Table 2.

Regression estimates for part-worth utility

Attribute and level	Part-worth utility, mean	Part-worth utility, SD	Part-worth utility < 0
Proportion of individuals identified
40/100	−2.29*	2.24*	84.7%
60/100	Reference	-	-
80/100	1.14*	0.89*	10.1%
90/100	1.66*	1.94*	19.6%
Number of tests	0.05	0.40*	45.3%
Total wait time (Months)	−0.15*	0.16*	83.3%
Cost ($)	−0.0011*	-	-
Opt out of testing	−7.02*	-	-
Opt in for testing	0 (assumed)	7.90*	-

Part-worth utilities represent the marginal preference-based utilities associated with each attribute level. A positive mean estimate indicates that, on average, patients expressed positive personal utility for the attribute. A negative estimate indicates that, on average, the attribute caused disutility, or a reduction in well-being. Part-worth utilities can be summed to indicate the overall preference-based utility of a good and the ratio of any two part-worth utility estimates shows the marginal rate of substitution between attributes. The estimated SD characterizes the heterogeneity of individual part-worth preference-based utility values in the sampled population

SD  standard deviation

*p  < 0.05