Comparing care at walk‐in centres and at accident and emergency departments: an exploration of patient choice, preference and satisfaction

Abstract

Objectives

To explore the impact of establishing walk‐in centres alongside emergency departments on patient choice, preference and satisfaction.

Methods

A controlled, mixed‐method study comparing 8 emergency departments with co‐located walk‐in centres with the same number of “traditional” emergency departments. This paper focuses on the results of a cross‐sectional questionnaire survey of users.

Results

Survey data demonstrated that patients were frequently unable to distinguish between being treated at a walk‐in centre or at an accident and emergency (A&E) department and, even where this was the case, opportunities to exercise choice about their preferred care provider were often limited. Few made an active choice to attend a co‐located walk‐in centre. Patients attending walk‐in centres were just as likely to be satisfied overall with the care they received as their counterparts who were treated in the co‐located A&E facility, although walk‐in centre users reported greater satisfaction with some specific aspects of their care and consultation.

Conclusions

Whereas one of the key policy goals underpinning the co‐location of walk‐in centres next to an A&E department was to provide patients with more options for accessing healthcare and greater choice, leading in turn to increased satisfaction, this evaluation was able to provide little evidence to support this. The high percentage of patients expressing a preference for care in an established emergency department compared with that in a new walk‐in centre facility raises questions for future policy development. Further consideration should therefore be given to the role that A&E‐focused walk‐in centres play in the Department of Health's current policy agenda, as far as patient choice is concerned.

A new wave of National Health Service (NHS) walk‐in centres opened during 2004, with many of them co‐located with accident and emergency (A&E) departments. The expansion of the walk‐in centre initiative in this way reflects the UK health policy determination that people should have greater choice in where they access health care, who should be providing services and how they should be delivered by providing convenient alternatives which meet the needs of different sectors of the population.¹,²,³ It is also linked with national policy to reduce waiting times in emergency departments.

As discussed in the companion paper, there is evidence to show that co‐locating walk‐in centres alongside emergency departments has little impact on the rates of attendance, duration of visits, process, costs or outcomes of care.⁴ This paper, therefore, seeks to explore the impact of establishing walk‐in centres alongside emergency departments from a different perspective—focusing not on organisational targets but on patients' experience of this new healthcare setting, with particular reference to choice, preference and satisfaction.

Methods

Overview of design

A controlled, mixed‐method evaluation was conducted comparing a sample of eight emergency departments with new co‐located walk‐in centres in England with the same number of emergency departments without additional facilities, matched according to performance against the national 4 h emergency access target, size and case‐mix. This paper draws primarily on data from a cross‐sectional patient survey, linking them with other observational and service‐level data obtained in the course of the wider evaluation. Since full details of the various data sources informing this study are outlined in the companion paper, this section describes only the additional survey methods employed.

Sampling

An initial sampling frame of 3200 potential survey respondents was constructed from the detailed anonymised patient data provided by all 16 study sites—200 per site—as part of the wider study. These data related to a specified time period after the opening of walk‐in centres in sites with co‐located facilities, and to a corresponding time period at the matched A&E departments. In order to focus the survey on patients with less serious conditions (which were more likely to be suitable for walk‐in centre care) and to avoid undue distress, all those patients who were admitted to hospital were excluded from the sampling frame. Patients aged <16 years were also excluded.

Data collection

The questionnaire was developed as a means of gaining information about the characteristics of service users, their reasons for attending particular facilities, their a priori expectations of these facilities, and their satisfaction with the services provided following treatment or care. As far as possible, the questionnaire was based on that used in the NHS Acute Trust Emergency Department Survey 2003,⁵ but with additional exploration of the notions of patient choice and problem resolution. Since both of these concepts are difficult to define and measure accurately, they were operationalised using the proxy measures of convenience of obtaining help and re‐consultations for the same healthcare problem, respectively.

