Abstracts

Rationale: Nocebo side effects (the undesireable counterpart to the positive placebo response) arising from negative patient expectations of their medication may lead to nonadherence and has caused dropout from controlled clinical trials (Preston, 2000). Patient perception of medication effectiveness and tolerability may be influenced by pill color (Srivastava, 2010). Whether patients with epilepsy have strong pill color preferences that could affect medication compliance, either in routine care or clinical drug trials, is unknown. We examined the acceptability and order of preference for a range of capsule colors in patients with epilepsy. Methods: We surveyed adult patients with epilepsy who were receiving their care at the University of Maryland Medical Center. Patients were shown standard AA size capsules (Capsugel BDcaps; Greenwood, SC) in five global colors (white, yellow, gray, caramel, maroon) and asked to select any color(s) considered 'unacceptable'. They were then instructed to rank-order their preference for all five colors and were allowed to give multiple colors the same rank.Results: Capsule color surveys were completed in 80 adult patients with epilepsy. There was an avoidance preference (Cochran Q test, p<.001) with patients showing the greatest avoidance, about equal, for gray, caramel, and maroon (G 17/80, 21%; C 15/80, 19%; M 16/80, 20%); few patients found white or yellow to be unacceptable (W 4/80, 5%; Y 3/80, 4%). These results were similarly reflected in the rank-order of colors in which patients demonstrated a clear preference for some colors (Fr. Test; p<0.0001). More patients preferred white or yellow than preferred gray, caramel, or maroon (Figure 1). From Dunn s multiple comparisons, white and yellow were equally favorably preferred. Preference for yellow and maroon were indistinguishable, though preference for white and maroon differed significantly. Maroon, gray, and caramel were equally not preferred (Figure 2). Overall, there was consistency in patient preference for white and yellow where both colors were equally preferred. Meanwhile, maroon, gray, and caramel were not preferred and most frequently unacceptable. Interestingly, maroon was a controversial color, with relatively large numbers of patients selecting it as most preferred but also with an equally large number ranking it as least preferred (Figure 1). These differences in patient opinion were not associated with gender, age, or race.Conclusions: We found strong pill color preferences in patients with epilepsy, with some colors even considered 'unacceptable' by a proportion of patients. This finding may play a practical role in the marketing of antiepileptic drugs and in the design of clinical drug trials with a common aim to improve patient compliance by reducing the likelihood of a nocebo effect related to drug appearance.