Development of an instrument to assess expectations of and preference for an insulin injection pen compared with the vial and syringe

Background: Before using a product, patients form expectations regarding the extent of a product's desirable attributes. These expectations can be used to understand their preference and anticipate potential satisfaction with the product. Objective: The aim of this study was to produce a valid and reliable data collection instrument (the Insulin Injection Preference questionnaire [IIP-q]) to measure expectations of and preference for the insulin injection pen compared with the vial and syringe. Methods: This study was initiated at the University of Mississippi (University, Mississippi). The IIP-q was developed to determine the extent to which respondents' prepurchase expectations of a product's attributes relate to preference for an insulin injection pen compared with the vial and syringe. Instrument development began with item generation related to product attributes important to patients who inject insulin. Items originated from an extensive search of the peer-reviewed Internet-based literature, marketing reports, clinical studies, and existing instruments. Content validity also was assessed using expert panel and focus group review. The resultant instrument (the IIP-q) was mailed to 1200 patients known to have type 1 or type 2 diabetes mellitus who either did or did not use insulin. Subscales were identified through exploratory factor analysis. Reliability and validity were assessed using Cronbach α for subscale items. Product-moment correlations between subscale dimensions and 2 global measures of preference were used to test the relationship between attribute expectations and preference. Results: Seventeen of the questionnaires were returned as undeliverable, leaving 1183 in the sample population. Questionnaires were received from 302 individuals, 55 of whom failed to complete ≥ 85% of the items and thus were not included in the final analysis. Of the 247 respondents (135 women, 112 men; mean [SD] age, 52.4 [13.2] years (range, 18–83 years]), 99 (40.1%) were current insulin usersand 143 (57.9%) were not using insulin. Exploratory factor analysis resulted in a 13-item solution (Cronbach α = 0.92), accounting for 73.6% of the total explained variance. Ease of use, activity interference, and social acceptability emerged as expectation subscales from exploratory factor analysis. Cronbach α for items comprising the subscales ranged from 0.82 to 0.92. The 3 subscales were significantly correlated with patient preference (ease of use, r = 0.520, P < 0.001; a