Box 1: Questionnaire content

The questionnaire collected data on:

• sociodemographic characteristics

• presenting complaint

• where the patient said they would have gone had the walk‐in centre/A&E department not been available

• improvement in presenting complaint

• consultations with any health service professional since the initial consultation

• satisfaction with access to care, the waiting time, facilities, treatment and advice received at the A&E department and/or walk‐in centre

• the extent to which patients felt the facility was a convenient way of obtaining care

• whether the service they attended accorded with their choice of facility

• patient pathway—that is, whether the patient was transferred to the walk‐in centre from the emergency department or vice versa.

Survey administration

The survey was conducted between January and June 2005. In order to maintain patient confidentiality, each person listed as eligible for the survey was sent the questionnaire directly from his or her local study site, some 4 weeks after their original consultation, along with a covering letter explaining the reasons for their inclusion and a pre‐paid return envelope, in which they could send their reply directly to the research team. A reminder and duplicate questionnaire were sent 3 weeks after the original mailing, if a response was still outstanding at that time. Questionnaires were marked with the patient's unique identifying number, but were otherwise anonymous. In order to minimise any seasonal effects, the survey was undertaken in each matched pair of sites—that is, co‐located facility and “traditional” A&E department at the same time.

Analysis

The data were coded and analysed using SPSS and Stata. Comparisons were made between patients attending intervention sites (ie, those with co‐located emergency department and the walk‐in centre facilities) and those attending the control sites (ie, those without co‐located facilities), using appropriate regression models, allowing for clustering and sampling probability. All percentages cited are weighted to take into account the probability of sampling. A series of dichotomous “problem scores” were also created from each of the relevant variables in the dataset, following the protocol described in the development of the NHS Acute Trust Emergency Department Survey 2003.⁵

Results

Response rate

A total of 2017 patients were identified as potential participants in the patient survey, by reason of being aged ⩾16 years and having not been admitted to hospital following their initial consultation. Subsequently, 65 potential respondents were deemed ineligible on a variety of grounds including mental incapacity, having no known address or having died. Of the 1952 eligible service users, 704 successfully completed and returned a questionnaire, which equates to an overall survey response rate of 36.1%. This response rate varied slightly between groups (32.9% at intervention sites and 39.7% at control sites) and considerably within groups (between 14.1% and 43.2% at intervention sites and between 21.6% and 51.6% at control sites).

Data indicated that survey respondents were marginally more likely to be female and older than non‐respondents, although none of the differences were statistically significant. There was no difference between respondents and non‐respondents in terms of whether a doctor or a nurse was consulted.

Reasons for attending

Survey data showed that the greatest proportion of patients attending all types of facility presented with an injury, with relatively few presenting with illness or some other kind of problem (table 1).

Table 1 Reasons for attending the hospital.

Reason patient attended hospital	Type of facility attended
	Intervention A&E	Intervention walk‐in centre	Intervention combined	Control A&E	p Value*
	n = 112	n = 219	n = 331	n = 360
Injury	47 (39.5)	118 (55.1)	165 (44.7)	194 (53.4)	0.39
Recent illness	12 (12.1)	29 (12.0)	41 (12.1)	37 (10.6)
Illness for >2 weeks	15 (15.4)	23 (9.9)	38 (13.6)	30 (8.5)
Other problem	38 (32.9)	49 (23.0)	87 (29.6)	99 (27.5)

A&E, accident and emergency.

Values are expressed as count (%).

*Comparison between intervention combined and control sites, using appropriate regression models, allowing for clustering and sampling probability. Percentages in the table also take account of probability of being sampled.

Route of access to care

Table 2 shows how the majority of patients attended the A&E department first, even at sites where there was a co‐located walk‐in centre. Most of the people recorded as having been seen in a walk‐in centre (79%, 170/220) had initially chosen to attend an A&E department and had subsequently been re‐directed to the walk‐in facility.

Table 2 Data relating to where patients attended initially.

Type of facility patient attended initially	Type of facility where patient was recorded as being seen
	Intervention A&E	Intervention walk‐in centre	Intervention combined	Control A&E	p Value*
	n = 113	n = 220	n = 333	n = 362
A&E	95 (84.4)	170 (79.3)	265 (82.7)	333 (92.3)	0.001
NHS walk‐in centre	15 (12.7)	40 (14.9)	55 (13.4)	12 (3.0)
Somewhere else	3 (2.9)	10 (5.8)	13 (3.9)	17 (4.7)

A&E, accident and emergency; NHS, National Health Service.

