THE IMPACT OF SF-6D HEALTH UTILITY ON PRODUCTIVITY LOSS: A POPULATION-BASED OBSERVATIONAL STUDY
Author(s)
An Tran-Duy, MSc, PhD.
Associate Professor, University of Melbourne, Melbourne, Australia.
Associate Professor, University of Melbourne, Melbourne, Australia.
OBJECTIVES: Productivity losses are an important component of the societal perspective in economic evaluation, yet they are often not measured in health research. This study examined the relationship between SF-6D health utility and productivity loss among Australian workers and explored the potential for estimating productivity costs when direct measures are unavailable.
METHODS: Data were obtained from the Household, Income and Labour Dynamics in Australia (HILDA) Survey. Productivity loss was measured as absenteeism (sick-leave days in the previous 12 months) and presenteeism (days worked while unwell in the previous four weeks). Absenteeism analyses used an unbalanced panel of 12,481 individuals (39,639 observations) from Waves 18-22. Due to excess zeros, absenteeism was modelled using a two-part approach comprising correlated random-effects (Mundlak) panel logit and negative binomial models, combined to estimate expected sick-leave days. Presenteeism data, available in Wave 21 only, were analysed using a negative binomial model. Models were adjusted for demographic, socioeconomic, occupational, behavioural, and health-related factors.
RESULTS: Higher SF-6D was significantly associated with lower absenteeism and presenteeism. A one-unit increase in SF-6D was associated with reduced odds of any sick leave (OR 0.23), fewer sick-leave days among those taking sick leave (IRR 0.46), and lower presenteeism (IRR 0.17). A 0.1-point increase in SF-6D was associated with a 14% reduction in odds of sick leave, an 8% reduction in sick-leave days among those taking sick leave, 0.24 fewer expected sick-leave days, and a 16% reduction in presenteeism. Corresponding reductions for a 0.5-point increase were 52%, 32%, 1.04 days, and 59%, respectively. Long-term health conditions, full-time employment, age, and current smoking were important predictors of productivity loss.
CONCLUSIONS: Using nationally representative Australian data, this study demonstrates a strong association between SF-6D health utility and productivity loss. The findings provide a basis for indirectly estimating productivity costs in economic evaluations when direct measures are unavailable.
METHODS: Data were obtained from the Household, Income and Labour Dynamics in Australia (HILDA) Survey. Productivity loss was measured as absenteeism (sick-leave days in the previous 12 months) and presenteeism (days worked while unwell in the previous four weeks). Absenteeism analyses used an unbalanced panel of 12,481 individuals (39,639 observations) from Waves 18-22. Due to excess zeros, absenteeism was modelled using a two-part approach comprising correlated random-effects (Mundlak) panel logit and negative binomial models, combined to estimate expected sick-leave days. Presenteeism data, available in Wave 21 only, were analysed using a negative binomial model. Models were adjusted for demographic, socioeconomic, occupational, behavioural, and health-related factors.
RESULTS: Higher SF-6D was significantly associated with lower absenteeism and presenteeism. A one-unit increase in SF-6D was associated with reduced odds of any sick leave (OR 0.23), fewer sick-leave days among those taking sick leave (IRR 0.46), and lower presenteeism (IRR 0.17). A 0.1-point increase in SF-6D was associated with a 14% reduction in odds of sick leave, an 8% reduction in sick-leave days among those taking sick leave, 0.24 fewer expected sick-leave days, and a 16% reduction in presenteeism. Corresponding reductions for a 0.5-point increase were 52%, 32%, 1.04 days, and 59%, respectively. Long-term health conditions, full-time employment, age, and current smoking were important predictors of productivity loss.
CONCLUSIONS: Using nationally representative Australian data, this study demonstrates a strong association between SF-6D health utility and productivity loss. The findings provide a basis for indirectly estimating productivity costs in economic evaluations when direct measures are unavailable.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE145
Topic
Economic Evaluation, Study Approaches
Topic Subcategory
Cost/Cost of Illness/Resource Use Studies, Work & Home Productivity - Indirect Costs
Disease
No Additional Disease & Conditions/Specialized Treatment Areas