Why seek a second consultation at an emergency centre? A qualitative study

Background The inappropriate use of emergency centres (ECs) is an expanding problem globally. The high attendance of non-urgent return presentations to ECs is recognised as part of the problem, placing an unnecessary demand on limited staff and resources. Of unscheduled returns 34% of cases had no change to diagnosis or treatment with the conclusion that 80% of re-attendance could be attributed to deficiencies in the initial consultation. This study aimed to evaluate the reasons why patients sought an early second consultation for the same complaint at a hospital EC in South Africa, by exploring the patient’s experience and shortcomings in the first consultation. Method A qualitative study was conducted using in-depth, semi-structured interviews with 20 purposively selected participants who presented to a rural regional provincial hospital’s EC within 7 days of a prior consultation for the same complaint. Verbatim transcripts were analysed using the framework method. Results The main reasons for a second consultation were symptom related factors and the need for diagnostic certainty. The major themes around patient experience of the initial consultation were shortcomings in effective evaluation and management of pain, diagnostic uncertainty including poor examination, poor explanation, uncertain access and follow-up and societal encouragement to utilise a hospital EC. Conclusion Further interventions should explore pain as a presenting symptom of the illness experience, and promote competence in addressing physical and psychological causative factors within a patient-centred approach for all health staff, especially in primary care services.

consultation. 14,15 Dissatisfaction is a multi-factorial concept involving elements of too little consultation time, the need for better explanation, communication and shared decisionmaking. 14 Patient perceptions have been shown to predict outcomes of consumer satisfaction and are recognised as an acceptable measurement of service quality. 16,17 South Africa is a resource-poor, developing country, where overcrowding of public ECs is commonly experienced, with critical consequences. 5,18 The four-fold burden of disease, which includes HIV and TB, non-communicable diseases, violence and maternal and child illnesses, compounded by mental illness and substance abuse, often overwhelms the public primary health care service and spills over into hospital ECs. 19 The majority of South Africans do not have health insurance and make use of the public health system, where primary health care is free and nurse-driven. Many of these patients also pay out of pocket to utilise the private health system. About 20% of South Africans do have health insurance and make use of the private health system of general practitioners (GPs) and specialists.
The National Health Act of 2003 emphasises person-centred health care, while the Western Cape 2030 Health Plan's vision is person-centred quality care with the focus on patient experience of health care as first principle. 20,21 Previous work performed in Eden showed that 65.0% of patients presenting after hours to the hospital EC had routine non-urgent complaints, of which 47.0% required primary level care. 18 Becker found that 88.9% of George hospital EC users were selfreferred of which only 4.9% were considered appropriate for EC care. 5 Understanding patients' reasons to seek a second consultation and exploring their experience of the initial consultation are fundamental to address unscheduled returns.
The aim of this study was to explore patients' experience of their first consultation and any shortcomings that led to seeking an early second consultation at a rural regional hospital EC. We were specifically interested in the main reason for the second consultation, the socio demographic, logistical and personal attributes of the patients seeking a second opinion and the possible shortcomings of the first consultation.

Study design
This was a qualitative study using in-depth, semi-structured interviews. Qualitative research was chosen as this was deemed the best way to obtain a deep understanding of patients' experience of the initial consultation and reasons for seeking a second consultation. 16,17

Study setting
The study was conducted in the EC of George provincial hospital, a rural regional state hospital in the Eden district of the Western Cape province of South Africa. It is the only public regional hospital in the Eden and Central Karoo districts and provides a higher level of care for 10 surrounding primary health care clinics (PHCs) and 10 district hospitals. The EC attends to approximately 3400 patients monthly. No referral is necessary to access the emergency services and being the only public health facility open after hours in the George sub-district, all patients, including those with less urgent complaints, utilise this service.

