International Variation in Time on Peritoneal Dialysis

Globally, a vexing problem in the care of individuals with kidney failure is how best to allocate limited health care resources to provide dialysis treatment to the greatest number of patients while maximizing survival and optimizing other patient-important outcomes, such as health-related quality of life and access to kidney transplantation. One strategy to partially address this problem is to promote public policy initiatives and payment strategies that encourage adoption of KRTs that are more cost effective than “conventional” facility-based hemodialysis. Home-based peritoneal dialysis (PD), which in most cases, is associated with lower annualized per-patient health care costs relative to center-based hemodialysis, is available in approximately 80% of countries globally and is associated with similar or lower rates of adverse clinical outcomes (1–3). Although only 11% of individuals worldwide who receive dialysis are treated with PD, this overall figure masks the dramatic variation in utilization by country and region (1). For example, just under 4% of all patients on dialysis in Japan are currently treated with PD, whereas in Thailand, that figure exceeds 30% (4,5). The most recent data from the United States Renal Data System (USRDS) show that in 2019, just over 11% of patients on prevalent dialysis in the United States were treated with PD (6). Future accelerated expansion in the adoption of PD in the United States is likely, at least in part driven by the 2019 launch of the US Department of Health and Humans Services’ Advancing American Kidney Health Initiative, which set as a broad policy aspiration to have 80% of patients with incident kidney failure treated with home dialysis or kidney transplantation by 2025 (7).

The prevalence and duration of PD treatment may be augmented by either (1) increasing the number and/or proportion of new patients on PD (as either patients on truly incident dialysis or through patients transitioning from hemodialysis) or (2) reducing transitions off PD (via death, transfer to in-center hemodialysis, or kidney transplant). In the United States, data from the USRDS show that all-cause mortality for patients on PD has fallen from 171 to 135 per 1000 patient-years from 2009 to 2019 (6). In contrast, conversion from PD to in-center hemodialysis has shown very little change over the past decade, with a cumulative incidence of approximately 25% at 2 years after dialysis initiation (6). To date, there have been only limited cross-country comparisons assessing these important outcomes for patients on PD worldwide.

In this issue of CJASN, Lambie et al. (8) report the results of a large study evaluating international variation in PD time on therapy by country using data from the Peritoneal Dialysis Outcomes and Practice Patterns Study (PDOPPS). PDOPPS is a 3-year international prospective cohort study that began participant recruitment in 2013 in seven countries around the world, including the United States, Canada, Japan, Australia/New Zealand, the United Kingdom, and Thailand. Countries were selected at least in part to provide diversity in geography, practice patterns, health care infrastructure, and public policy priorities pertaining to kidney failure. Extensive patient-level and center-level baseline information was obtained, and annual assessments were completed by patients and centers to assess patient-reported outcomes as well as center-specific practices and metrics. The primary aims of the study by Lambie et al. (8) were to use PDOPPS data to quantify and compare time on PD therapy by country and to assess the incidence and risk for hemodialysis transfer, death, and kidney transplantation, accounting for potentially confounding variables. Specific patient-level factors assessed included typical demographic characteristics, but also the presence of psychiatric illness, body mass index, and caregiver involvement. Facility-level factors assessed were extensive and included the proportion of patients with weekly Kt/V over 1.7, facility size, and operational features such as patient-nurse ratio.

There were a number of important findings in this study. First, although the median time on PD was 2.3 years overall, there was substantial country to country variation, with the longest time being in Japan (3.2 years) and the shortest time being in the United Kingdom (1.7 years). Second, for patients who transitioned off PD prior to the end of the study period, there were dramatic differences in their relative dispositions. For example, compared with other countries, patients in Japan and Thailand were much less likely to receive a kidney transplant, and this contributed greatly to the longer observed time on PD relative to other countries. In contrast, transplantation rates were very high in the United Kingdom, and this experience is reflected in the short overall time on PD in the United Kingdom of only 1.7 years. Third, despite long-standing recognition that infections are the Achilles’ heel for patients on PD and despite advances in exit site care, antimicrobial regimens, fungal prophylaxis, PD cycler technology, and flush before fill protocols, infection remains the most common reason for transfer to hemodialysis for patients treated with PD, irrespective of country. Finally, despite the granularity of facility-level data available for participating PDOPPS dialysis facilities, there were few significant associations between dialysis facility characteristics and clinical outcomes.

