EXAMINING POTENTIAL DOUBLE COUNTING WHEN COMBINING PATIENT AND CARER QUALITY OF LIFE - A MEDIATION ANALYSIS
Author(s)
Jan Faller, MHealthEc1, Cathy Mihalopoulos, PhD1, Gang Chen, MSc, PhD2, Brendan Mulhern, PhD3, Lidia Engel, PhD1;
1Monash University, Melbourne, Australia, 2University of Melbourne, Melbourne, Australia, 3University of Technology Sydney, Sydney, Australia
1Monash University, Melbourne, Australia, 2University of Melbourne, Melbourne, Australia, 3University of Technology Sydney, Sydney, Australia
OBJECTIVES: Double counting becomes a potential issue when both care recipients’ and carers’ health-related quality-of-life (HRQoL) impacts are considered simultaneously in economic evaluations. This study explores the extent of double counting during HRQoL outcome assessment.
METHODS: Using data from the Child Health CheckPoint Study embedded in the Longitudinal Study of Australian Children (LSAC), we applied mediation analysis to evaluate hypothesised pathways described by the Basu and Meltzer framework. Three variables were used: 1) child health condition as exposure variable, 2) child HRQoL (Child Health Utility instrument [CHU9D]) as outcome variable, and 3) parent (carer) HRQoL (Assessment of Quality of Life [AQoL-8D]) as mediating variable. We hypothesised full mediation occurs if the exposure’s effect on the outcome disappears after incorporating the mediator, indicating double counting. Partial mediation occurs if the effect weakens, but remains. If the mediator has no effect, double counting is absent.
RESULTS: 1817 parent-child dyads were included. Three-quarters of the children had health conditions. Univariate analysis indicated a significant direct effect of the health condition on child HRQoL (β = -0.093, p<0.001). After including the mediator (parent HRQoL), the direct effect of the health condition on child HRQoL decreased (β = -0.077, p<0.01), suggesting parent HRQoL is a mediating variable. The relationship was classified as a partial mediation as the direct effect decreased, but was still significant.
CONCLUSIONS: This study found the presence of potential double counting of HRQoL outcomes, which is a concern when combining care recipient and carer HRQoL. Future research should use a longitudinal setting and with other outcome measures, including different items and construct, as well as in different care recipient populations to check for the robustness of the findings. Additional research to advise and arrive at a consensus regarding how to address double counting when aggregating care recipient and carer HRQoL is warranted.
METHODS: Using data from the Child Health CheckPoint Study embedded in the Longitudinal Study of Australian Children (LSAC), we applied mediation analysis to evaluate hypothesised pathways described by the Basu and Meltzer framework. Three variables were used: 1) child health condition as exposure variable, 2) child HRQoL (Child Health Utility instrument [CHU9D]) as outcome variable, and 3) parent (carer) HRQoL (Assessment of Quality of Life [AQoL-8D]) as mediating variable. We hypothesised full mediation occurs if the exposure’s effect on the outcome disappears after incorporating the mediator, indicating double counting. Partial mediation occurs if the effect weakens, but remains. If the mediator has no effect, double counting is absent.
RESULTS: 1817 parent-child dyads were included. Three-quarters of the children had health conditions. Univariate analysis indicated a significant direct effect of the health condition on child HRQoL (β = -0.093, p<0.001). After including the mediator (parent HRQoL), the direct effect of the health condition on child HRQoL decreased (β = -0.077, p<0.01), suggesting parent HRQoL is a mediating variable. The relationship was classified as a partial mediation as the direct effect decreased, but was still significant.
CONCLUSIONS: This study found the presence of potential double counting of HRQoL outcomes, which is a concern when combining care recipient and carer HRQoL. Future research should use a longitudinal setting and with other outcome measures, including different items and construct, as well as in different care recipient populations to check for the robustness of the findings. Additional research to advise and arrive at a consensus regarding how to address double counting when aggregating care recipient and carer HRQoL is warranted.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
MSR32
Topic
Methodological & Statistical Research
Topic Subcategory
PRO & Related Methods
Disease
No Additional Disease & Conditions/Specialized Treatment Areas