Study design
The economic burden associated with ADPKD was estimated for 2018 and comprised direct healthcare costs, direct non-healthcare costs, and indirect costs. Cost components were identified from scientific publications, governmental agencies, and non-governmental organizations. Costs were stratified by disease stage when information was available.
Costs associated with ADPKD were calculated using a prevalence-based approach with a prevalence of 0.043% and the estimated 2018 US adult population from the US Census Bureau [14, 15]. The number of individuals with ESRD-RRT due to a primary diagnosis of ADPKD in the US was derived from the United States Renal Data System (USRDS) [16]. The distribution of the prevalence of ADPKD for chronic kidney disease (CKD) stages 1–5 was estimated based on the relative proportion of individuals with each disease stage (excluding individuals with ESRD-RRT) based on the literature [17]. The incremental costs associated with ADPKD were estimated based on the average cost difference between an individual with ADPKD and an individual from the US population (comparison group), and thus constitute the “excess” costs attributed to ADPKD; similar methodology has been used in several previous publications [18,19,20,21,22,23,24,25,26]. All costs were expressed in 2018 US dollars, with direct healthcare costs adjusted using the Consumer Price Index for All Urban Consumers (CPI-U): Medical Care and direct non-healthcare and indirect costs adjusted using the CPI-U: All Items [27]. Adjustments for population growth were made for parameters estimated prior to 2018 using the population estimates from the US Census Bureau [15].
Direct healthcare costs
Excess direct healthcare costs associated with ADPKD were calculated based on the average annual excess direct healthcare costs incurred by an individual with ADPKD compared to those incurred by an individual from the US population, stratified by disease stage, multiplied by the estimated number of individuals with ADPKD by disease stage in the US [16, 17]. Excess direct healthcare costs incurred by payers for an insured individual were based on estimates from Knight et al. [10]. As reported by Knight et al., the comparison group of individuals from the US population consisted of a random sample of sex- and age-matched patients without polycystic kidney disease (PKD) or ADPKD [10]. To estimate the direct healthcare costs from a societal perspective, out-of-pocket expenditures were added as a proportion of the absolute costs based on estimates available in the literature for patients with non-dialysis dependent CKD, patients with ESRD, and the US population [16, 28,29,30]. In the absence of published estimates for the excess costs incurred by publicly insured and uninsured individuals with ADPKD, excess costs were assumed to be the same for privately and publicly insured individuals with ADPKD, and uninsured individuals with ADPKD were assumed to incur no excess costs compared to uninsured individuals in the US population. Excess direct healthcare costs by insurance type were weighted by the proportion of individuals by insurance type [16, 29] to obtain the estimated average excess direct healthcare costs for an individual with ADPKD compared to an individual from the US population, stratified by disease stage.
Direct non-healthcare costs
Direct non-healthcare cost components included research, training, advocacy, matching of donors and recipients for kidney transplant, and transportation to and from dialysis centers. Costs of research, training, and advocacy were drawn directly from the estimates of funding for research on PKD reported by the National Institute of Health (NIH) and from the annual report of the PKD Foundation [31, 32]. Costs of matching donors and recipients for kidney transplant among individuals with ADPKD were based on the Organ Procurement and Transplantation Network (OPTN) fees and the United Network for Organ Sharing (UNOS) fees multiplied by the number of individuals waiting for a kidney transplant due to ADPKD [16, 33,34,35]. Costs of transportation to and from dialysis centers among individuals with ADPKD in the US were based on the average transportation costs for individuals using in-center dialysis multiplied by the number of individuals with ADPKD using in-center dialysis in the US [12, 16]. The number of individuals with ADPKD using in-center dialysis in the US was computed as the number of individuals with ESRD-RRT requiring dialysis due to cystic kidney disease in the US multiplied by the proportion of individuals with ESRD-RRT due to ADPKD among individuals with cystic kidney disease and the proportion of individuals requiring in-center dialysis among individuals requiring dialysis due to cystic kidney disease [16]. Of note, it was assumed that direct non-healthcare costs incurred by patients with ADPKD would be zero if these individuals did not have ADPKD. Therefore, the direct non-healthcare costs associated with ADPKD were assumed to be equal to the excess direct non-healthcare costs.
