Purpose: Treatment for hepatitis C (HCV) is a complex, value-based, decision. Studies have demonstrated the potential value of using adaptive conjoint analysis (ACA) as a tool to support shared medical decision-making (SDM); however, there is little known regarding the psychometric properties of ACA. The purpose of this study was to assess the criterion validity of this decision support tool. Methods: We recruited consecutive patients eligible for treatment of HCV. Baseline data were collected in face-to-face interviews with a research assistant. Participants then completed an ACA task designed to help patients evaluate the pros and cons related to treatment of HCV with pegylated interferon and ribavirin before seeing their physician. Attributes for the ACA questionnaire were chosen based on patient testimonials obtained from focus groups. Preferences were measured for two choices: 1. Treatment associated with mild side effects versus no treatment, and 2. Treatment associated with severe side effects versus no treatment. Criterion validity was assessed by measuring the association between preferences predicted by ACA and the treatment plan, as well as by ascertaining the correlations between preferences predicted by ACA and patient values. Results: 65 patients have been evaluated to date; mean age = 53±5 (range 34-64); 99% male; 55% Caucasian; 59% were college educated; 31% were currently employed; and 23% reported having very good or excellent health status. Patients strongly preferring treatment (as determined by the ACA questionnaire) were more likely to be prescribed treatment during the subsequent visit whether treatment was associated with mild (62% versus 9%, p=0.001) or severe side effects (85% versus 13%, p=0.001). Correlations between baseline measures and preferences also support the validity of this tool and are presented in the table below.

 

	Treatment with Mild Side Effects		Treatment with Severe Side Effects
	Correlation Coefficient	P value	Correlation Coefficient	P value
Choice Predisposition	0.63	0.0001	0.6	0.0001
Perceived risk of cirrhosis	0.26	0.03	0.31	0.01
Importance of flu symptoms	-0.36	0.003	-0.39	0.001
Importance of fatigue	-0.35	0.005	-0.37	0.002
Importance of depression	-0.27	0.03	-0.44	0.0003
Uncertainty	-0.26	0.04	-0.32	0.009
Value clarity	-0.26	0.03	-0.27	0.03

Conclusions: The above data provide some support for both the convergent and predictive validity of ACA as a tool to facilitate SDM for complex medical decisions such as treatment for HCV.