Background: The 29-item 'Medical Interview Satisfaction Scale' (MISS-29) was developed in the USA to assess patient satisfaction with individual doctor-patient consultations. It has been used in studies from British general practice. However, there is limited evidence for its psychometric properties in this population.
Objectives: The present study was designed to examine the validity, reliability and applicability of the MISS-29 in British general practice populations.
Methods: The study was divided into two phases. The first investigated the properties of the MISS-29 in a UK general practice population and resulted in a modified MISS (MISS-21). The second investigated the properties of the MISS-21 in a wider UK general practice population. In phase 1, 150 patients over 16 years were recruited sequentially from patients attending a large group practice in suburban north London. Patients completed a questionnaire which collected demographic data and the MISS-29. In phase 2, 159 patients with a new problem were recruited from patients over 16 years consulting 18 GPs in north London, Essex and Suffolk. Patients completed a questionnaire while waiting to see the doctor; this collected demographic data and included six separate items, designed by the author, intended to measure patient satisfaction with previous consultations with the doctor. The patients completed the MISS-21 when they left the doctor's consulting room.
Results: The response rates for the phase 1 and phase 2 studies were 76.9 and 72.6%, respectively. Factor analysis, using principal component analysis with a varimax rotation, of the data collected in phase 1 resulted in a 21-item scale with the same four subscales as the original MISS-29. Correlations between subscales range from 0.46 to 0.65. Values of Cronbach's alpha between 0.67 and 0.92 suggest that the subscales are internally consistent under the conditions of the study. In phase 2, 92.1% completed all the items in the MISS-21 and there were no significant relationships between patients' demographic variables and the proportion of completed MISS-21 questionnaires, nor were there any differences in the proportion of completed MISS-21 questionnaires between type of practice or between practices serving different geographic populations, suggesting that the items were acceptable to patients. There were highly significant positive correlations (0.21-0.63) between scores on the MISS-21 and all aspects of satisfaction with previous consultations, providing supportive evidence for the construct validity of the MISS-21.
Conclusion: This study has demonstrated that the use in British general practice of the 29-item MISS developed in the USA should be treated with caution. However, a new 21-item version with the same four subscales as the 29-item MISS was developed which has satisfactory internal reliability. The correlations between subscales suggest that they represent fairly discrete but overlapping aspects of satisfaction. Evidence is produced suggesting that patients have less difficulty completing the MISS-21 and that it is applicable for assessing satisfaction with the consultation in different practice types and populations in the UK. Limited data supporting the construct validity of the MISS-21 are presented. While this study does not provide a full assessment of the MISS-21, we believe it provides evidence for its psychometric properties, which suggests that it is a valid and reliable instrument for the assessment of patient satisfaction with individual consultations in British general practice.