Indirect costs
The studied components of indirect costs were reduced productivity at work, loss of productivity due to unemployment, loss of productivity from premature mortality, and caregiver burden, as previously identified by Cloutier et al. [18]. The latter included loss of productivity as a result of caregiving as well as incremental healthcare costs incurred by caregivers.
Unemployment
Costs associated with productivity loss from unemployment were estimated based on the excess number of unemployed individuals with ADPKD multiplied by the age-adjusted average annual wage in the US employed population [9, 16, 36]. The excess number of unemployed individuals with ADPKD was computed based on the number of individuals with ADPKD aged 15–64 years old by disease stage, the employment-to-population ratio in the ADPKD population by disease stage, and employment-to-population ratio in the US population given the average age of working-age individuals with ADPKD in each disease stage.
Reduced productivity at work
Reduced productivity at work was assessed based on the productivity weight in the ADPKD employed population by disease stage compared to the US population. Costs associated with this reduced productivity were further calculated based on the number of employed individuals with ADPKD by disease stage and the average annual wage in the US employed population given the average age of working-age individuals with ADPKD in each disease stage. The number of employed individuals with ADPKD was based on the number of individuals with ADPKD by disease stage and the employment-to-population ratio in the ADPKD population by disease stage [9, 36,37,38].
Premature mortality
Premature all-cause mortality is associated with a productivity loss for the society. These costs were estimated using the following components by age group: (1) all-cause excess mortality rate in the ADPKD population compared to the US general population [15, 39], (2) the average number of years of productive life (i.e., based on retirement age) in the US general population [40], and (3) the average annual wage in the US employed population [41]. In order to calculate the net present value of costs arising in the future, a 3% discount rate was applied to the productivity loss from premature all-cause mortality [42].
Caregiver burden
The productivity loss from caregiving was defined as the number of unpaid hours of care received by individuals with ADPKD [9]. The associated costs were calculated for each disease state by multiplying the number of hours lost by the average hourly wage in the US [36]. Of note, it was assumed that productivity loss due to caregiving would be zero if individuals with APDKD did not have ADPKD. Accordingly, for this component, the total costs were assumed to be equal to the excess costs.
Moreover, to obtain excess healthcare costs incurred by caregivers of individuals with ADPKD, the estimated excess direct healthcare costs incurred by a family caregiver compared to those of a non-caregiver were multiplied by the number of individuals with ADPKD receiving assistance from family caregivers by disease stage [29, 34, 43, 44]. The proportion of excess direct healthcare costs incurred by caregivers was derived from Albert et al. based on stratified analyses comparing caregivers and non-caregivers within groups defined by age, gender, and white-collar versus blue-collar status [43]. The excess direct healthcare costs incurred by a family caregiver were then computed by multiplying the proportion of excess costs among caregivers with the average direct all-cause healthcare costs per individual per year in the US population reported in the Medical Expenditure Panel Survey (MEPS) survey [29].
Sensitivity analyses
Four sensitivity analyses were conducted to assess the impact of using different sources and assumptions on the estimates. First, an adjustment was made to the estimated excess direct healthcare costs among publicly insured individuals to reflect potentially higher excess direct healthcare costs compared to those of commercially insured individuals [45]. Second, direct healthcare costs for uninsured individuals with ADPKD were estimated based on the average medical costs reported in the literature for the US uninsured population and the ratio of the direct healthcare costs for individuals with ADPKD versus without ADPKD, as measured in Knight et al. [10, 18, 46]. Third, the number of hours per year devoted to caregiving for an individual with ADPKD was varied based on estimates for the CKD population [13, 47]. Finally, the economic burden of ADPKD was calculated based on the distribution of the prevalence of ADPKD per disease stage from Neumann et al., without adjustment for the number of individuals with ESRD-RRT based on the USRDS [17].