Evaluating a game-based randomized experiment to increase melanoma identification among adults living in the U.S.

Background: Although cutaneous melanoma diagnoses are rising, morbidity and mortality can be reduced through early detection. This investigation seeks to improve melanoma identification accuracy, attitudes, and intentions among a lay population by comparing the effectiveness of different melanoma identification training strategies and the effect of real-time decisional feedback on a melanoma identification task. We developed an innovative, game-based approach and hypothesize differences among frequently used melanoma identification training modalities (i.e, the Asymmetry/Border/Color/Diameter [ABCD] rule, the Ugly Duckling Rule [UDS], and a modality that combines them both, ABCDF (where the F stands for 'funny looking"), and investigate differences in types of immediate feedback on a melanoma identification task.

Methods: We conducted a national online randomized experiment to test a 4 (melanoma training strategies: ABCD, UDS, ABCD-F, control) × 3 (feedback: Dermatological, Dermatological + Motivational, control) factorial design on melanoma identification, skin cancer beliefs (perceived susceptibility, severity, response efficacy, self-efficacy), attitudes, and prevention intentions.

Results: ABCD training (p < .001) and UDS training (p = .05) resulted in significantly higher melanoma identification than the control. All training types resulted in significantly higher self-efficacy than the control (p = .02). Both Dermatological (p = .02) and Dermatological + Motivational feedback (p = .01) elicited significantly lower melanoma identification than the control condition, although this effect may be due to differences observed among participants who received UDS training. There was a significant main effect of feedback on self-efficacy (p = .002), where both Dermatological and Dermatological + Motivational feedback elicited higher levels of self-efficacy than the control.

Conclusions: Our results suggest that game-based ABCD and UDS training strategies could increase melanoma identification accuracy. Real-time feedback reduced accuracy, but was associated with increased self-efficacy related to melanoma detection outcomes.