Abstract
BACKGROUND: Well-designed pharmaceutical pictograms may improve patients' understanding of medication instructions. However, the iterative participatory design process required to produce effective pictograms can be costly in terms of money, time, and effort. Crowdsourcing has been applied to bring down the costs of the participatory design process, but the feasibility of using this approach with older adults remains largely unknown. OBJECTIVES: To evaluate the feasibility of using Amazon Mechanical Turk (MTurk), a leading crowdsourcing platform, for participatory pictogram evaluation with older adults (55+) and to evaluate the comprehensibility of USP pictogram, identify common misinterpretations, and explore the relationship between selected participant characteristics and their pictogram comprehension performance. METHODS: 108 older adults (56.5% female; 57-80 years of age) were recruited via MTurk to complete a cross-sectional online survey that asked them to interpret 15 USP pictograms and answer questions about their health and health literacy. RESULTS: It was feasible to perform pictogram evaluation with older adults on MTurk, as shown by ease of recruitment and high data quality. Of the 15 pictograms tested, seven (46.7%) resulted in a comprehensibility score below the threshold established by the American National Standards Institute (ANSI), eight (53.3%) elicited common misinterpretations, and two (13.3%) resulted in ANSI-defined "critical confusion." Age (P = 0.04) was associated with pictogram comprehension performance. Certain issues with the pictogram subtitles emerged during the evaluation. CONCLUSIONS: MTurk is a feasible platform for participatory pictogram evaluation, even for a sole target of older adults. The USP should develop a pictogram user manual, redesign pictograms confusing to older adults, and establish policies and procedures to ensure that pictogram subtitles conform to evidence-based best practices and standards for patient-centered written drug information.