Search for a command to run...
Objective Patients with high treatment satisfaction are known to have better treatment compliance and persistence, and the Treatment Satisfaction Questionnaire for Medication (TSQM) is a questionnaire widely used to assess treatment satisfaction. A Japanese version of this questionnaire is available; however, its application is limited. In this study, we aimed to culturally adapt and assess the psychometric properties of a newly developed Japanese version of the TSQM in patients with diabetes or migraine who were prescribed oral and injectable medications in Japan.Methods The study included two phases; phase I, retranslation and cultural adaptation of the Japanese TSQM, and phase II, psychometric validation of the updated Japanese TSQM. Japanese patients with diabetes or migraine receiving oral and/or injection treatment were involved in the cognitive debriefing interview and psychometric validation. The retranslation and cultural adaptation of the Japanese TSQM were conducted following the process recommended by the ISPOR Task Force. The updated Japanese TSQM, finalized after amendments based on cognitive debriefing interviews, was distributed to patients for psychometric validation. The psychometric properties of the finalized Japanese TSQM were assessed using classical test and Rasch measurement theories.Results Minor modifications were made to the Japanese TSQM after cognitive debriefing interviews with eight patients. In total, 512 patients completed the final version of the Japanese TSQM. All versions of the Japanese TSQM had moderate reliability, with Cronbach's alpha ranging 0.6-0.9 for each domain and intraclass correlation coefficients for test-retest reliability between 0.5 and 0.6 for most domains. High item-convergent and item-discriminant validity were observed, in which all items had higher correlations with their respective domains than with other domains of the Japanese TSQM. The Rasch measurement theory added additional support to the reliability and validity of the measures as they fit the Rasch models. Additionally, high internal consistency reliability, unidimensionality of the domains, and scale invariance were supported by the values of the person separation index, Chi-square value of each domain, and differential item functioning across patient subgroups, respectively.Conclusion The Japanese TSQM was successfully updated by revising the complex wording and removing wording specific to oral medications, making it usable for patients regardless of medication type. The updated Japanese TSQM had sufficient reliability and validity and could be used regardless of the disease and mode of administration of medication.