Article Text

Understanding patient experience of out-of-hours general practitioner services in South Wales: a qualitative study
  1. J N Egbunike1,
  2. C Shaw2,
  3. S Bale3,
  4. G Elwyn1,
  5. A Edwards1
  1. 1
    Department of Primary Care and Public Health, School of Medicine, Cardiff University, Cardiff, UK
  2. 2
    Department of Care Sciences, Faculty of Health, Sports and Science, University of Glamorgan, Pontypridd, UK
  3. 3
    Gwent Healthcare NHS Trust, Llanfrechfa Grange Hospital, Cwmbran, Torfaen, UK
  1. Professor A Edwards, Department of Primary Care and Public Health, School of Medicine, Cardiff University, 2nd floor, Neuadd Meirionnydd, Heath Park, Cardiff CF14 4YS, UK; edwardsag{at}


Background: In the light of recent changes in the structure and provision of out-of-hours service in the UK, there is a need to re-assess the quality of care. One way to assess the quality of care is through patient experience.

Objectives: This study aimed to explore patient expectations and help-seeking behaviour, in order to understand their relationship with satisfaction and experience of out-of-hours care.

Methods: 30 semistructured telephone interviews were carried out with users of the general practitioner out-of-hours service in Gwent, South Wales. The interviews explored users’ experiences of using the service. A thematic analysis of transcripts was carried out using NUD*IST software. Comparison of data within and across codes facilitated the identification of explanatory constructs. Double coding of a sample of transcripts and discussion of emerging themes by members of the research team ensured the reliability of findings.

Results: Most (n  =  25, 83%) respondents reported satisfaction with the overall service received but a few (n  =  5, 17%) were dissatisfied. Patients generally had specific expectations of their consultation and there was a mismatch between patients’ expectations of the service and what the service actually provides in some specific user groups. Unmet expectations resulted in subsequent, and in some cases, multiple consultations.

Conclusions: Users’ views and expectations may be used to inform service design and improve services, but the data also indicated a need to address user expectations of services, for example by enhanced information provision. Any such interventions to improve patient experience of out-of-hours care will need to be evaluated.

Statistics from

“Out-of-hours service” in primary care is a term used to denote health services provided outside the normal general practitioner (GP) working hours. In the UK, there have been changes in the structure and provision of out-of-hours services since the 2003 “new contract” for GPs.1 Out-of-hours is defined as between 18:30 and 08:00 hours for weekdays, and all day at the weekend. Current out-of-hours healthcare services are intended to offer simple, convenient access to services.2 Changes to service provision are anticipated, however, such as changing skill mix between the numbers of GPs and nurses offering frontline care. It is important to assess the effects of these and other changes on the quality of care provided.

One way to assess the quality of care is through consideration of patient experience. Previously, patient experience has been given low priority and has been routinely measured only in terms of a simple satisfaction score. More recently, the importance of satisfaction as an outcome measure in determining the quality of healthcare has gained support and its measurement is being encouraged3 in ways that focus on achieving a patient-centred service. Consumers’ attitudes towards healthcare are also important in evaluating health services.4 Both research and audit projects have examined user satisfaction with out-of-hours care, with findings that most (approximately 80%) users were satisfied, but a significant minority were dissatisfied with the service provision.46 The more recent audit project found that 20% of users who were dissatisfied with the quality of care were specifically concerned about waiting times for consultation, continuity of care, management of information and integration with other services.6 Conversely, however, other studies have found such issues (eg, information transfer) to be well managed by services.2 The communication skills of clinicians have also been found to be variable—sometimes associated with positive experiences2 and elsewhere being poor.7 The specific context for some studies (eg, palliative care),2 makes it difficult to generalise about their applicability to other user groups.

This project therefore aimed to explore users’ experiences of the GP out-of-hours service in Gwent, South Wales, and to establish user preferences for service development. It sought to explore patient expectations and help-seeking behaviour in detail, to obtain a better understanding of the relationship between these aspects and satisfaction with their experience of out-of-hours care.