Sampling and data collection
Study participants were purposively selected from all adult (18 years of age or older) patients presenting to the EC during working hours between 08:00 and 16:00, during the month of October 2014. 22 Inclusion criteria were: adults, 18 years and older, who presented within 7 days of a prior consultation for the same complaint at any health facility, who was not referred or scheduled for a follow-up, who could converse comfortably in Afrikaans or English and who was not clinically unstable. Prior consultations included a consultation at a primary health care facility, private GP or any EC. The mean time period for returns for a related complaint has been found to be 4.5 days, and 7 days is advocated to capture all data of related returns. 8 The researcher approached patients during their triage procedure in the EC with the question, 'Have you visited a health care practitioner within the last seven days for the same complaint?' Eligible patients were consulted in private and the aim and objectives of the study explained to them. Patients who agreed to participate proceeded to sign an informed consent form. In-depth, semi-structured interviews, as well as demographic questionnaires, were conducted with each participant in a private room within the EC. Twenty-one participants were interviewed by the researcher, with interviews lasting up to 25 min each. An interview guide was used and volunteered themes were further explored. The interviews were audio taped and field notes were made by the researcher. One interview was not included as the interview was discontinued as a result of language barriers.
Continued discussion with a senior co-researcher during the interview phase led to revision of the interview guide, formulating extended questions to further explore prominent themes and research objectives. The interviews stopped when a point of information saturation was reached. Verbatim transcription of the audio recordings was performed.

Analysis
Twenty audio recordings, verbatim transcriptions, demographic questionnaires and the reflective field notes of the researcher were analysed using the framework method. 23,24 After familiarisation, themes were inductively identified. Thereafter an index of numerical codes was assigned to the themes in the transcripts, and finally a chart of major and minor themes was developed, with interpretation of contradictions and associations between major and minor themes explored. 24 To improve credibility, the data from the interviews were triangulated with findings in the literature and by involving a co-researcher in the analysis. The discussion of themes and interpretation of data between two independent researchers helped to limit bias. Confirmability and objectivity was sought by inclusion of a reflexive paragraph.

The role of the researcher
Personal reflexivity of the self as research tool to acknowledge the impact of her education, Caucasian heritage, privileged upbringing and lived experience as a doctor working in the EC of George hospital needed to be explored to caution against pre-judgemental assumption regarding the study question and objectives. The researcher attempted to limit biases by carefully constructing openended questions in the interview guide, together with the co-researcher who encouraged spontaneous discussion of topics and themes. She was aware that the doctor title and being a female, White student from the university tended to position her in a 'patriarchal' role relative to the participants. Although introduced by her first name with the explanation that she is a postgraduate student who used to work in the EC before, it was clear that the participants still associated her with the institution and the larger health system. The researcher had to consciously refrain from the directive, closed question interview style taught during undergraduate studies. She was taught to be a reductionist detective of symptoms and signs and to focus on the search for the most likely diagnosis, as opposed to a more constructivist approach of allowing for 'new truths' to emerge from another's world view.
Interpersonal reflexivity was sought in compiling demographic outlines of the participants and exploring cues from open-ended answers to better understand their lived experiences. The researcher found it difficult to put down the doctor's detective magnifying glass in search of presumed theories as stated in previous literature. As she matured in her interview style and reassessed what she brought to the equation, she learnt how to allow the participants to truthfully explain their experiences. Soon it became more natural to walk with them through their memories and try to observe their reality. The search for the assumed correct answers faded and she understood that she needed to listen to the spoken and unspoken words to make sense of their lived experiences. Being able to converse with them in Afrikaans and English improved the limitations as would have been set by using an interpreter. The researcher kept a reflexive journal of all interviews to alert herself to her own feelings, emotions and concerns during the interviews, and discussed these with the co-researcher, to allow for perspective.

Ethical consideration
Ethics approval was obtained from the University of Stellenbosch Higher Research Ethics Committee (Nr S13/07/126), the Western Cape provincial health department and George Provincial Hospital.

