The effects of interactions between patient characteristics on patients' opinions of general practice care in eight European countries

Abstract

The literature shows that patients' positive opinions of general practice care are associated with being older, of lower educational level, lower social status and being married, but there is no association with the sex of the patient. The health related characteristics of patients, such as self-reported health status, chronic condition and utilization of care, also predict judgements of the quality of some aspects of general practice care. This study investigated whether, in addition to individual effects, combinations of patient characteristics can also influence opinions of care, and whether these findings are valid in different European countries.

A self-completion questionnaire was developed and given to consecutive patients in five general practices in each of eight European countries. There were 48 questions covering five aspects of care: availability and access, organisation and co-operation, medical care, relationship and communication, information and support and an overall opinion. There were a further 16 questions on patient characteristics. One way analysis of variance, using SPSS software, was used to compare mean scores for each aspect of care for different values of the patient characteristic. Multivariate analysis was used to investigate the interaction effects between pairs of patient characteristics, which, individually were found to influence opinion. Further tests were made to identify any interaction between the effect of each patient characteristic and the country of residence.

1008 (response rate 63%) questionnaires were completed. The patient characteristics which influence patients' opinions were found to be age, levels of education and professional qualification, length of relationship with the practice, and whether they have a chronic condition. However, there are no interaction effects between these characteristics. Importantly these effects were found to be valid in all countries in this study. The patient characteristics which do not appear to influence opinions of care are sex, levels of utilization of care and self-reported health status.

When surveying opinions of patients about general practice care, since patients of different characteristics will give different views of the same service, the sample of patients whose views are surveyed should aim to account for these predictable differences. Decisions on changes in care provision should be based only on the results of patient opinion surveys in which the characteristics of the sample of patients responding to the survey were taken into account.