Study area

The service had six treatment centres at the time of the study, located across Gwent, an area with a mixed geography of large urban populations in the south, large post-industrial valley towns in the north and west, and rural more affluent areas in the east. Pockets in Caerphilly and areas in Blaenau Gwent have some of the worst levels of social deprivation, morbidity and mortality in the United Kingdom. Since October 2004, Gwent Health Care NHS Trust has been the provider of the GP out-of-hours service, previously provided by two separate cooperatives and one private company across the whole of Gwent. The population served is approximately 630 000.

Service structure

Individuals seeking out-of-hours GP advice or healthcare call into the service centre staffed with trained non-medical call handlers who collect patients’ details and prioritise calls based on the symptoms described. Calls are then returned by doctors or nurses who may be based either at the call centre or in the treatment centre (see fig 1). At most times, a nurse-led triage takes place and calls may be re-prioritised. At this point the triage clinician decides what type of consultation the patient should receive: telephone advice, a treatment centre visit or a home visit.

Figure 1 A flow chart of the out-of-hours service process.


A series of in-depth, semistructured interviews with users was undertaken. This gives a focused approach to the interviews, allowing the interviewer to explore potential areas of interest further. Individuals who had contacted the service on two randomly selected weeks in July and August 2005 (chosen to be as soon as possible after the commencement of the service when the study was operational) were identified from its database. As sampling occurred during the summer when service use was relatively low, participants’ experiences may be different from times when there are increased demands on services and such experiences may not have been captured in the present study. A total of 3376 (1524 in the first and 1852 in the second) users had contacted the service in the 2 weeks, of which 12% (n  =  406) were home visits, 33% (n  =  1114) telephone advice and 44% (n  =  1485) treatment centre consultations; 11% (n  =  371) categorised as information only were excluded. The sample was screened to exclude all deaths and those known to be unable to participate due to cognitive impairment or terminal illness. Letters of invitation and details of the study were sent by the service provider with consent for participation returned to the research team. Patients between the ages of 11 and 15 years were excluded for confidentiality reasons. We aimed to include representatives from all age groups and type of service/consultation received. Purposive sampling8 9 was used to include invitations to: patients seen or advised by a doctor or nurse; both young (<11 years) and adult (>15 years); both male and female; individuals with or without significant co-morbidity; individuals with different modes of consultations, ie, telephone advice, home visits or base visits at the six treatment centres in Gwent.

Invitations were sent to 221 individuals, with reminders sent at 2 weeks to non-responders. For users under 11 years of age, invitations were addressed to their parents or guardians. Further consent was obtained from each respondent at the end of each session for their interview to be used for the study. Data collection and analysis was carried out concurrently so that participants could be selected purposively from those who agreed to take part, based on emerging findings, in order to allow the final sample size to be determined by data saturation.

Data collection

A flow chart of the service process was mapped out as shown in fig 1.

From the chart and literature on patient satisfaction with services, the research team (CS, JNE and AE) developed an interview guide to address the different pathways through the service. The guide consisted of a series of open-ended questions with prompts to encourage respondents to develop their accounts and present their perspectives as appropriate, consistent with standard qualitative methods.10 One question, however, asking about overall satisfaction with the service, took a closed format and was asked consistently of all participants. The interview schedule is included as a supplemental appendix published online only.

The telephone interview explored people’s experiences of using the out-of-hours service, including their evaluations of the staff, availability of services required and the outcome of the episode. Comments were sought on the attitude/manner of the individual clinician they consulted primarily in the episode and suggestions for improvements to the service.

The interviews lasted approximately 30 minutes and were simultaneously audiotaped for transcription and analysis. Transcription was carried out after each interview and proof read by the researcher (JNE) carrying out the interview to ensure the reliability of the transcripts. Sampling and analysis were carried out concurrently so that sampling was terminated on data saturation,8 when no new themes were emerging.

Box 1 Patient preference for the out-of-hours service

1. User no 30: preference for the GP out-of-hours service over NHS Direct:

“I don’t think NHS Direct necessarily gives you the right advice. I have rung them with my son when he’s had asthma and they just turn around and say I think we’d better phone an ambulance and they phone an ambulance whereas if I can take him to the out-of-hours service and he goes on the nebuliser, more often than not, he can come home again. We’ve often found with them, that they call an ambulance and we’ve ended up spending a night at the hospital. I found the out-of-hours service much better than that”.

