General Practice

Comparison of out of hours care provided by patients' own general practitioners and commercial deputising services: a randomised controlled trial. II: the outcome of care

BMJ 1997; 314 doi: (Published 18 January 1997) Cite this as: BMJ 1997;314:190
  1. R K McKinley, senior lecturera,
  2. D K Cragg, lecturera,
  3. A M Hastings, lecturera,
  4. D P French, research associatea,
  5. T K Manku-Scott, research associatea,
  6. S M Campbell, research associateb,
  7. F Van, research associateb,
  8. M O Roland, professor and director of research and developmentc,
  9. C Roberts, senior research fellowc
  1. aDepartment of General Practice and Primary Health Care University of Leicester Leicester General Hospital Leicester LE5 4PW
  2. b Department of General Practice University of Manchester Rusholme Health Centre Manchester M14 5NP
  3. c National Primary Care Research and Development Centre Williamson Building University of Manchester Manchester M13 9PL
  1. Correspondence to: Dr McKinley
  • Accepted 22 November 1996


Objective: To compare the outcome of out of hours care given by general practitioners from patients' own practices and by commercial deputising services.

Design: Randomised controlled trial.

Setting: Four urban areas in Manchester, Salford, Stockport, and Leicester.

Subjects: 2152 patients who requested out of hours care, and 49 practice doctors and 183 deputising doctors (61% local principals in general practice) who responded to the requests.

Main outcome measures: Health status outcome, patient satisfaction, and subsequent health service use.

Results: Patients seen by deputising doctors were less satisfied with the care they received. The mean overall satisfaction score for practice doctors was 70.7 (95% confidence interval 68.1 to 73.2) and for deputising doctors 61.8 (59.9 to 63.7). The greatest difference in satisfaction was with the delay in visiting. There were no differences in the change in health or overall health status measured 24 to 120 hours after the out of hours call or subsequent use of the health service in the two groups.

Conclusions: Patients are more satisfied with the out of hours care provided by practice doctors than that provided by deputising doctors. Organisation of doctors into large groups may produce lower levels of patient satisfaction, especially when associated with increased delays in the time taken to visit. There seem to be no appreciable differences in health outcome between the two types of service.

Key messages

  • Between 24 and 120 hours after a request for out of hours care patients cared for by deputising services and practice doctors show no difference in health status

  • There is no difference in health service use between the two groups in the two weeks after a request for out of hours care

  • Patients are more satisfied with out of hours care provided by their own practice doctors

  • Though patients are more satisfied with out of hours care provided by practice doctors, their health outcomes are no better than when care is provided by a deputising service


It has been suggested that out of hours primary medical care provided by practice and deputising service doctors differs in quality1 but there has never been a randomised controlled comparison. Three descriptive studies provided some evidence that patients are less satisfied with care provided by deputising doctors,2 3 4 though in two studies there were long delays between the study visit and data collection2 3 and one was of a single practice.4 All used satisfaction instruments of uncertain reliability and validity,5 so the validity of these measures was questionable. We report a randomised controlled trial comparing the outcome of out of hours care given by practice and deputising doctors. The outcome measures used were patient satisfaction, health status, and subsequent use of health services. Our accompanying paper reports on process measures used in the trial.6

Subjects and methods

Details of the recruitment of participating practices, study design, and randomisation are given in our accompanying paper.6 Data on patient satisfaction and health status were collected by interview between 24 and 120 hours after the out of hours call. The time to interview was balanced between the arms of the study; two thirds of the interviews took place between 24 and 72 hours after the call.

A modified version of the anglicised short form 36 (SF-36) questionnaire7 was used to gather health status information on all patients aged 16 or more. With permission from the Health Institute, Boston, question stems were modified to ask about health “since you saw the doctor” instead of “in the last four weeks.” If the patient or informant was unable to complete the questionnaire the standardised interviewer version of the SF-368 was used with the stems changed in the same fashion. For children aged 5-15 we used a development 30 item, nine subscale version (short form 2.01) of the child health questionnaire parent form to measure health status (J Langraff, personal communication, 1993). For children under 5 we used four scales (overall health, physical ability, temperament or mood, and impact of illness on the carer) to measure health status. Each scale had five points. Change in health was also measured for all patients by the transition question, “Compared with how you felt when you called the doctor, how do you feel now?” Responses were recorded on a graduated visual analogue scale, on which zero corresponded with “much worse,” 40 with “no difference,” and 100 with “completely better.”

Patient satisfaction was assessed with a questionnaire9 developed by established qualitative and quantitative methods.10 Scale scores were calculated by scoring questions from one to five (five always representing high satisfaction), summing them, and expressing the total as a percentage of the maximum score for the scale. Results for five scales are presented. The remaining scales (satisfaction with access, the person who answered the telephone, and telephone advice) are not reported, as few patients received telephone advice from the deputising services and all but one practice used the same telephone answering service whether practice doctors or the deputising service provided care. Data on health service use in the two weeks after the out of hours call were extracted from the medical records of those patients who gave consent.

Analysis was carried out by multilevel modelling techniques,11 as described.6


A total of 1046 requests for care to practice doctors and 1106 to deputising doctors were studied. Details of the calls studied and response rates are given in the accompanying paper.6 Though intradoctor cluster correlation coefficients were negligible for most health status measures (<0.01), they were between 0.019 and 0.068 for the satisfaction scales.

Health status–Table 1 gives the scores for adults on the SF-36 subscales. There were no differences between the scores of patients cared for by practice and deputising doctors. For infants and children none of the health status scales showed differences between the groups. After adjustment for age, sex, ethnic group, and access to a car the mean transition question scores were 69.1 (95% confidence interval 67.4 to 70.9) and 69.2 (67.7 to 70.8) for practices and deputising services respectively. This represents scores midway between “a bit better” and “much better.”

