Sampling design

A telephone survey was conducted with a random sample of the adult population (aged 18 and over) of British people residing in France. The fieldwork was organised by IFOP, a French-based market research institute, using computer-assisted telephone interviewing (random-digit dialling). The target population was identified through a telephone directory using a list of names provided by the Office for National Statistics. The study was conducted in March 2011.

We contacted 4057 individuals. People who were not natives of Great Britain (from the United States, from South Ireland, or from another country) (n=2119), whose main residence was not in France (n=147) and who had never lived in Great Britain before (n=11) were excluded. Of the remaining 1780 eligible participants, 1004 agreed to participate (56%). Among these participants, 144 had not consulted a general practitioner (GP) at least once during the past 12 months and therefore were unable to respond to the questionnaire. Thus, a total of 860 respondents participated in the study.

Questionnaire

The Europep questionnaire was used [5]. The core of the questionnaire is a set of 23 questions evaluating different aspects of general practice care, using a five-point answering scale with the extremes labelled as “poor” and “excellent”. The items are organised into 4 domains: doctor-patient relationship (6 items), information and support (4 items), medical-technical care (5 items) and organisation of service (8 items) [6]. The respondents were asked to evaluate care provided in France within the past 12 months. The interviews were conducted in English. General and demographic data were also collected from the participants.

Statistical analysis

Reliability analysis was used to verify the internal consistency (Cronbach’s coefficient alpha) within the Europep instrument. The items were dichotomised according to 1 versus 2–5 (excellent versus less than excellent). This procedure is recommended as a standard for Europep users [7]. Frequency distributions were used to describe the participants’ characteristics and their responses to the 23-items questionnaire.

We used chi-square tests to compare their characteristics and responses to the questionnaire according to whether respondents were confident (very/somewhat confident) or not (not very confident/not at all) in the NHS when they were still living in the UK.

The association between scoring an item “excellent” and the degree of confidence in the NHS was also investigated for all items after adjusting for age (18 to 39 years / 40 to 54 / 55 and over), gender (M/F) and number of years of residence in France (2 years or less / 2 to 5 / more than 5). Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. We used SAS® software.

General characteristics

Responses to the Europep questionnaire

Degree of confidence in the NHS and responses to the Europep questionnaire

Responses to the Europep questionnaire

We observed that British people living in France were generally satisfied with the general practice care provided. The aspects that were rated the highest were related to the doctor-patient relationship. Some aspects of the organisation of services, such as “preparing you for what to expect from specialists, hospital care or other care providers”, “being able to talk to the general practitioner on the telephone” and “waiting time in the waiting room” received relatively negative evaluations. This first result is consistent with a previous report of Grol et al. [3] who compared the opinions of patients in various European countries and found that “patients are generally very positive about the care provided, but improvements in practice management in some countries are requested”.

Degree of confidence in the NHS and responses to the Europep questionnaire

The main result of our work was to show that a favourable opinion on certain aspects of general practice care in France was independently and positively associated with the lack of confidence (not very/not at all) in the NHS. We should say again that this variable was used as a proxy of ‘the experience of care’ as it is usually the case in the literature [4]. Most of these aspects concerned the doctor-patient relationship and information and support during the consultation. Few of these favorable opinions were related to the organization of practice. In our sample, 27% of the respondents were not confident in the NHS before moving to France. This figure is close to the one found in a recent survey on adults’ views of care systems overall in five Commonwealth countries in which 23% of respondents in the UK reported that they were not (or not at all) confident in the NHS [4]. The design of our study does not enable us to draw any causal link between a favourable opinion on certain aspects of general practice care in France and a lack of confidence in the British health system in general, and it was not the purpose of the study. However, the association that we have highlighted raises our curiosity because it appeared for all of the care items, even if it was not always statistically significant. It can be argued that results could have been different if we had made a comparative analysis by domains. However, as far as we know, the issue of the unidimensionality of the domains is debated and it remains unclear how to use the proposed domains [8]. This result is quite interesting since the issue of to what extent patient experience explains satisfaction with the health-care system is currently debated. Although it is more balanced in the study by Bleich et al. [9], the literature suggests that much of the remaining variation in health system satisfaction after adjusting for factors commonly used to measure the concept is a reflection of patient experience [10–12].

Limitations

This study has got some limitations. Our sample was not representative of all British citizens living in France because individuals under the age of 40 were under-represented. Individuals in this age group were difficult to reach by telephone, particularly during the day, despite efforts to call them between 7pm and 9pm. However, this under-representation was not an obstacle with regard to the objectives of the study, which focused on the users of the health care system over the past 12 years, not many of whom are in this age group. A second limitation concerns the degree of confidence in the NHS. This was not a degree of confidence in the NHS expressed at the same time t for all the respondents, since they migrated at different times, and the NHS has been reformed over the past ten years. This limitation emphasises that one must be cautious about finding any causal link between a lack of confidence in the NHS and a favourable opinion of certain aspects of health care in France. A third limitation concerns a potential recall bias. Actually, almost 70% of the respondents left the UK at least 5 years ago. However, we did not ask respondents to recall precise details about the NHS, just to rate whether they were confident or not in the NHS when they were still living in the UK which did not represent a very important memory effort. Especially as these individuals had for most of them (three quarters of the individuals being older than 55 years old) a long experience of care provided by GPs before moving to France and had probably the time to form their opinion. Finally, due to time constraints, we did not ask people about their living conditions in the UK. This factor could have been linked to their degree of confidence in the NHS and could therefore influence the interpretation of our results.