Data analysis

Interview transcripts were imported into NUD*IST software for thematic analysis. The texts were read and re-read to identify broad category codes based on the aims of the research study and on themes emerging from the data. The main category codes were broken down further into subcodes, creating a hierarchical coding frame. Comparisons of data were made within codes and also across codes and cases in line with a constant comparison technique,11 in order to explore the data in more depth and identify explanatory constructs. The research team consisted of experts (both medical and non-medical) within the field of primary care service provision and evaluation. Double coding of a sample of transcripts by members of the research team (JNE, CS) and discussion of the emerging themes ensured reliability.

Frequency data are presented for the question concerning overall satisfaction with the service.


A total of 221 users were invited to take part in telephone interviews. Those invited for interview were representative of the characteristics identified in the purposive sampling strategy and include a range of users as shown in table 1.

Table 1 Characteristics of the initial sample

Those seen or advised by district nurses include individuals with significant co-morbidity. By the close of recruitment, there was a response rate of 26%. Of the 58 who had responded, 35 (60% of respondents) consented to be interviewed. Twenty-three (40%) people declined for health or other reasons. A total of 30 interviews was carried out before the point of data saturation was reached. Those interviewed were representative of the characteristics identified in the purposive sampling strategy and include a range of users as shown in table 2.

Table 2 Characteristics of study participants (sample interviewed)

Overall, 17% (n  =  5) of the sample had been advised by nurses (including also hospice/district nurses), the remainder by doctors. Their ages ranged from 9 months to 91 years. With the exception of age ranges, the characteristics of the sample broadly reflect the demographics and service provision of the service in general, as well as reflecting the characteristics likely to influence views of the service.

Several themes were identified from the data; they are shown in fig 2.

Figure 2 Flow chart of themes identified from the data.

We present descriptions and data from these themes, with respondents identified by a “user number” from the series of 30 interviews.

Box 2 Patient expectations of the consultation

1. User no 17: woman, 52 years old, felt ill at work, received telephone advice

“… I had been quite worried because it was such a nasty experience that I’d had. I would have preferred it if he’d said come down and I’ll examine you”.

2. User no 3: mother of 9-month-old girl, treatment centre visit

“… I was really worried, she wasn’t herself, she was really whingey, so we took her up to the doctors and they wouldn’t give anything for her eyes. He didn’t even tell us what this heat rash thing was on her body, and he said it was eczema on her neck ….”

Overall satisfaction

Most respondents (n  =  25, 83%) reported satisfaction and only a few (n  =  5, 17%) reported dissatisfaction with the overall service received. Timeliness, perceived quality of the service and matching user expectations were the principal factors influencing overall reported satisfaction. Preference for the GP out-of-hours service was expressed when other services such as NHS Direct, A&E and their named GPs were felt to be unnecessarily time consuming in comparison (see box 1).

Reasons for dissatisfaction

Themes emerging from the data that were associated with dissatisfaction were as follows.

Expectations of waiting

Generally, respondents reported a maximum wait of approximately 20 minutes for the first call back from a clinician, although those who had telephone advice or a home visit could wait for an hour or more. Those seen at the treatment centre reported shorter waits of 20 minutes or less, even when attending the centre without an appointment. Despite the shorter waits at the treatment centre, there appeared to be an expectation that this was an “instantly available” type of service. Below are excerpts from an interview with user no 3: mother of a 9-month-old girl who visited the treatment centre without an appointment:

Interviewer: Did you have to ring into the out-of-hours service call centre?

Mother: No we didn’t, we just went up there. We usually do, but we just went up there.

Interviewer: How long did you have to wait to see the doctor?

Mother: “About 10–15 minutes but he wasn’t doing anything. There was no one in with him”.

Unmet expectations

Before contacting the service most individuals appeared to have an expectation of the type of consultation or level of care that they sought, and dissatisfaction was associated with a mismatch between expectations and service received, eg, patients who wanted to be seen and were given telephone advice only (see box 2). This is likely to have influenced the subsequent use of services (see table 3).

