Sample
Questionnaires were given to 1000 consecutive patients attending one inner-London general practice with a multicultural patient group containing high numbers of black and Vietnamese patients. This sample size was chosen to enable a final sample, which would enable analysis by ethnic group and would be sufficient to detect a significant difference in experiences and expectations between ethnic groups.
Patients were excluded if they were blind, identified as having a serious mental health problem that would preclude them from completing the questionnaire, or aged under 16 years. Vietnamese patients were given a questionnaire that had been translated by the practice interpreter into Vietnamese. Patients who could not complete the questionnaire on their own were helped by the interpreter or by a receptionist.
Completed questionnaires were received from 604 patients (response rate = 60.4%); 240 patients refused to complete the questionnaire — this was mostly due either to a lack of time or to a sense of unfamiliarity with the research process; 156 questionnaires were either not completed to a satisfactory level or had been discarded by the patient.
Measure
Accessing non-English-speaking patients is problematic and each methodology has its weaknesses. Although an interview enables the individual to chose their own agenda and frame their views using their own language and perspective, such a methodology requires an interpreter and several stages of translation that can transform the interviewees' original meaning. In contrast, a questionnaire constrains the participant in their answers and requires the agenda to be set by the researcher. However, although this approach requires some translation, the final quantitative data is more open to objective interpretation and is less contaminated by problems of language. For this reason the present study used a quantitative methodology involving a questionnaire.
There are several existing measures of patient experiences of care in the form of satisfaction scales23-25 and those specific to particular aspects of the consultation such as patient centredness.16,26,27 The present study did not use one of these existing validated measures as we were interested in measuring both experiences and expectations in a way that would enable these two different constructs to be directly compared. This required a new questionnaire in which questions relating to experiences could be matched with questions relating to expectations. In line with this, we developed a new measure designed with this specific requirement in mind.
This new measure was based upon the existing literature concerning what constitutes quality general practice17,21,22 and reflected the recent emphasis on shared decision making, patient choice and good communication. The questionnaire was translated into Vietnamese by the practice's interpreter. The English and the Vietnamese versions were then given to a small sample of patients for their feedback on the language used and the face validity of the questions. The reliability of the measure was assessed using Cronbach's α and total scores were created by summating the items. All Cronbach's α results were higher than 0.7, indicating an acceptable level of internal reliability.
Experiences of care. Patients were asked ‘How often have you experienced the following in a consultation with your GP?’. They were then asked to rate a series of 16 statements using a five-point Likert scale ranging from ‘never’ (1) to ‘very often’ (5). These statements were summated to reflect five main constructs:
Treatment — three items; for example, ‘felt that the GP prevented you from having the medicines when you wanted them’;
Communication — three items, for example, ‘felt that your real reason for coming had not been discussed’;
Consistency — three items, for example, ‘felt that your GP has prescribed different medicines compared with other GPs you have seen’;
Patients' agenda — three items, for example, ‘felt that the GP did not take all of your symptoms seriously’;
Patients' choice — four items, for example, ‘felt that the GP prevented you having the medicine of your choice’.
Higher scores reflected experiences of general practice involving less control over treatment options, poorer communication, more inconsistent medical practice, less focus on the patients' own agenda and less respect for the patients' choices.
What patients expect from a consultation.Using a five-point Likert scale ranging from ‘not at all’ (1) to ‘totally’ (5), patients rated a series of 16 matched statements relating to the same five aspects of care following the statement: ‘To what extent do you want your GP to do the following’:
Treatment — three items, for example, ‘felt that the doctor prevented you from having the medicines when you wanted them’;
Communication — three items, for example, ‘search for the real reason you have come’;
Consistency — three items, for example, ‘prescribe the same medicines for your symptoms as other doctors you have seen’;
Patient agenda — three items, for example, ‘take all of your symptoms as equally seriously’;
Patient choice — four items, for example, ‘allow you to chose the medicine you want’.
Higher scores reflected an expectation of greater control over treatment options, clearer communication, more consistent medical practice, a focus on the patients' own perspective and a respect for the patients' choices.
Profile characteristics. Patients described their profile characteristics in terms of age (years), sex, whether English was their first language (yes/no), ethnic group (white British, white European, black African, black African Caribbean, Vietnamese, Chinese, other) and whether they used an interpreter when seeing the GP (yes/no, if yes: family/friend/interpreter).