Values are expressed as count (%).

*Comparison between intervention combined and control sites, using appropriate regression models, allowing for clustering and sampling probability. Percentages in table also take account of the probability of being sampled.

Patients' choice of facility

Relatively few people reported having made an active choice to attend a walk‐in centre. When asked where they would have preferred to be seen, 35% (70/215) of those seen in a walk‐in centre said they would rather have been seen in an A&E department, whereas only 13% (13/110) of patients seen in a co‐located A&E facility or 12% (38/260) of those attending a stand‐alone A&E department would have chosen to attend an NHS walk‐in centre. More than one‐third of patients in each healthcare setting did not express any preference.

Table 3 shows that 55% (117/215) of those attending a walk‐in centre did not even realise that they were seen at that kind of facility, stating in their survey response that they had been treated in an A&E department. This is consistent with the finding from the site observations that, in some locations, the walk‐in centre was a nominal concept, with very little to indicate to patients that they were being treated in something other than a “traditional” A&E department.

Table 3 Date showing where patients stated they were seen in relation to where they were recorded as being seen.

Type of facilty patient stated they were seen	Type of facility where patient was recorded as being seen
	Intervention A&E	Intervention walk‐in centre	Intervention combined	Control A&E
	n = 109	n = 215	n = 324	n = 355
A&E department only	84 (75.9)	117 (55.1)	201 (69.0)	324 (91.4)
Walk‐in centre only	15 (13.0)	35 (12.7)	50 (12.9)	11 (2.7)
A&E, then walk‐in centre	8 (9.3)	52 (26.1)	60 (14.9)	7 (2.2)
Walk‐in centre, then A&E	–	6 (3.1)	6 (1.0)	1 (0.3)
Other	2 (1.7)	5 (3.0)	7 (2.1)	12 (3.4)

A&E, accident and emergency.

Values are expressed as count (%).

Convenience of obtaining care

Generally speaking, satisfaction related to accessing care was high, with arrangements being described as “fairly convenient” or “very convenient” by the majority of respondents. Moreover, no significant differences were observed, with regard to convenience of obtaining care, between patients who presented at the various healthcare settings.

Reason for attendance at a particular type of facility

Comparing types of facility, it appears that more people initially chose to attend sites with both an A&E department and a co‐located walk‐in centre, rather than a “traditional” A&E facility, due to an expectation of a shorter wait for treatment or because it would be quicker than getting a GP appointment (table 5). This may reflect the fact that walk‐in centres have often been specifically established in those areas where people were known to have difficulty accessing primary care services. However, when considering both types of facility available to patients at intervention sites, there was a suggestion that people choosing to go to the walk‐in centre did so because they felt it was quicker than getting a GP appointment, whereas people attending the co‐located A&E department initially did so because they thought it was the most appropriate place for their problem.

Table 5 Reasons for choosing the first facility attended.