Table 3 Further consultations and reasons for consultations

Unmet expectations of the service and consultation were a particular issue for the following three groups of respondents. They include:

Patients with no previous experience of particular illnesses

Respondents in this group appeared more worried and anxious during waiting periods, as shown by the following excerpt:

Interviewer: Can you tell me how long it was from the time you made the first call to when you actually spoke to the doctor?

Respondent: It had to be 25 minutes. It must have been that length of time, because I was beginning to panic and wonder what to do with myself. I know now what to do; to just keep drinking. (user no 18)

Mothers with young children under 5 years old

Mothers with young children reported anxiety during waits and wanted more time during consultations regardless of whether or not their children had previously had a particular illness. For example, one mother of an 11-month-old boy commented:

“… I am probably a slightly neurotic mother. I think neurotic mothers want more time to ask questions that are not really relevant to …” (user no 14)

This observation, however, was not a consistent finding across all cases—reported by two out of four mothers.

Individuals who live alone

Individuals who live alone appeared to have certain expectations of the service. One respondent, a 52-year-old woman who lives alone, had a reaction to tablets that she had been given by her doctor. The report of her experience highlighted feelings of helplessness and dissatisfaction with the service she received. Below are excerpts from her interview:

“… I live alone and I was too ill to get myself to a hospital … I wasn’t satisfied with the way I was being treated but I was too ill to do anything about it. I felt that I was having to look after myself the whole time”. (user no 28)

After the consultation

Participants who did not get the level of care that they expected reported that they had then accessed other aspects of unscheduled primary care, making follow-up, and in some cases multiple, consultations for the same problem (see table 3).

What users value

Respondents identified various aspects of the current service they felt could be improved. In analysing their responses, the following themes emerged:


The aspects of continuity highlighted were twofold:

During the consultation

Suggestions to improve continuity involved the “stepwise” process experienced in contacting the service with the view that the process consisted of too many, potentially unnecessary, contacts, which prolonged the consultation process. According to user no 18:

“… There were so many questions it would have been better if I could have said all this to one person, instead of going over it with the operator, then with the nurse and then with the doctor.”

For onward referrals

Respondents who had chronic illnesses or required a specialised level of care through a referral network made comments about the service not meeting their needs efficiently. One respondent (case no 3 in table 3) reported having made six out-of-hours contacts on behalf of his wife before she was eventually referred to a specialist.

Attitude/manner of health professionals

Comments concerning the clinicians’ attitude and manner were both positive and negative. Attributes such as “caring” and interest in the patient as an individual were associated with evaluations of clinicians as good or bad. A female respondent, who had had a reaction to strong pain relief, made the following contrasting comments about the two health professionals she had consulted:

“I felt that the doctor I spoke to on the phone was good, and gave the impression of being a caring professional”.

She then commented on the consulting clinician at the treatment centre:

“He didn’t seem to be interested in me as a patient. I got the impression that he was very laid back and that really I was disturbing his evening. But that’s just the impression that I got”. (user no 28)

Confidence in the clinician’s ability

Confidence or a lack of it in the ability of the health professional was also an important aspect of the participant’s evaluation of the clinician, which influenced compliance with advice. User no 19, who declined the doctor’s advice to go into hospital, commented:

“She seemed very young and I think she was trying to be over cautious … I don’t know whether she was fully qualified or whether she was still training or something”.

User no 6, who also had a home visit but agreed to go into hospital at the doctor’s advice, commented:

“I was confident; I could tell by his voice that he was confident so I was”.


Principal findings

Although most respondents (n  =  25, 83%) reported satisfaction with the overall service received, a significant minority (n  =  5, 17%) commented on issues with which they were dissatisfied. Dissatisfaction resulted in follow-up consultations (sometimes numerous), with other aspects of unscheduled care such as accident and emergency, ambulance, NHS Direct (UK-wide telephone advice line), their named GPs and the GP out-of-hours service itself. However, when experiences were positive, respondents expressed preference for the GP out-of-hours service over the other unscheduled care services, including their named GPs. Before contacting the service, respondents had specific expectations of the type of consultation or service they sought, be it a home visit, base visit, a hospital referral, or a prescription. Unmet expectations, anxiety during waits and resultant dissatisfaction were particularly evident in specific user groups. They include patients without previous experience of their illness, mothers with children under 5 years, individuals who live alone and those requiring a specialised level of care through a referral network. Users in these groups reported that they had made follow-up and in some cases multiple consultations when their expectations had not been met.

