OBJECTIVES

Standardised integration of productivity costs into health economic evaluations is hindered by equity and distributional concerns. Our aim was to explore the distributive impact of productivity cost methodological variation by socio-demographic characteristics, describing the consequences for different groups.

METHODS

527 prostate cancer survivors (2-5 years post diagnosis) completed questions on work patterns since diagnosis. Productivity loss, categorised into temporary/permanent absenteeism, reduced hours and presenteeism, was costed in €2012. Valuation approaches included -the human capital approach (HCA), the friction cost approach (FCA) and the wage multiplier approach (WMA). Both national and self-reported wages were used. Costs were compared across socio-demographic and economic characteristics using non-parametric tests.

RESULTS

The estimated base case (HCA, using national wages) total productivity cost was €44,201 per prostate cancer survivor. Permanent absenteeism accounted for the largest cost (€18,537), followed by reduced work hours (€11,130), presenteeism (€8,148) and temporary absenteeism (€6,386). Productivity costs estimated according to the alternative approaches deviated widely from the base case, ranging between -89% (FCA: €4,625) to +49% (WMA: €65,764). Higher productivity costs were estimated compared to the base case using self-reported wages (€53,678); this finding was consistent across all valuation approaches. Statistically significant differences in productivity cost were found across four of the six survivor socio-demographic and economic characteristics by valuation approach, despite no significant difference in their physical unit equivalents.

CONCLUSIONS

Our results indicate that the distributional impact of productivity costs varies by socio-economic and demographic characteristic, in addition to valuation approach. We advocate that: productivity loss should be reported in physical units where possible; cost estimation should be subject to sensitivity analysis; and only where this is not feasible, that the HCA and national wages be used to value productivity loss where equity concerns are paramount.