Reason patient attended facility	Type of facility patient first attended
	Intervention A&E	Intervention walk‐in centre	Intervention combined	Control A&E	p Value*
	n = 260	n = 55	n = 331	n = 356
Convenient location	73 (27.5)	10 (21.6)	87 (26.1)	67 (18.6)	0.18
Convenient opening hours	35 (14.1)	13 (23.4)	49 (14.8)	45 (13.2)	0.55
Quicker than getting GP appointment	70 (25.3)	22 (40.4)	93 (27.0)	49 (14.7)	0.01
Would be a shorter wait	22 (8.0)	2 (4.9)	24 (7.2)	9 (2.8)	0.04
Best place for my particular problem	119 (50.3)	10 (12.2)	134 (44.5)	136 (37.2)	0.21
Not registered with a GP	10 (3.5)	6 (14.8)	17 (5.4)	9 (2.6)	0.20
Wanted a second opinion	9 (4.7)	0 (0)	10 (4.0)	8 (2.4)	0.17
Didn't want to bother my GP	7 (1.6)	0 (0)	7 (1.3)	9 (2.5)	0.14
My GP wasn't available	36 (15.2)	10 (17.8)	46 (14.7)	29 (7.9)	0.04
No NHS walk‐in centre nearby	8 (2.6)	2 (4.5)	11 (3.3)	29 (8.7)	0.03
Sent there by my GP	21 (7.5)	9 (18.2)	34 (9.5)	48 (13.7)	0.04
Sent there by NHS direct	8 (4.0)	0 (0)	9 (3.6)	26 (8.0)	0.004
Sent there by someone else	21 (7.8)	7 (14.8)	30 (8.8)	47 (12.6)	0.54
Didn't think about going anywhere else	37 (13.6)	8 (8.1)	46 (13.1)	56 (16.0)	0.37

A&E, accident and emergency; GP, general practitioner; NHS, National Health Service.

Values are expressed as count (%).

*Comparison between intervention combined and control sites, using appropriate regression models, allowing for clustering and sampling probability. Percentages in the table also take account of the probability of being sampled.

Satisfaction with handling of main problem

When asked to give feedback on the way in which the problem they had attended the hospital with was handled, 59% (410/690) of respondents felt that the problem had been dealt with to their complete satisfaction, with a further 31% (212/690) believing that it had been resolved to a lesser extent but still satisfactorily. There was no evidence of any significant difference between satisfaction levels at co‐located or stand‐alone facilities, nor between walk‐in centres and A&E departments within co‐located sites in terms of reported patient outcome.

Overall rating of care received

Survey data indicated that 65% (446/691) of respondents described the care they received as being “very good” or “excellent”, with no significant differences in reporting between types of site. Similarly, patients attending walk‐in centres were just as likely to be satisfied overall with the care they received as their counterparts who were treated in the co‐located A&E facility.

However, when the data were recoded according to the protocol used in the NHS Acute Trust Emergency Department 2003 and re‐analysed as a series of dichotomous “problem scores”, there were clear differences in the views expressed by patients attending the various healthcare settings in relation to some aspects of care and patient consultation. Patients attending a co‐located A&E department were more likely to report dissatisfaction than patients attending the co‐located walk‐in centre in relation to visit duration, cleanliness of the facility, time given to discuss healthcare problems, involvement in decision‐making, discussion of fears and anxieties, and privacy during the consultation.

Discussion

Principal findings

This survey demonstrated that service users were frequently unable to distinguish between being treated at a walk‐in centre or at an A&E department, perhaps due to low visibility of the new walk‐in centre facilities or due to their high degree of integration with the co‐located emergency facility. Even where services were distinct, opportunities for patients to exercise choice over their preferred care provider were often restricted by triage and streaming practices, since this relies on staff, rather than patients, deciding on the most appropriate source of care. Taking into account the similarities around treatment and process of care at sites with and without walk‐in centres,⁴ it is not surprising that few differences were observed in patients' perceptions of their care. Survey respondents who presented at sites with walk‐in centres did not find their care any more convenient than those who presented at the A&E, nor were they more likely to be more satisfied with their visit to the hospital as a whole. A small proportion of patients treated in co‐located walk‐in centres, however, reported greater satisfaction with a number of particular aspects of their care and consultation. It is not possible though, from this research alone, to judge which individual component of that care or consultation was responsible for the increase in satisfaction—for example, new facilities or dedicated staffing—nor whether the same effect could have been reproduced in surroundings other than walk‐in centres.

Table 4 Convenience of obtaining care.

Assessment of patient regarding convenience of obtaining care	Type of facility patient attended
	Intervention A&E	Intervention walk‐in centre	Intervention combined	Control A&E	p Value*
	n = 113	n = 221	n = 334	n = 356
Very convenient	70 (63.6)	134 (61.0)	204 (62.7)	198 (55.1)	0.15
Convenient	36 (30.4)	67 (29.0)	103 (30.0)	122 (34.8)
Not very convenient	6 (5.2)	12 (6.5)	18 (5.7)	21 (5.6)
Not at all convenient	1 (0.7)	8 (3.5)	9 (1.6)	15 (4.5)

A&E, accident and emergency.