Findings in context

As in other studies,12 13 participants in this study reported having various desires and expectations from the service. Some users seemed to expect more, and in some instances much more, than they could otherwise obtain from a particular type of consultation. Expectations were particularly high regarding waiting times in the group seen at the treatment centre—even though the actual (reported) waiting times were shorter than for telephone advice or home visits. In effect, these users appeared to desire and expect a “walk-in” treatment centre—clearly analogous to the “walk-in centre” recently developed and evaluated in NHS England.14 15

A mismatch between users’ expectations and the type of service received results in dissatisfaction, which influences help-seeking behaviour, thus leading to the further use of unscheduled care services or the use of different services, perhaps inappropriately, in subsequent illness episodes. This observation is supported by one recent study12 but contradicted by another older study,13 although the latter had a specific sample of older male patients. McKinley et al12 recommend actively managing user expectations to improve satisfaction. External communications with the users may be an effective strategy for this.16

Unmet expectations were evident in specific user groups—those living alone, parents of children under 5 years and those without previous experience of their illness—similar to a previous study.17 The service needs to account for users in these and other particular groups such as those requiring referral networks for specialist care, in order to cater for their needs more holistically. This may help curb the issue of unmet expectations and repeated access of unscheduled care and other services. The pattern of further service use, when patients are dissatisfied or their perceived needs are not met, indicates suboptimal quality of care, which may impact on safety (by not choosing the most appropriate service).

The responses here give some indication of what might be changed to improve services from the patient’s perspective—users’ expectations, communication skills of clinicians, continuity and coordination in referrals. However, they also indicate a need to examine these issues in more depth and to understand the process of seeking and receiving help from users’ perspectives; also to identify interventions to develop services that address these needs and challenges. The ability of such interventions and developments to achieve the desired goals—improved quality, safety and efficiency—will require evaluation in well-designed research.

Study limitations

Sampling occurred during the summer when service use is relatively low. Participants’ experiences may be different at times when there are increased demands on services, which may not have been captured in the present study. Nevertheless, the qualitative method of the study expands the scope of the research by providing an insight into patients’ experiences, which in turn sheds light on patient expectation as it influences their satisfaction and help-seeking behaviour during and after contact with the out-of-hours service.


Most respondents (n  =  25, 83%) in this study were satisfied with the service, but a significant minority (n  =  5, 17%) were dissatisfied. This dissatisfaction often resulted in subsequent re-access of unscheduled care services, representing a poor experience for the patient and inefficiency for the service. Areas of dissatisfaction highlight potential areas for intervention. Whereas some users’ views and expectations may be used to inform future design to improve services, information and education to others through public awareness initiatives is important to enable people to make the best use of the available services. More training is needed for administrative and support staff on how to explain how the service operates and what users can expect. Other suggested areas for improvement include building upon pre-existing interpersonal skills among clinicians, thus enabling them to address patients’ expectations and concerns better. Such service developments require evaluation, but could improve the patient experience and patient satisfaction, and may also reduce the incidence of multiple contacts with the service for the same episode, thus ultimately improving overall resource use efficiency.


The authors would like to thank the service users who participated in the study and the staff at the Gwent Healthcare NHS Trust GP out-of-hours service for their help in the administration of the project. They also thank Miss Gwennan Thomas for her work in transcribing the interviews and Dr Linda Ball who reviewed the article and its revision for the journal.


Supplementary materials


  • ( The interview structure is published online only at

  • Funding: This study was part of a larger project funded by the Gwent Healthcare NHS Trust and Blaenau Gwent, Caerphilly, Monmouthshire, Newport and Torfaen Local Health Boards.

  • Competing interests: None.

  • Ethics approval: UK Multi-centre Research Ethics Committee application and approval was gained for the study in June 2005 (05/MRE09/35).

  • Patient consent: Obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.