Development of a French inpatient satisfaction questionnaire

Objective. To develop a brief French-language, generic, self-administered q uestionnaire to measure inpatient satisfaction. Design. Issues relevant to patients were identified using three open-ended questions designed in accordance with the disconfirmation paradigm. The con tent of patients' responses was analysed and then supplemented by items tak en from published instruments in order to generate a pool of 93 items. Twen ty-nine items were selected following a strict procedure. Content validity was judged by comparing the questionnaire to existing instruments. Construc t validity was supported by testing specific hypotheses derived from the li terature and by performing principal component analysis. Reliability was es timated by calculating Cronbach's alpha. Setting. A 2200-bed French teaching hospital. Subjects. A mail survey was carried out on a random sample of 1000 inpatien ts within 2-4 weeks of discharge. Eligible subjects were medical, surgical and obstetrics inpatients who had stayed in the hospital for more than 24 h ours. Results. The participation rate (71%) and the completion rate (95%) were in dicators of acceptability. There were modest differences between the questi onnaire and published instruments (financial aspects, amenities). Construct representation by principal component analysis consisted of six scales whi ch accounted for 58% of the variance in total satisfaction scores. The reli ability estimates of internal consistency ranged from 0.67 to 0.86. Conclusion. We propose that the self-administered multidimensional inpatien t satisfaction questionnaire provided encouraging preliminary psychometric information. This instrument is intended to involve patient feedback in a c ontinuous quality health care improvement strategy.