Examination of Methods to Estimate Productivity Losses in an Economic Evaluation: Using Foodborne Illness as a Case Study

Background Cost-utility analyses commonly use two primary methods to value productivity: the human capital approach (HCA) and the friction cost approach (FCA). Another less frequently used method is the willingness-to-pay (WTP) approach, which estimates the monetary value individuals assign to avoiding an illness. In the context of foodborne illnesses (FBI), productivity loss represents one of the most significant economic impacts, particularly in developed nations. These losses arise from factors such as missed workdays, reduced workplace efficiency due to illness, and long-term health complications that can limit an individual’s ability to work. As a result, accurately quantifying productivity loss is critical in understanding the broader economic burden of FBI. Aim Our aim was to compare the impact of valuation methods used to measure productivity loss in an economic evaluation, using a hypothetical intervention for FBI caused by campylobacter as a case study. Cost effectiveness from three perspectives is examined: health care system, employee, and employer. Method A Markov model with a 10-year time horizon was developed to evaluate the morbidity and productivity impacts of FBI caused by campylobacter. The model included four health states: ‘healthy’, ‘acute gastroenteritis’, ‘irritable bowel syndrome and being unable to work some of the time’, and ‘irritable bowel syndrome and unable to work’. Five approaches to valuing productivity loss were compared: model 1 (cost-utility analysis), model 2 (HCA), model 3 (FCA), model 4 (FCA+WTP to avoid illness with paid sick leave), and model 5 (WTP to avoid illness without paid sick leave). Health outcomes and costs were discounted using a 5% discount rate. Costs were reported in 2024 Australian dollars ($AUD). Results Model 1, which did not include productivity losses, yielded the highest incremental cost-effectiveness ratio (ICER) at $56,467 per quality-adjusted life-year (QALY) gained. The inclusion of productivity costs (models 2–5) significantly increased the total costs in both arms of the models but led to a marked reduction in the ICERs. For example, model 2 (HCA) resulted in an ICER of $11,174/QALY gained, whereas model 3 (FCA) resulted in $21,136/QALY gained. Models 4 and 5, which included WTP approaches, had ICERs of $19,661/QALY gained and $24,773/QALY gained, respectively. Conclusion These findings underscore the significant impact of different modelling approaches to productivity loss on ICER estimates and consequently the decision to adopt a new policy or intervention. The choice of perspective in the analysis is critical, as it determines how the short-term and long-term productivity losses are accounted for and valued. This highlights the importance of carefully selecting and justifying the perspective and valuation methods used in economic evaluations to ensure informed and balanced policy decisions.