Eliciting Patient Preferences for Hepatocellular Carcinoma Screening: A Choice-Based Conjoint Analysis.

Abstract

Abbreviated MRI (AMRI), proposed as an alternative imaging modality for hepatocellular carcinoma screening, provides higher sensitivity than ultrasound. It is, however, unknown how patients weigh the higher sensitivity of AMRI against its higher cost and potentially less desirable testing experience. To assess patient preferences for hepatocellular carcinoma screening test attributes including sensitivity, false-positive rate, test-related anxiety, cost, and need for intravenous catheterization and contrast use, measured by choice-based conjoint analysis. This was an ancillary study to two prospective dual-center studies designed to compare the hepatocellular carcinoma detection rates by ultrasound versus AMRI. Of the 135 eligible participants, 106 (median age 63, range 25-85; 56% male) completed the choice-based conjoint analysis survey and were included in this substudy. Participants' preference for individual screening test attributes was assessed using a 12-item, web-based choice-based conjoint analysis survey administered in person at the screening visit. Conjoint analyses software and hierarchical Bayes random-effects logit model were used to calculate the relative importance of each attribute. The most important attribute driving patient preferences was higher test sensitivity (importance score 39.8%), followed by lower cost (importance score 22.8%) and lower false-positive rate (importance score 19.4%). The overall estimated participants' preference for ultrasound and AMRI were similar when assuming the same specificity for both modalities. Higher screening test sensitivity and lower cost were the leading patient preference drivers. This study has important implications for understanding patient preferences for specific screening test characteristics as potential determinants of adherence.