Mixture models for consumers’ preferences in healthcare

This paper aims to explain preferences behaviour by a sample of Sardinia residents with respect to
combined choice of attributes related to cardiology services. The rating of proposed cards, containing
a combination of several attributes to qualify the services, are examined in terms of intrinsic
components and main drivers to determine the ordinal choice: location, screening mode, cost, waiting
time for the visit and subjects’ covariates. The topic is relevant in telemedicine as experienced in
Sardinia, a region with a mobility and a socio-economic disadvantage. The innovative approach allows
for effective visual support to interpret and compare results and it is useful also to predict the
respondents’ profile with respect to their individual characteristics. Empirical evidence supports policy
interventions and suggests the usefulness of the implemented statistical procedure.