OBJECTIVES: New adjunctive drug therapies in severe hypercholesterolemia considerably reduce apheresis frequency and thus improve quality of life. So far it is unknown, if this reduction generates highest benefit for all patients. Prior analyses using mixed logit showed high standard deviations in patients’ preferences.

METHODS: A discrete choice experiment (DCE) analyzed the impact of treatment attributes on patients’ preferences and how preferences varied according to socio-demographic, attitudes or experiences. The DCE included seven attributes with three levels each. A latent class analysis was used to model heterogeneity in preferences.

RESULTS: N=348 patients participated. Based on the review of the model accuracy, a model with 3 classes was identified as suitable to reflect heterogeneities. For one class (N=126) "reduction of LDL-C level" (Level Difference/LD: 5.097) was most important and dominated all other attributes. This class focuses solely on this treatment outcome independent of apheresis frequency or additional injections. Preference patterns of the second class (N=106) are characterised by a focus on “frequency of apheresis” (LD: 2.323), "risk of myopathy" (LD: 1.577) and "reduction of LDL-C level" (LD: 0.893). Respondents clearly consider higher frequency of apheresis to have a negative impact. The objective ranking of attributes in the third class (N=116) is identical to the second one. However, the principle "the more the better" seems to apply - the highest frequency of apheresis is preferred most (LD: 0.750). These patients have adjusted to apheresis for more than ten years and prefer it.

CONCLUSIONS: The three groups show differences in the assessment of treatment attributes. As assumed, reduction of apheresis is important, but only for a sub-segment (30%) of patients. One third wants effective LDL-reduction by whatever means necessary. Most strikingly, another 30% postulate higher frequencies of apheresis and seem to reject adjunctive drug therapies even if outcomes are the same.