Results
Twenty participants, 9 female and 11 male, between the ages of 20 and 63 years (mean 42.6 years), fulfilled the inclusion criteria and participated. Seventy per cent of the participants' first language was Afrikaans, 25% isiXhosa and 5% English, which is a true reflection of the Eden district demographics (70.8% Afrikaans, 18.3% isiXhosa, 7.5% English). 25 Thirty-five per cent of participants relied on public transport to access health care. Two participants bypassed the primary health care services by utilising ambulance transport to the hospital EC. Eleven participants' first consultations were at their local PHC, seven presented themselves to the EC after a prior visit and only two participants sought medical care at a private GP. The distance from home to the EC ranged from 2.5 km to 23.0 km.
Six major themes with closely connected minor themes emerged from the interviews ( Table 1).

I expect the pain to be stilled
The majority of participants reflected that ineffective pain relief was the single most important factor motivating a second consultation. The pursuit of adequate pain management drove them to the hospital EC. ' Closely linked with the expectation of effective pain relief was the minor theme of general dissatisfaction with the receiving of 'only pills' for the pain. Participants were looking for a diagnosis with an effective management plan and expected more than merely the routine prescription of pills that they received at their initial consultation.
'I thought that they would do further investigations to look where the pain is coming from, but nothing of that.' (Participant 12, male, 27 years) 'I think they could have given me an injection.' (Participant 3, female, 53 years) They perceived this as an inability of health care workers to adequately diagnose and manage their pain, which led to distrust and a second consultation. This emerged from participants whose first consultations were at the primary health clinic, hospital EC and private GPs. This was influenced further by financial limitations of the participants which played a large role in access to transport and seeking health care. Proximity to a health establishment determined access as explained by participant 20: 'I just go to the place that I'm close to'.
Transportation is costly and some participants utilised the ambulance services as free transport: 'Transport is expensive for us, we don't have the money to pay transport … so I decided we must try the ambulance again.' (Participant 11, male, 39 years) The uncertainty of being helped at the clinic made some participants access a private GP, but still was left dissatisfied, and because of financial constraints were forced to access further public health care.
'So I kind of need someone to sort it out and clearly my GP couldn't do it because he needs more and that means more money and I can't afford that right now.' (Participant 19, female, 38 years)

They told me I must rather go to the hospital
Family members, friends, employers and even health care providers tended to show more confidence in health care access and management at the hospital EC as opposed to public or private PHC. Access of care was understood and defined from the patient's experience, and not from an understanding of the health system structure around appointments, triage criteria and referral pathways.
In contrast, this regular user of her PHC clinic and the hospital EC, had the following viewpoint:

Discussion
The main reasons for seeking a second consultation were persistence of symptoms, particularly pain, closely associated with the need to understand the cause of the pain. This is in keeping with previous research as the main reason for unscheduled return visits to the emergency department being symptom related with inadequate pain management proven to be a significant problem. 1,7,8,9 Milbrett identified that six of seven major reasons to visit an EC as having some relationship to pain, and White described pain-related diagnoses to be a major reason for unscheduled returns. 2,4 Overdramatisation of pain, sometimes to the extreme of fear of death, raises the question of how well health care workers are addressing the presumed underlying physical or psychosocial contributing factors, and how this can be managed better.
The reasons for a second consultation for the same complaint at the EC can be understood as issues relating to the patient, the first consultation, the health system and society. These issues cannot be analysed as separate entities but should rather be seen as a ripple effect starting at the patient, constantly interlinking with broader issues (Figure 1).
Patients, at the centre of the consultation, had preconceived perceptions about the severity of their illness, their preference for health access and possible treatment outcomes. Patients often viewed their complaints as more serious compared to the perceived opinions of health care workers, similar to findings in other settings. 3 These perceptions were influenced by their beliefs, fears and prior experiences, as well as those of family and community members. 26 Kleinman makes a clear distinction between 'disease', which is the usual focus of the health care workers, and 'illness', which is the patient's experience of their symptoms. Herein lies a big assumption, namely that patient and health care worker find common ground during the consultation. We found the patients' illness experiences were strongly influenced by the initial consultation. The lack of diagnostic certainty and a clear explanation were the biggest shortcomings in the initial consultation and served as internal motivation to seek a second consultation. 13,14,15 The urge to understand and take more ownership of their diseases and illness behaviour was an important missed opportunity in the first consultation.
It has been shown that patients want a patient-centred approach. They expect health care practitioners to explore their main reason for the visit, their concerns, seeking an integrated understanding of the whole person and find common ground on what the problem is and mutually agree on management. 27 Stewart, widely regarded as the pioneer of patient-centred care, agrees that patients should be the ones to define patient-centred care. 16 From the health system point of view, it has been shown that patient-centred care is associated with reduced health care utilisation, reduced referral for special investigations and improved health outcomes. 28,29 Family medicine worldwide and also in South Africa emphasises the importance of the bio-psycho-social approach to patient care in order to understand the patient's illness experience, which has proven to be essential for accurate diagnosis, health outcomes and humane care. 30,31 This is supported by the HealthCare 2030 Plan of the Western Cape Department of Health, which lists person-centred quality of care as the first principle in their vision statement. 21 The holistic approach to the patient consultation incorporated into a shared management plan to improve health outcomes is stressed in policy but is not easily practised. This may be a big reason for many repeat consultations and failed patient expectations.
Access to emergency health care is a constitutional right in South Africa and health care providers or establishments may not refuse a person emergency care. 32 This encourages  and almost entrenches a sense of entitlement in the community to present to the hospital EC, especially if there is a perception that their symptoms are severe and they may not be sufficiently managed at a primary care facility. 5 Paradoxically, we found that patients felt disempowered to access care as they expected in terms of investigations and a comprehensive, shared management plan. Participants had to 'find the way' of least resistance and personal cost for their expectations to be met. This translated into phoning an ambulance and going to the facility with the perceived least risk of failed expectations, which usually meant the EC.
The South African public primary health care service was designed to be run by clinical nurse practitioners as the first contact health care providers. This is often contradicted by the way society understands it, for example advocating to 'see a doctor' in case of illness. Expecting a doctor as the preferred health care provider, together with the consumerist world view of the need for special investigations to confirm a diagnosis, led to a negative appraisal of participants' first consultation. 6 We found the primary health care providers used the lack of resources as a reason to justify their diagnostic uncertainty. This reduced the credibility of primary care's clinical diagnostic skills and a loss of confidence in primary health care in the minds of the patients. 3 Dell's study conducted at George hospital resonated this loss of confidence in PHC by stating the three main reasons for patients to attend the EC to be that the clinical treatment was not helping, they perceive hospital treatment to be superior and the lack of PHC services after hours. 5 Being told to come back 'tomorrow' or 'another day' added to the participants' uncertainty and was not perceived as sufficient safety netting in a comprehensive management plan.
The reasoning behind the pursuit of health care in most of the interviews was usually not a logical consideration of the health system and its appropriate entry point for a specific clinical complaint, but rather a response to a 'socially shared custom', as described by Beache and Guell. 3 Health system factors enhanced and encouraged this custom. The habitual use of the hospital EC is not only appealing because of convenience, accessibility, staff and resource availability, but especially because of the likelihood of being attended to by a doctor, with the expectations of an examination and access to some investigations. 3,5 In order for the primary health care service and the primary care component of hospital ECs to meet the expectations of patients and fully implement the holistic patient-centred model that is envisaged, it is important that consultations follow the bio-psych-social approach, together with a physical examination and a shared decision plan that clearly addresses the patient's complaints, especially those relating to the experience of pain. Without this, patients' perceptions and expectations of primary care will continue to be disappointed, with the default of access to hospital as perceived higher quality of care. 6

Conclusion
The aim of this study was to explore patients' experience of their first consultation and any shortcomings that led to seeking an early second consultation at a regional hospital EC. The main themes included shortcomings in effective evaluation and management of pain, diagnostic uncertainty including poor explanation, poor examination, uncertain access, uncertain follow-up and societal encouragement to utilise the hospital EC.
Areas of possible intervention include exploring pain as a presenting symptom of the illness experience, and promoting competence in addressing physical and psychological causative factors within a patient-centred approach for all health staff, especially primary care services.