Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

2007 HSR&D National Meeting Abstract

Printable View

National Meeting 2007

1039 — Overcoming the Limitations of Health States in Decision Analysis

Lenert LA (VA San Diego Healthcare System) , Sturley A (San Diego Hospice), Athale N (VA San Diego Healthcare System)

Objectives:
Methods for elicitation of preference weights for cost-effectiveness-analysis (CEA) ask subjects to imagine and rate states of health that continue unchanged until death. This abstraction is inherently unrealistic, particularly in the setting of temporary health conditions or clinical disorders where progression is the norm. It also leads to complexities due to large numbers of states that need to be valued in CEA models. The objective of this study was to assess the feasibility of measuring values using descriptions of life courses where disease effects wax and wane over time.

Methods:
We developed patient-centered descriptions of the quality of life effects of hypertension starting with diagnosis and proceeding with treatment and ending with complications (strokes) leading to death. We embedded these in a multimedia utility elicitation program (iMPACT4). We then recruited a convenience sample of 229 veterans with hypertension from general medicine clinics and measured each subject’s values for the life courses using a visual analog scale (VAS), a time trade-off variant (TTO-LC) and the standard gamble. We then examined the reliability, logic errors, and the responsiveness of elicitations.

Results:
VAS, TTO-LC, and SG elicitations were all reliable. Correlation coefficients among similar states ranged from 0.75 for the VAS to 0.91 for the TTO-LC. The rate of logical errors ranged from 10.6% -13% for the VAS to 1%-2.6% for the TTO and 9.7%-13% for the SG. VAS, TTO, and SG ratings were all responsive (significantly different) in ratings of life courses lower overall quality of life due to more rapid progression of illness and higher adherence effort. VAS and TTO ratings also were responsive to ongoing reductions in quality of life due to mild and moderate medication side-effects.

Implications:
The psychometrics of utility elicitations using life course methods are generally good. The optimal procedure for assessment for utility assessment for life courses appears to be the TTO-LC.

Impacts:
Use of life-course methods may make preference weights in CEAs more realistic, as a result making the results more valid. Simpler methods for describing the health impacts of disease may make it easier to tailor health care resource allocations to the preferences of veterans using CEA.