done to investigate the issue from a social–psychological angle. To address this gap, we aimed to explore the underlying factors influencing the health-seeking behaviour of patients for non-urgent conditions. METHODS Five focus group discussions (FGDs) were conducted in English by four facilitators in Changi General Hospital (CGH) between March and April 2016. The facilitators had no previous relationship with the participants. FGD was chosen, as the interactive discussions enable the generation of unique data that is not accessible through individual interviews.[28,29] We purposively sampled across two groups: (i) the A&E group and (ii) the GP group. The A&E group comprised patients who were self-referred or referred by their GP to A&E and not admitted. Individuals were approached at A&E while waiting for medication and payment after consultation with the doctor. Not admitting the patients after an emergency consultation suggests that they have non-urgent conditions that can be managed by a GP. Individuals referred to A&E by the GP were included in the sampling strategy to enable us to understand the reasons for the initial choice of GP over A&E. The GP group comprised specialist outpatient clinics (SOCs) patients who did not visit A&E but saw a GP within the last 3 months. We used a semi-structured, pretested topic guide for consistency. An exploratory approach involving experts’ inputs (e.g. A&E clinicians) was used to develop the guide, focusing on adaptive questions about the decision-making processes that prompted participants’ choice of healthcare providers. The inclusion criteria were individuals aged ≥21 years with no cognitive impairment and English-speaking patients. Each FGD lasted approximately an hour, and discussions were audio-recorded, transcribed verbatim, anonymised and analysed using NVivo for Mac, Version 11.0[30] (Lumivero, Denver, CO, USA) by the first author, based on Braun and Clarke’s thematic analysis (TA) approach.[31] Themes were derived from codes and conceptualised based on the patterns identified in the data. The FGDs and analysis occurred iteratively, and data and thematic saturation was reached by the fifth FGD. This study was approved by the SingHealth Centralised Institutional Review Board (reference no. 2016/2620) and the Singapore University of Technology and Design Institutional Review Board (reference no. 15-088). RESULTS Twenty individuals participated in five FGDs. Majority were male (65.0%), Chinese (55.0%), had at least a pre-university education level (70.0%) and resided in public flats (55.0%). We identified five main themes and 12 subthemes [Figure 1 and Table S1, see Supplemental Digital Appendix at https://links.lww.com/SGMJ/A21].Figure 1: Themes and sub-themes from the analysis.Theme 1: The need for continuity of care Sub-theme 1.1: Availability of past medical records Some participants felt that the availability of documented medical records and the health provider’s knowledge of their condition influenced them to choose GPs instead of A&E. Sub-theme 1.2: The type of doctor–patient relationships Some participants chose GPs because they had a positive doctor–patient relationship and valued the familiarity with their regular doctor. Some of them highlighted that they were likely to seek treatment from a healthcare provider with whom they had previous experience. “I went to my nearest uh GP, is also my family doctor because I know she have the records of my family including my children so in terms of bonding wise, patient and doctor. [The GP] knows what happening too. She got the case so I prefer uh GP rather than to A&E.”—(ID35) Theme 2: Patient-perceived quality of care Sub-theme 2.1: Presence of specialists and medical expertise at A&E to handle various medical conditions Most participants expressed confidence in A&E as a place with the medical expertise to handle a variety of medical cases. “Because they [A&E] have a lot of experience I assumed, all sorts of situations? I meant most, I would assume that most GPs mostly see the standard flu and cough. But if it is something more out of the ordinary, then I would trust A&E rather than the GP.”—(ID15) Sub-theme 2.2: Comprehensive medical facilities and expertise provided in one place The availability of onsite ancillary services (e.g. X-rays, blood tests) offered participants a sense of assurance that they would be able to receive the required quality of care for their condition. They regarded the lack of such ancillary facilities in the GP clinic, coupled with the unavailability of same-day test results, as reasons for choosing A&E over GPs. “The reason I would go to A&E rather than a GP is because uh, I think the A&E can give me my blood test results, X-Ray, everything. It’s like a one-stop thing. As for GP, you need to wait for a few days, at least one day you know, before you know what’s happening in the blood test.”—(ID23) Theme 3: Patient-perceived severity of medical condition Sub-theme 3.1: Perceived severity of condition influences the health-seeking behaviour. Participants explained that they would head straight to A&E if they perceived their conditions to be severe. Sub-theme 3.2: Urgency to seek medical assurance and relief from pain and discomfort Participants expressed the urgency for immediate medical attention to allay their concerns over the uncertainty of their condition’s severity and to be relieved from pain and discomfort as soon as possible. “If I deem my condition requires immediate attention and I want quick, quick answers you know. Because the main worry is you want to know what’s the cause of it, is it something very life threatening or I mean you get an assurance lah, and answers straightaway. I will go to A&E.”—(ID31) Sub-theme 3.3: Reliance on different information sources for judgement of severity Participants were aware that A&E is meant for urgent conditions and reported seeking health information from multiple sources, such as friends and family members, printed materials, personal experiences and the internet, to determine their condition’s severity and the choice of healthcare provider. “Based on the situation and based on our life experience. And our knowledge, judge. So the two examples I have given for myself just now, the hernia ok, is not so important, must have it done immediately. Unless great pain. Burst! But I know my situation not, so is not necessary to go to A&E. Alright, the hand so swollen and getting worse, I think I have to go to A&E, especially the finger is already twisted.”