Preference for a general practitioner and patients' evaluations of care: a cross-sectional study

Introduction

PERSONAL continuity no longer seems to be evident in general practice. Although reviews have shown reasonable evidence for the positive effects of personal contin-uity,1^,2 there appears to be denial of its importance. Patients make trade-offs between availability and accessibility,3 and consider contact with their personal general practitioner (GP) relatively unimportant for minor illness or regular checks.4^,5 GPs seem to agree with them,5 and policy makers, health services and doctors now put emphasis on general practice as a primary care service, in which a personal doctor is considered less important than previously. This may have repercussions on patients' perceptions and evaluations of care. We therefore explored whether dissent or indifference towards the GP contacted was related to patients' evaluations of consultations.

Methods

In the Groot Gelre district in the east of the Netherlands, 30 GPs from 17 practices participated in a project on contin-uity of care. Fourteen GPs worked with combined patient lists and sixteen with personal lists. We developed a two-part questionnaire, with pre- and post-consultation questions (see Supplementary information). Practice staff distributed 80 questionnaires per GP to consecutive patients in the waiting room, starting on the first day of the week. This impeded GPs from causing selection bias. Before the consultation, patients indicated whether they would have preferred to see another doctor that day, had it been possible, with a ‘yes’, ‘no’, or ‘no preference’ answer. Patients could seal this pre-consultation part. After the consultation, we measured patients' evaluations of consultations using the following four 5-point Likert questions: overall satisfaction with the consultation, the feeling of being helped forward, trust in this GP, and clearness of management plans. Responses were dichotomised by means of clustering ‘positive’ and ‘very positive’; and ‘neutral’, ‘negative’ and ‘very negative’. Patients with more than three missing values on a questionnaire were excluded from analysis. We used univariate analysis to explore the data, and the mixed model procedure (SAS), with the practice as a random effect, to calculate odds ratios with confidence intervals and corrected for practice type (personal or combined list), reason for encounter, perceived seriousness of symptoms, anxiety, age, and sex.

Results

From a total of 2400, we received 2142 completed questionnaires with fewer than three missing values. One hundred and four patients (5%) did not know which GP they would be seeing. Of the remaining 2038 respondents 6% stated that they would have preferred to see another GP, and 18% said that they had no preference. Patients from practices working with combined lists had no preference significantly more often than those from practices with personal lists (for combined lists 22% were indifferent, for personal lists 14% were indifferent; P<0.001). This also applied to younger patients (for patients aged 20–39 years 19.7% were indifferent, for 40–59 years 15.8%, and for 60–79 years 11.1%; P<0.001), and to patients who were less worried (using a 5-point scale from ‘much worried’ to ‘not worried at all’: 10.8%, 14.6%, 19.8%, 19.2%, 16.2% were indifferent; P = 0.032). Perceived seriousness of symptoms, sex, and reason for encounter were not found to be related to indifference. Most responders gave positive evaluations of care. Table 1 shows that patients who had no preference as to which doctor they would be seeing were significantly less satisfied, had less trust in the GP, felt less helped forward, and felt that management plans were less clear compared with patients who contacted their preferred GP. To a greater degree, this also applied to patients who would have rather seen another GP.

Discussion

Patients who were indifferent to the GP they would be seeing and patients who would have preferred another GP evaluated their consultations relatively less positively than patients who saw their GP of preference. The high numbers of patients who saw a GP of preference are not unusual for the Netherlands — yet. However, the outcome of this study questions the tendency in general practice to consider personal continuity unimportant, as this may lead to patients being indifferent as to which doctor to consult. Our results show clearly that less personal care is less efficient care from the patient's point of view.

As yet, the results should be interpreted with caution. Our design could not show causality, as patients with a strong bonding to their GP will probably make efforts to see this personal GP. These patients may, in any case, be more positive about consultations. On the other hand, patients from practices working with combined lists were more often indifferent as to which doctor they would be seeing, and this was related to poorer outcome. An explanation for the observed difference between practices with personal and combined lists may be mediated by both patient and practice factors. In practices with combined lists, patients may have less bonding with individual GPs and thus be indifferent more often. On the other hand GPs may feel less responsible for patients and induce indifference. In the Netherlands, patients generally are not inclined to change practice, or to choose their practice intentionally on the basis of practice organisation. Therefore, it is unlikely that bias occurred before the study through patient diversity. More studies on the value of personal continuity are needed to verify our findings. Until then, developments towards promoting less personal care should be reconsidered carefully.

Supplementary information

Additional information accompanies this paper at: http://www.rcgp.org.uk/journal/index.asp