Influence of Reimbursement Policies on Dialysis Modality Distribution around the World

Population aging and increasing prevalence of diabetes worldwide are fueling the growth in numbers of people requiring dialysis round the world. The growth is most marked in lower-income countries, where there are often inadequate resources to screen for, and manage earlier stages of CKD (1). The cost of the global dialysis market is huge and was valued at $87 billion in 2016, with a projection of reaching $123 billion by 2023 (2). Given limitations to health care resources in all countries, it is important to understand what determines the relative use of hemodialysis (HD) and peritoneal dialysis (PD), given the lower costs of PD and its advantages for patient independence and experience. It has been recognized for a long time that in higher-income countries, the considerable variation in PD use between different countries can partially be explained by the health care system, with PD being utilized more in public compared with private health care systems (3). Determinants of dialysis modality distribution, however, are very complex. At the same time as PD use is declining in many Western health care systems despite its economic advantage, there is a significant growth in Asia as countries become more affluent (4,5), and use of PD is highest in countries that have a “PD first” (e.g., Hong Kong, Thailand) or “PD favored” (e.g., United States, Canada, China) policy (6). In low-income countries, particularly those in Africa, availability of dialysis of any type is very limited, and paradoxically there are economic advantages to HD as staff costs are low and PD fluid not only has to be imported, but also has to be transported long distances often across several country borders (7). It is helpful to divide potential factors favoring individual dialysis modality selection into those directly related to the individual treatment facility, and those related to the broader economic and political environment, as shown in Figure 1.Figure 1.: Dialysis modality distribution is determined by factors at individual facility level and factors related to the economic and political environment.The study by van der Tol et al. (8) in this issue of the Clinical Journal of the American Society of Nephrology is an important one because it attempts to assess the impact of government reimbursement for HD and PD and incentives for self-care dialysis on actual prevalence of the different modalities, using current financial data. The study includes both high- and lower-income countries, including those in which dialysis has only become more readily accessible in the past few years. It is not surprising that information about government support for dialysis is not readily available for many countries. The authors therefore used a rather unorthodox approach to extract the data: namely, asking high-profile nephrologists worldwide to assess health system payment policies and government reimbursement for their country. Heads of nephrology divisions of university hospitals who were considered to be opinion leaders in 94 countries were personally invited by one of the study investigators, with one nephrologist being selected for each country. Annual costs for reimbursement of HD were calculated using the amount refunded for one public-sector hospital HD session by 156 (3×52), unless reimbursement was only for a limited number of sessions (e.g., 90/yr in the Philippines). Annual reimbursement for PD was on the basis of 365×1 day of continuous ambulatory PD. Other modalities such as hemodiafiltration, home HD, or automated PD were not considered. Using the total number of patients on HD and PD, it was then possible to calculate the prevalence of each modality (using the country’s population) and the percentage of public health expenditure on dialysis provision. From this data, the authors attempted to analyze incentives to increase use of PD by calculating the HD-to-PD reimbursement ratio per country, and then used a rather arbitrary definition of financial incentive to increase use of PD as reimbursement for PD that was >110% of HD reimbursement per patient, per year. Impressively, information was obtained from 90 out of the 94 nephrologists contacted, making this the largest study of its kind aiming to assess factors influencing global dialysis provision, compared with a previous study estimating global HD-to-PD costs that included 46 countries (4) and used cost data extracted from publications dating mostly from 2005 to 2012, although some were even earlier. This study showed that the cost of HD was higher than PD by 1.25–2.35 times in high-income countries, whereas in lower- and middle-income countries, the cost of HD compared with PD was mostly similar. The HD-to-PD ratio was considerably lower in eight low-middle income countries, reflecting the low staff costs in HD and the high cost of importing PD fluid. In the study by van der Tol et al. (8), financial data are current (midpoint of data collection was October 1, 2016). The principal findings from the study are that prevalence of HD and PD correlate with national gross domestic product (GDP) per capita in low- and middle-income, but not high-income countries; the amount of government reimbursement per annum increases with GDP per capita for HD and PD; government reimbursement is sufficient to cover dialysis costs in three quarters of countries overall, but only sufficient in four out of 14 lower-middle and low-income countries; and percentage of health expenditure spent on dialysis is higher in low- or middle-income compared with high-income countries. Finally, eight countries were considered to have a governmental PD-first policy with a considerably higher percentage use of PD. Interestingly, these countries had a similar HD-to-PD reimbursement ratio to countries with no incentive and a much lower PD prevalence. Total dialysis costs were therefore considerably lower in PD first countries. The strengths of this study have already been mentioned: the large number of countries surveyed and the immediacy of the financial data and patient numbers. The problems are recognized by the authors and relate to the data being collected by one person in each country, with no checking by a second person; dialysis costs may not have included all of the same components in the different countries (e.g., nephrologist fees, transport costs, medications etc); and in some countries, reimbursement differs between regions, which may not have been accurately reflected by a single nephrologist. The main conclusion from this study would appear to be that government reimbursement will determine the numbers of patients on dialysis, but it is government policy that directly influences the distribution of dialysis modalities. However, the failure of the Ontario initiative in 2006 to increase PD prevalence shows that government policy on its own will not change dialysis modality distribution (6). The subsequent plan, involving the Ontario Renal Network with more patient-focused targets related to education, dialysis care plans, and choice, has been more successful (6). Similarly, the PD-first policy in Thailand has been achieved in part by the abolition of import duties for PD fluid to decrease costs but, at the same time, focused training of medical personnel, policy and guideline development, enabling insurance cover for PD, and establishing a registry to monitor practice and patient outcomes have all contributed to the huge increase in patients on PD (6). In contrast to the global analysis of the study by van der Tol et al. (8), there is some evidence that financial incentives do play a role in promoting the use of PD both at country and institution levels. Thus in the United States, where dialysis is publicly funded by Medicare but within a private health care system, a change to reimburse HD and PD to the same amount, thereby increasing the profit on PD, has resulted in an increase in patients on PD and PD facilities (9). In the United Kingdom, which is an entirely public health care system, the financial incentive is at institution level where a lump sum is top-sliced from the institution budget and is only given back if an agreed expansion in PD use has occurred; in my own institution, this approach has resulted in an increase in PD numbers from 65 to 138 in the past 4 years, whereas HD numbers have remained static. As shown in Figure 1, the environmental tapestry influencing dialysis modality distribution is much more complex than simply government policy or reimbursement. Even within the United States example of financial incentives increasing PD use, Golper has argued that this increase occurred on the background of rising PD numbers after years of educational efforts aimed at patients and nephrologists (10). His analysis also includes more recent years with a slowing down in the increase of PD numbers after implementation of the financial incentive that was caused by a shortage of PD fluid in 2013, as the manufacturers had not increased supply sufficiently to meet the demand. Curtailing costs of dialysis is essential to enable dialysis provision to grow over coming years. Understanding the factors affecting the use of PD is therefore important, but it is simplistic to think that government policy or reimbursement can result in increased use without changing the culture of local health care provision or finding more economical ways of manufacturing and delivering PD fluid than currently exist. One example of this can be found from recently published experience in Sri Lanka, where a curriculum focusing on PD management was developed and delivered to all health care workers, resulting in a doubling of PD numbers and halving of the peritonitis rate (11). Disclosures None.