—(ID11) Theme 4: Minimising out-of-pocket (OOP) cost Sub-theme 4.1: Various forms of medical coverage that improves healthcare affordability Participants mentioned that their entitlement to company medical benefits, insurance coverage or financial assistance enticed them to seek care at providers where they could claim for their medical bills and save costs. “…the MSF (Ministry of Social and Family Development) helping me. They give me a card and say that in this 6 months, I don’t have to pay anything if I go for medical lah. Then in that case I go to A&E… then I can get the free medicine lah.”—(ID22) Sub-theme 4.2: Costs incurred from seeking treatment at GPs Participants expressed the fear of being charged twice if they were referred to A&E after visiting a GP. Moreover, participants felt that GP fees are costly, especially after-hours, and any additional treatment or tests ordered would charged separately. “Like at night, you go to the 24 hours GP clinic, you pay even more and after that they say you have to go to the A&E immediately. So with the letter you go the A&E and you pay another set of fees.”—(ID23) Sub-theme 4.3: Willingness to endure longer wait times for lower OOP costs Some participants preferred lower OOP costs even if they had to endure longer waiting times. “Hmm… For me ah, lower cost. [Interviewer: So it doesn’t matter how long you have to wait] No.”—(ID13) Theme 5: Minimising time to access care Sub-theme 5.1: Minimising waiting time to medical consultations Participants highlighted a need to be attended to by the respective healthcare providers in the shortest time possible. Individuals who chose to consult GPs did so due to perceived shorter waiting. “I don’t like to wait uh because when I sick and I have to wait for hours at A&E, I rather go to the GP and get information and they straight away tell me what I am suffering from.”—(ID64) Sub-theme 5.2: Time wasted before receiving appropriate care Primary care visits which preceded referrals to A&E were deemed as time consuming by most participants, and they disliked the idea of navigating through channels like hospitals’ hotlines for information on where to seek treatment. “And if I go [primary care] and I wait and I, I… in the end I still get referred to the A&E, I must (might as) well go straight to A&E.”—(ID31) DISCUSSION The analysis of the FDGs revealed five themes: the need for continuity of care; patient-perceived quality of care; patient-perceived severity of medical condition; minimising OOP cost; and minimising time to access care. Continuity of care is defined as the patient’s experience of a smooth and coherent progression of care.[32-35] Typically, this requires a relationship between providers and patients that is based on trust and familiarity.[33,34] The findings of our study were consistent with those published in literature, whereby a positive physician–patient relationship is reported as one of the factors influencing the choice of GPs over A&E.[36-39] Also, previous positive experiences were positively associated with future choices of the same type of provider.[40,41] Hence, building a positive relationship between patients and their GPs could encourage patients to first seek treatment from GPs for non-urgent conditions. Patients who perceived their conditions to be severe preferred to visit A&E, which was deemed to provide better reassurance.[38,39,42-47] Nevertheless, there is evidence noting that patients were not always able to accurately assess their condition’s severity,[38,46-48] highlighting a potential need for interventions. Previous educational campaigns in Singapore resulted in a decline of up to 67.3% for non-urgent A&E visits;[14] however, this effect wore off subsequently.[14] To mitigate this tapering effect, tapping on GP staff for word-of-mouth dissemination, launching regular health literacy and publicity awareness campaigns and the use of digital outreach methods[49] for IT-savvy adults could be considered. Better onsite ancillary services and medical specialists at A&E and minimising OOP cost and time to access care have been regarded as key contributing factors for non-urgent A&E visits.[20,21,37,39,43,50-52] Northington et al. observed that non-urgent A&E visits were driven by a belief that A&Es are better equipped to provide better quality of care compared to GPs.[20] Individuals who were financially insured for A&E visits contributed to increased non-urgent A&E usage.[45,53] In some cases, A&E was preferred over primary care providers, as no upfront payment was required.[39] Thus, policymakers could consider improving access to ancillary services and redesigning the financial coverage structure for GPs to incentivise visits to GPs first. Long wait times for GP appointments was a contributing factor for A&E visits in other studies.[21,39,54-56] We posit that this will be absent from our findings, as these studies were conducted in countries where the primary care system operates on a by-appointment-only model. Singapore’s primary care facilities accept walk-ins,[57] thus removing the need to wait for an appointment. Despite a longer waiting time at A&E, some participants still preferred to proceed to A&E first, as they perceived it to be of better value in terms of convenience and cost-saving. This is a unique finding that was not mentioned by other qualitative studies.[15,37,38] Hence, to further examine such trade-offs, a discrete choice study could be designed to explore and quantify how certain attributes are valued over others. This study is not without its limitations. We did not recruit participants who were non-English speaking. This may limit the ethnic-specific perspectives that could potentially surface during FGDs. Bias arising from analysis of the data is possible, as the first author was the only coder and no inter-rater coding was performed; also, member checking[58] was not performed as we did not have access to participants’ contact details. Therefore, the generalisation of results to other groups should be done with caution. In conclusion, the present study showed that multiple reasons influenced an individual’s choice of healthcare provider. Therefore, a multipronged approach involving tailored regular outreaches, enhancement of the patient–GP relationship and provision of financial coverage might be effective in reducing A&E non-urgent cases in a multi-ethnic Asian population. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.