Psychometric Evaluation of Instruments Used to Measure Patient Satisfaction with Pharmacist Services

OBJECTIVES: Evaluate the practical and psychometric properties of the instruments used to measure patient satisfaction with pharmacist cognitive services, including medication therapy management and patient counseling. METHODS: After a comprehensive literature review, instruments were included based on: at least one peer-reviewed publication using the instrument, reference lists, instrument availability, and a focus on pharmacist cognitive services. The six measurement criteria used were: practical features (administration time), breadth (multidimensional construct), depth (floor and ceiling effects), internal consistency (coefficient alpha > 0.7) and test-retest reliability (r > 0.7), and validity. Measurement evaluation was based on McHorney and Tarlov’s (1995) criteria for evaluation of outcome measurements and the recent Food and Drug Administration guide on patient reported outcome measures (2006). RESULTS: Of the 22 instruments identified, five were excluded because satisfaction was not measured as a multidimensional construct. Of the remaining 17 instruments, none met all six study criteria. A majority of the studies focused on patients’ interaction with the pharmacist in a community pharmacy setting. Instruments were notable for the lack of reported psychometric data, especially test-retest reliability, which was reported in only one instrument. Only nine of 17 instruments reported internal consistency measures, but the nine reporting met study criteria (alpha > 0.7). The series of instruments developed by MacKeigan and Larson (1989, 1994, 2002) were supported by adequate reliability data (coefficient alpha > 0.7) and multiple validation studies. Overall, when data was available, in a majority of the cases, measures met study standards. CONCLUSION: Evaluation of patient satisfaction with pharmacist cognitive services is critical to optimizing patient care in a clinical or community setting. To accomplish this, precise and valid measures are vital; however, no one instrument met all criteria. Lack of reported data warrants further research documenting an instrument’s psychometric properties, ease of administration, and ability to capture the multidimensional nature of patient satisfaction.