Values are expressed as count (%).

*Comparison between intervention combined and control sites, using appropriate regression models, allowing for clustering and sampling probability. Percentages in the table also take account of the probability of being sampled.

Study limitations

This patient survey has a number of limitations. Firstly, triage and streaming of patients on arrival at intervention sites invariably results in systematic differences between the characteristics and case‐mix of patients attending the different healthcare settings under observation. This may explain some of the differences observed when comparing aspects of care or patient consultation at the different facilities available at intervention sites. Secondly, the response rate (36.1%) is slightly lower than that achieved previously in similar surveys⁵ within the same setting. This is partly due to the fact that, although two survey reminders were planned, this was deemed coercive by the ethics committee and only one reminder was sent. As a result, the generalisability of the respondents' experiences reported here is necessarily limited, and these findings should be interpreted with caution since non‐respondents to the questionnaire may have had different experiences from responders. However, it is unlikely that there will have been different effects at intervention and control sites.

Policy implications

It is arguable that, for patients, the overall impact of this new wave of A&E‐focused walk‐in centres is limited, since few actively chose to attend a walk‐in centre for advice or treatment. Indeed, at present, the majority of the walk‐in centre population is redirected there from the co‐located A&E department, and, as a result, the case‐mix of the walk‐in centre is essentially selected by A&E staff. This has led to many of the new walk‐in centres both resembling and functioning as a “minors” stream of A&E—something borne out by the high proportion of survey respondents reporting an injury rather than an illness as their reason for attendance.

From a patient's perspective, an active decision to attend a walk‐in centre in preference to its co‐located A&E department would seem rational only if there were clear benefits—for example, in proximity or waiting time. The policy decision to locate walk‐in centres immediately next to existing A&E departments rules out, by its very definition, any differences with regard to the first of these, leaving waiting time as the only feasible advantage. The policy aim of establishing walk‐in centres to improve access to care appears to have been lost or subsumed by a more immediate demand to reduce A&E workload and waiting times.

Table 6 Dissatisfaction reported when comparing walk‐in centres and A&E departments within colocated sites and control sites.

Reason for dissatisfaction expressed as a “problem score”*	Type of facility patient attended
	Intervention A&E	Intervention walk‐in centre	Control A&E	p Value†
Length of visit	28 (23.5)	23 (11.5)	70 (19.5)	0.03
Cleanliness of facilities	49 (43.6)	57 (27.7)	137 (38.0)	<0.001
Time available for discussion of the problem	49 (43.6)	55 (26.5)	135 (37.3)	<0.001
Involvement in decision–making	34 (32.5)	51 (24.1)	112 (31.4)	0.01
Privacy when discussing the problem	35 (30.9)	47 (23.1)	96 (26.9)	0.01
Privacy during examination or treatment	31 (28.8)	35 (18.0)	68 (19.2)	0.02

A&E, accident and emergency.

Values are expressed as count (%).

* “Problem scores” were calculated for numerous aspects of care or patient consultation. For the sake of brevity, only those found to have a statistically significant difference are reported here.

†Comparison between intervention A&E and intervention walk‐in centre facilities within intervention sites, using appropriate regression models, allowing for clustering and sampling probability. Percentages in the table also take account of the probability of being sampled.

Conclusions

There is no evidence, from the data available, that walk‐in centres co‐located with A&E departments have achieved the aim of increasing patient choice, preferences or satisfaction with care received. This is probably related to the finding that, at present, these facilities have a low public profile, with most activity arising out of staff‐initiated re‐direction rather than through patient‐expressed preference and choice. Further consideration should therefore be given to the role that A&E‐focused walk‐in centres might play in the Department of Health's current policy agenda, particularly in relation to patient choice, preference and satisfaction.

Abbreviations

A&E - accident and emergency

NHS - National health service