Development of a value-based scoring system for the MobQoL-7D: a novel tool for measuring quality-adjusted life years in the context of mobility impairment

Abstract

Purpose: To create a preference-based value set scoring system for the MobQoL-7D outcome measure, and to examine differences in the health state preferences of the general population and individuals with impaired mobility.

Methods and materials: A preference elicitation study was undertaken to ascribe utility weights to all health states (i.e., all unique combination of answers) described by the MobQoL-7D. The elicitation exercise was developed using the Online Elicitation of Personal Utility Functions (OPUF) tool. Two UK sample groups were recruited; firstly a representative general population sample (N = 504), secondly a balanced sample of individuals with impaired mobility (N = 368). Distinct preference-based value sets were developed for each sample. Differences in dimension ranking, weighting, and overall utility values were assessed.

Results: The general population sample considered most health states, especially the more severe states, to be worse than the mobility impaired sample comparatively. Statistically significant differences between the samples were observed in four of the seven MobQoL-7D dimensions.

Conclusions: This study is the first to provide preference-based value sets for the MobQoL-7D, ready for use in economic evaluations, QALY calculation, and other clinical or research applications. The study demonstrates how the general public and individuals with impaired mobility value health states differently.

Keywords: Disability; QALY; health-related quality of life; mobility impairment; patient reported outcomes; quality of life; utility.

Figure 1.

Frequency of utility values across all MobQoL-7D health states, by sample. On the X axis the utility space is plotted from 0 to 1, where 1 represents the best possible health state and 0 represents the worst possible health state. On the Y axis the frequency of health states is plotted. Comparing the MI sample depicted in yellow, and the GP sample depicted in purple, we see that the MI sample rated more health states higher than the GP sample, particularly between utility values of 0.50 and 0.75, and were less likely to rate health states as having utility values below 0.50.

Figure 2.

Distribution of MobQoL-7D utility values from best (i.e., 1111111) to worst (i.e., 4444444), by sample. On the X axis all 16,384 MobQoL-7D health states are plotted form the worst (i.e., 4444444) to the best (i.e., 1111111). On the Y axis the utility values for each health state are plotted, from 1 (highest possible utility value) to 0 (lowest possible utility value). Comparing the MI sample depicted in yellow, and the GP sample depicted in purple, we see that the MI sample rated most health states higher than the GP sample. The differences between the two samples become more apparent as the health states worsen. The GP data line appears smooth because health states are ordered along the x-axis according to the preference ordering of the GP sample; the MI data line shows variance (compared to GP sample) due to the different relative ranking of health states.

Figure A1.

Participants were asked to consider a pairwise comparison between the worst MobQoL-7D health state “444444” (scenario A; a state in which all seven dimensions are at the worst level) and ‘being dead’ (scenario B).

Figure A2.

If participants preferred the scenario A (the worst health state), over “being dead” in the S1task, participants were asked to locate the position of scenario A on a VAS between “no mobility problems” (=100) and “being dead” (=0).

Figure A3.

If participants indicated that they preferred “being dead” over scenario A in the S1 task, they were asked to locate the position of “being dead” on a VAS between “no mobility problems” (=100) and scenario A (=0).