How can these findings help us identify best practices for extending patients’ time on PD in ways that are patient centered and meaningful? One lesson that is obvious yet nonetheless important to emphasize is that not all “time on PD” is alike. A patient’s time on PD may be extended by systems or providers that work to enhance patient self-efficacy and activation through providing critical support to patients and caregivers or, alternatively, by long transplant listing and wait times due to an insufficient supply of available donor kidneys. Given the demonstrated patient-centered benefits of kidney transplantation, efforts to increase transplant evaluation, listing, and living kidney donation should be pursued as a high clinical and policy priority, even though if successful, such efforts will invariably result in shorter time on PD. In a similar vein, a patient’s PD experience may end not with transplant but with either death or transfer to center-based hemodialysis. Although death and transfer to hemodialysis may sometimes be grouped together as a composite outcome in research studies (as was done for a number of the analyses in this study), these individual outcomes are not, in fact, equivalent. For patients, maximizing time on PD is only part of the equation; enhancing access to kidney transplantation, reducing overall mortality, and increasing access to alternative dialysis modalities, such as home hemodialysis and hybrid therapies, are at least as important, if not more so. Additionally, in countries such as Thailand, where “PD-First” policies have been implemented, or in places where there may be limited geographic access to center-based dialysis, providing greater access to in-center hemodialysis for patients struggling with the burden of PD or who have lost a critical caregiver may reduce time on PD but would nonetheless represent a patient-centered policy priority.

Another key lesson from the study by Lambie et al. (8) is that infection continues to represent a vitally important and potentially modifiable risk factor for transfer to hemodialysis in patients treated with PD. The granularity of the data presented shows in stark relief that this is true irrespective of country, region, or specific dialysis provider. In every PDOPPS-participating country, infection represents at least approximately one third of all primary reasons for transfer to hemodialysis for patients treated with PD, and in some countries, such as the United Kingdom and Thailand, this figure exceeds 50%. These data confirm the importance of international efforts to refine our understanding of causes, treatment, and prevention of PD-related infections, as well as the importance of international clinical practice guidelines on the prevention and treatment of PD-related peritonitis, such as those recently released by the International Society of Peritoneal Dialysis (9).

Finally, the study by Lambie et al. (8) serves as an important reminder that traditionally used metrics of dialysis “adequacy” do not stand up as predictors of clinical outcomes in the real world. In particular, the investigators found that the proportion of patients with total weekly Kt/V ≥1.7 at a facility had no association with death or hemodialysis transfer. Despite this, in most countries, between 10% and 20% of patients transitioning to hemodialysis did so due to issues related to solute clearance, and of these patients, nearly 60% transferred primarily due to inadequate Kt/V. There are clear opportunities for international efforts to promote contemporary clinical practice guidelines that de-emphasize quantitative measures of small solute clearance and instead focus holistically on patient-reported outcomes, nutrition, fluid status, and other factors to define high-quality PD (10). A greater embrace of such an approach would minimize potentially unnecessary transfers off PD due to arbitrary small solute clearance targets and thus increase time on PD for patients who wish to do so. A tangible goal for the nephrology community for the next decade should be to connect patients with kidney failure to the KRT that best matches their values, preferences, and individual care needs and then maximize meaningful time on that therapy. For an increasing number of patients around the world, that therapy is PD.

Disclosures

M.B. Rivara reports serving as a member of the North American Advisory Group for PDOPPS, serving as a member of the Medical Executive Committee for Northwest Kidney Centers, serving as a member of the CJASN editorial board, receiving research funding from Satellite Healthcare, and receiving honoraria from Home Dialysis University.

Funding

None.

Author Contributions

M.B. Rivara conceptualized the study, wrote the original draft, and reviewed and edited the manuscript.