Assessing satisfaction with pain medication in primary care patients:: Development and psychometric validation of a new measure

Background: The measurement of patient satisfaction with pain medication (SPM) is a potentially useful aid for health care decision-making, but no validated measures for SPM are known. Objective: This study aimed to develop an instrument to assess this patient-reported outcome in primary care and to evaluate whether it satisfactorily fulfilled the required psychometric properties (ie, reliability, validity, and sensitivity to change). Methods: The measure's content was obtained from literature reviews, focus groups, and expert opinion. A preliminary version of 14 self-administered items was obtained and tested in a prospective study in patients receiving pain medication. Item-total statistics and factor analysis were performed to obtain the final version. The final version was psychometrically validated by assessing feasibility, internal consistency (Cronbach's α), convergent validity (multivariant methods), discriminant validity in patients presenting pain relief (receiver operating characteristic curves), test-retest reliability (using the intraclass correlation coefficient [ICC] in patients maintaining medication), and sensitivity to change (in patients changing medication). Results: A total of 1119 patients were recruited (626 women [55.9%]; mean [SD] age, 47.6 [15.36] years; primarily suffering musculoskeletal injury [50.7%]). Four items were deleted from the preliminary version owing to low item-scale correlation and/or factor loadings. The final factor analysis confirmed a 4-factor solution, which explained 81.4% of the variance in questionnaire scores. Adverse events, speed/duration of effect, functional benefit, and overall satisfaction dimensions were identified. Cronbach's α and ICC for the 10-item final version were >0.80 for the summary score and all dimensions. Pain-related characteristics (intensity, frequency, and degree of pain relief), but not patient-related characteristics, were independently associated with the summary score. The area under the curve was 0.78 for the summary score. Effect sizes and standardized response mean were ≥0.84 for the summary score and all dimensions. Conclusion: The SPM questionnaire appears to have good acceptability as well as satisfcctory psychometric properties, based on these analyses. Copyright © 2004 Excerpta Medica, Inc.