The Preventable Productivity Burden of Kidney Disease in Australia

Abstract

Background: Kidney disease is associated with impaired work productivity. However, the collective effect of missed work days, reduced output at work, and early withdrawal from the workforce is rarely considered in health-economic evaluations.

Methods: To determine the effect on work productivity of preventing incident cases of kidney disease, using the novel measure "productivity-adjusted life year" (PALY), we constructed a dynamic life table model for the Australian working-age population (aged 15-69 years) over 10 years (2020-2029), stratified by kidney-disease status. Input data, including productivity estimates, were sourced from the literature. We ascribed a financial value to the PALY metric in terms of gross domestic product (GDP) per equivalent full-time worker and assessed the total number of years lived, total PALYs, and broader economic costs (GDP per PALY). We repeated the model simulation, assuming a reduced kidney-disease incidence; the differences reflected the effects of preventing new kidney-disease cases. Outcomes were discounted by 5% annually.

Results: Our projections indicate that, from 2020 to 2029, the estimated number of new kidney-disease cases will exceed 161,000. Preventing 10% of new cases of kidney disease during this period would result in >300 premature deaths averted and approximately 550 years of life and 7600 PALYs saved-equivalent to a savings of US$1.1 billion in GDP or US$67,000 per new case avoided.

Conclusions: Pursuing a relatively modest target for preventing kidney disease in Australia may prolong years of life lived and increase productive life years, resulting in substantial economic benefit. Our findings highlight the need for investment in preventive measures to reduce future cases of kidney disease.

Keywords: burden; chronic kidney disease; dynamic life table model; end stage kidney disease; incidence; productivity.

Graphical abstract

Figure 1.

Markov model of the possible transitions and health states. The Australian population of working age was separated into those with and without kidney disease. People with kidney disease were further separated into those with stage 3 and 4–5 CKD and ESKD. Stage 3 CKD was defined as an eGFR of <60–30 ml/min. Stage 4–5 CKD was defined as an eGFR of <30 ml/min, but not undergoing RRT. ESKD was defined as those with an eGFR of <15 ml/min (stage 5 CKD) and undergoing RRT. With each annual cycle, those in the “alive without kidney disease” health state were at risk of developing CKD or dying, whereas those in the “alive with Stage 3 CKD” health state could remain alive, develop stage 4–5 CKD, or die. Those in the “alive with stage 4–5 CKD (but not ESKD)” health state could remain alive, develop ESKD, or die. Those in the “alive with ESKD” health state could remain alive or die.

Figure 2.

The impact of kidney disease on productivity in the Australian population of working age over the next 10 years (2020-2029), which amounts to US$91.2 billion in GDP. The economic burden is driven by kidney disease-related presenteeism, followed by early labor-force withdrawal. To assess each component (premature mortality, early labor-force withdrawal, absenteeism, and presenteeism), one factor at a time were inputted into the model while the others were “switched off.” The results were compared with the analysis that assumed kidney disease did not exist. Technical details are provided in Supplemental Table 3. All analyses were estimated in 2020 Australian dollars and converted into US dollars using the purchasing power parity rate of 1.440 (2019 data).

Figure 3.

Sensitivity analyses assessing the effect of variation in key model parameters on the number of PALYs, compared with the base-case analysis. The tornado diagram demonstrates the model is most sensitive to variations in the increased risk of early labor force withdrawal associated with kidney disease. Black bars denote the lower-bound estimate has been used, whereas the white bars indicate use of the upper-bound estimate of the parameter. Bars represent the difference in the number of PALYs compared with the base case. In the base-case analysis, the total PALYs among working age Australians in 2020–2029 was 89,783,023 (Table 2). Note that due to a lack of available data for the intervals, the HR of early labor-force withdrawal due to CKD (as reported by Klarenbach et al.) was assumed to vary by 30%.