Table 1

Health status scores (on SF-36 subscales) of patients between 24 and 120 hours after out of hours visit. Scores are adjusted for age, sex, ethnic group, and access to a car

View this table:

Patient satisfaction–Table 2 shows that adjusted mean scores for four of the five patient satisfaction scales were higher for practice doctors than for deputising doctors, the difference being most pronounced for the “Delay until visit” scale. Table 3 shows the adjusted percentages of patients who expressed clear dissatisfaction for each scale. Almost twice as many patients who received care from deputising doctors expressed clear dissatisfaction with overall care, delay until visit, and communication with the doctor. Only 34.3% (95% confidence interval 29.3% to 39.8%) and 17.1% (14.3% to 20.4%) of patients cared for by practice and deputising doctors respectively expressed clear satisfaction (adjusted scale scores greater than 62.5) with delay until visit.

Table 2

Mean satisfaction scores in patients seeing practice doctors and deputising doctors. Scores are adjusted for age,sex, ethnic group, access to a car, and (except for “delay until visit”) time between request and visit

View this table:
Table 3

Percentage of patients dissatisfied (mean item score < 37.5 on scale 0-100 (50=neutral)) among those who had seen practice doctors and deputising doctors. Scores were adjusted forage, sex, ethnic group, access to a car, and (except for “delay until visit”) time between request and visit

View this table:

Subsequent health service use–Data on subsequent health service use were obtained for 1389 (94.7%) of the 1466 patients who were interviewed. Table 4 shows that after adjusting for age there were no significant differences in the number of follow up visits or prescriptions between the groups. More patients who received telephone advice from practice doctors consulted compared with those who received a visit from practice doctors (53.8% (95% confidence interval 44.4% to 62.9%) and 44.7% (39.8% to 49.9%) respectively), though the differences were not significant.

Table 4

Differences between groups in health service use in two weeks after out of hours call. Proportions adjusted for age, sex, ethnic group, and access to a car

View this table:


These data confirm that patients were more satisfied with out of hours care provided by doctors from their own practice than with that provided by deputising doctors.3 4 This effect remained after correcting for visit delay.3 12 Though a minority of patients expressed dissatisfaction with both services, it is not possible to compare our findings with those of other studies because of the different instruments used to measure satisfaction.2 3 4 12 The median delay between the patient's request for care and the practice doctor's arrival was 35 minutes,6 yet only one third of patients were clearly satisfied with the delay until visit. This may reflect the current emphasis on waiting times in many quality standards, including the patient's charter. It may be unreasonable to expect any service to offer a faster response than this between the patient's decision to seek care and a consultation with a doctor. We must consider whether expectations have been inappropriately raised.

We found no difference in expressed satisfaction with continuity of care, which may reflect the patient's perception of urgent need overriding any desire for continuity of care.13 14 Indeed, a patient of a multidoctor practice will not usually see his or her personal doctor out of hours,15 so that there may have been little difference between the services in the chances of patients seeing a doctor they knew.

Despite the differences in satisfaction with the care given by the two groups of doctors there were no differences in any of the three indices of health outcome. Transition questions provide valid measures of health change.16 As the SF-36 could not have been administered shortly after the request for care as a basis for comparison for the scores found at interview, we could not calculate its sensitivity to detect change in this context.17 Nevertheless, it has been shown to be reliable, valid, and appropriate for primary care7 18 and responsive to change in health.19 Furthermore, the sample size was large enough to detect clinically important differences in SF-36 scores.20 In addition, there was no greater demand for care in either group over the next 14 days. It therefore seems that there was no difference in health outcome between the groups.

Possible biases

Several possible biases exist in this study, which if present will have tended to improve the apparent performance of both services. As it was not possible to recruit a random sample of practices and deputising services for the trial, the data may not be representative of all out of hours care. The pattern of weekend and evening work for practice doctors was unchanged when 72%6 of requests were made. However, during the study many practice doctors who normally used deputising services for night visits provided care at night. Participants may therefore have changed their behaviour (the Hawthorn effect) when on duty, though such behaviour change can attenuate rapidly.21 To minimise such potential biases, practice doctors were not informed about which calls would be studied and, as far as possible, which duty periods would be sampled. Deputising services were not informed about which practices participated.

These results are of particular relevance to general practitioner cooperatives, which are likely to provide more out of hours care in the future.22 As about half of the care provided by deputising services was by doctors who were also local principals,6 the greater delay and reduced satisfaction found for deputising services may reflect the system of care rather than the doctor providing care. This suggests that increased provision of out of hours care by cooperatives staffed mainly by local practitioners will not adversely affect health outcomes, though patients' perceptions of the care provided may deteriorate.

In conclusion, we have shown that when practice doctors provide their own out of hours care patients are more likely to receive telephone advice and, when visited, are seen more quickly. In addition, these patients receive fewer, cheaper prescriptions, which are likely to be more carefully considered,6 and are more satisfied. There was no noticeable health gain or overall effect on health service use in the next two weeks. The “cost” to the general practitioner of providing personal out of hours care is fatigue,23 stress,24 and the risk of suboptimal performance next day.25 There may be differing interpretations of these results but our view is that the advantages of practice doctors continuing to provide personal out of hours care are small in the context of these difficulties.


We thank the partners and staff of the practices, the deputising services, and the patients who participated.

Funding: MRC Health Services Research Board. Service support for the participating practices was provided by Trent and North Western Regional Health Authorities.

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
View Abstract