Measuring patient satisfaction with four items: validity of the client satisfaction questionnaire 4 in an outpatient population

INTRODUCTION

Patient satisfaction with mental health services has for several decades been considered an important component when evaluating service quality. It is often assessed in the context of monitoring quality of care, developing or evaluating newly implemented interventions or changes in practice. Because of this, patient satisfaction questionnaires are often added to longer questionnaire batteries, and shorter questionnaires are preferred to prevent respondent fatigue and non-compliance and to secure easy implementation. However, most studies use unvalidated patient satisfaction measures, making comparisons between studies difficult. Validation studies of short patient satisfaction measures are therefore warranted.

METHODS

The primary aim was to examine the construct validity and internal reliability of the Client Satisfaction Questionnaire-4 (CSQ-4) in a Norwegian outpatient mental health setting. A total of 467 patients were recruited from an outpatient psychiatric care clinic in Central Norway. The secondary aim was to examine an earlier proposed cutoff for classifying dissatisfied patients in this new population. A principal component analysis was conducted to evaluate factor structure, correlation analyses were conducted to test for predicted relationships, and Cronbach's alpha was calculated to examine internal reliability.

RESULTS

The CSQ-4 showed a clear unidimensional structure with one factor explaining 80% of its variance. Its internal reliability was very high, with a Cronbach's alpha of 0.92. As hypothesised this study found no statistically significant sex differences in satisfaction and no statistically significant association between age and satisfaction. Positive changes in symptoms during treatment and higher post-treatment functional impairment were associated with higher patient-reported treatment satisfaction scores, which indicates good construct validity.

CONCLUSION

This is the first study to evaluate the CSQ-4 in a psychiatric population. The CSQ-4 demonstrated good structural validity and internal reliability and was correlated with outcome variables in terms of symptom change and post-treatment functioning. In sum, this indicates that the CSQ-4 is a good short alternative for evaluating patient satisfaction in routine outpatient mental health care.