Aim: There is a growing literature comparing the application of the Human Capital Approach (HCA) and the Friction Cost Approach (FCA) to health economic calculations of productivity losses. However, the implications of choosing one model over another in comparing sub-groups of individuals (e.g. by gender or age) have not been examined. This study calculated the lost productivity associated with head and neck cancer (HNC) using both the HCA and FCA, and examined the implications of using each approach for the comparison of socio-demographic and clinical groups.
Methods: The study used survey data from a sample of individuals in Ireland employed at the time of HNC diagnosis. The survey included questions about the nature of work and working hours pre- and post-diagnosis. These data were combined with average wage rates specified by age and occupation type in Ireland from the Central Statistics Office. The costs of temporary and permanent work absence, reduced working hours and premature mortality were estimated up to age 65 years using both the HCA and FCA. National Cancer Registry Ireland data and Irish life tables were used to estimate premature mortality rates associated with HNC. National data sources were used to identify estimates of workforce participation and wage growth rates. Friction periods were assigned based on occupational group, and ranged from 9.9 to 13.3 weeks. Future costs were discounted at 4%. Comparisons by socio-demographic and clinical variables, including gender, age, occupation, medical card status, cancer stage and treatment, were performed to examine the impact of selecting one method over another on group differences.
Results: There were 251 survivors of head and neck cancer in the survey sample who were employed at the time of diagnosis. Total productivity losses per person of working age and employed at the time of HNC diagnosis were €253,800 using the HCA and €6,800 using the FCA. The primary driver of productively losses in the HCA was premature mortality (38% of total loss), while in the FCA it was temporary time off (73%). When comparing sub-groups of socio-demographic and clinical variables, the FCA produced proportionately smaller differences between groups than the HCA for age and education. For example, using HCA older individuals (over 50 years) had 52% the productivity losses of younger individuals (50 years and under), while it was 114% using FCA. In contrast, gender, occupation type, medical card status and radiotherapy status had FCA results which accentuated differences between the subgroups, in comparison with the HCA results. For example, females had 99% the productivity losses of males using HCA, but 85% using FCA.
Conclusions: The productivity losses following HNC diagnosis can be significant, and as expected the methodology used has a large impact on the overall estimate. However, the methodology selected may also influence the observed differences between the productivity impacts of cancer on different socio-demographic and clinical groups. The FCA may provide more generalisable results, while the HCA may be more appropriate for comparing sub-groups. The purpose of calculating productivity losses should inform the choice of methodology.