Use of peritoneal dialysis (PD) for the treatment of end-stage renal disease (ESRD) has decreased sharply in many developed countries during the past 15 years.1, 2 This decrease has been particularly marked in the United States, where the proportion of dialysis patients on PD therapy has decreased from 14% to 8%.2
In this issue of the American Journal of Kidney Diseases, Mehrotra et al3 use the US Renal Data System database to look at the interaction between PD use and ownership pattern of dialysis units during 1996 to 2004. They show a disproportionate increase in the percentage of patients treated by large dialysis organizations (LDOs). They report that LDO units generally use PD less than non-LDO units, and also that some LDO units have consistently lower PD use than others. They find that the LDOs with low PD use had greater mortality and technique failure rates with PD.
It is well recognized that modality distribution varies greatly among countries and is influenced highly by the economic structure of health care delivery in the various jurisdictions.4 In particular, in such countries as the United Kingdom and Canada, where public providers dominate dialysis delivery, use of home dialysis modalities is relatively high. In contrast, in countries where private providers dominate, such as the United States, Germany, and Japan, hemodialysis (HD) is overwhelmingly used and home dialysis use typically is less than 10%. Notwithstanding, the recent decrease in PD use is being seen in both public and private provider countries. For example, PD use in Canada has decreased from 32% in 1996 to 19% in 2005, and in the United Kingdom, it has decreased from 34% in 1997 to 23% in 2005.5, 6
Attempts to identify the factors underlying the decrease in PD use therefore need to look at trends across these different countries. However, we should start by considering why private providers make so little use of PD. This often is presented as a paradox because costing studies from developed countries consistently show that PD costs less than HD.2, 7, 8 The extra costs of providing sterile PD solutions are more than offset by the greater staffing costs involved in delivering HD. However, the reimbursement side of the balance sheet also needs to be considered. In many countries, there is notably greater provider and physician reimbursement for HD and thus it should not be too surprising that the modality is favored.7 However, in the United States, this explanation is not applicable. Provider and physician reimbursement, despite recent modifications, are largely modality independent.9 With lower costs and equivalent reimbursement, why do US providers and physicians continue to use so little PD? The question often is posed as if this provider behavior is economically illogical or even foolish. The answer is of course more complicated. Comparative costing studies do not capture all the economic realities of US practice.
First, in the United States and other private provider countries, in-center HD compared with PD allows greater billing opportunities for investigations and for administration of such parenteral medications as erythropoietin, iron, and vitamin D analogues. Anecdotally, it often is stated that the profitability of many US HD units depends on fees related to erythropoietin administration. This situation may change somewhat with possible moves toward a more capitated or bundled form of remuneration.
Second, absolute costs do not reflect marginal costs. In a situation in which there is unused capacity in HD units, the cost of treating additional patients is relatively less. Fixed costs are incurred whether an HD unit is half or completely full. This is a much greater issue with the unprecedented growth in the number of for-profit units across the United States that has occurred during the past decade. The incidence rate of ESRD is no longer increasing, but the number of dialysis units is, and the consequence is greater HD overcapacity.2
There is a third less tangible factor that needs to be considered: the workload required to manage HD versus PD patients in the US setting. This area is not well studied, and impressions come from anecdotal opinions. However, there appears to be a perception among many US nephrologists that it is easier to initiate patients on HD therapy, it requires less thought and effort to manage them, the degree of control over the patient is greater, and there is the added comfort of having the patient seen 3 times weekly by facility staff. All these factors are magnified in a setting in which the number of HD units is increasing and PD use is decreasing. In centers with small numbers of PD patients, experience of physicians and staff with the modality obviously will be less and discomfort will be greater. There is a growing literature indicating better outcomes on PD therapy with greater center experience.10, 11
Paradoxically, in large centers with high PD use, there may be the contrasting perception that managing extra numbers of PD patients is less work for the unit and the physician. In essence, it is a question of what infrastructure is available to support physicians in caring for the patients. In large PD programs, there are full-time PD nurses, often providing extensive on-call service, and cross-covering physicians are experienced with the modality. In small PD programs, there often is care from nurses who do PD only part time and who may not be regularly on call, and physician colleagues may be poorly trained in the modality. In other words, a real-life economic analysis needs to take into account not only the reimbursement received and costs incurred by providers and physicians, but also the perceived effort that will be expended to earn that reimbursement.
This analysis also can be extrapolated to public provider systems. A common feature in all developed countries in the past decade is the proliferation of HD units. Public provider countries, such as Canada, Australia, and the United Kingdom, previously had highly centralized dialysis delivery systems with large units typically located in academic centers. Now there are units in towns and smaller cities and the suburbs of large cities.12, 13 Many are in nonacademic centers in which infrastructural support for PD will be less. Some do not even provide PD.
The hypothesis then is that a common factor that has led to decreasing PD use in both private and public provider countries alike is the proliferation of relatively small dialysis units during the past 15 years. Some of these do not offer PD, and few have a sufficiently large PD-related infrastructure to maximize the efficiency and profitability of this modality. It may be that PD use is influenced more by the size of dialysis units and the presence of HD overcapacity than by whether the units are publicly or privately owned.
Of course, there are other factors that may have decreased PD use during the past decade and a half. The increasing age and comorbidity burden of ESRD populations have made home dialysis more challenging; however, Mehrotra14 has presented data to suggest that this is not a sufficient explanation. The negative influence of some US comparative outcome analyses or of studies questioning the adequacy of PD clearances also may have contributed, but the same decreases in PD use are seen in countries in which outcome studies were more favorable to PD.15, 16 Other possible contributors include increased use of living donor transplantation and shortages of nurses with skills in PD. Failure to even offer PD to patients and deficiencies in the education of nephrologists have also been invoked, but it is likely that these are secondary effects that compound the problem, and are not primary causes of it.17
Findings in the study by Mehrotra et al3 confirm that private for-profit units use less PD than not-for-profit units. The differences in PD use between LDOs are statistically significant, but not huge, with a range of 6% to 10% in 2004. It would be interesting to know whether these differences reflect smaller units or more HD overcapacity; however, this information is not provided. Outcome data showing worse outcomes on PD therapy in LDOs with lower PD use complements a growing body of literature indicating that greater experience with PD leads to better outcomes.10, 11
Can this decrease in PD use be reversed? There is no turning back the proliferation of dialysis units that has already occurred, although the recent stabilization of incidence rates for ESRD and the present adverse economic climate may slow this trend.2 Consolidation of PD provision into a small number of specialist units in a given region has been proposed and attempted by 1 US dialysis chain, but the feasibility of this approach on a larger scale is unproved.18 Educational initiatives for nephrologists and nephrology trainees in PD also are important, as are approaches to reimbursement that lessen the present advantage for HD.
PD is not a panacea for every patient; however, outcome studies from a variety of countries suggest that it is an effective modality for many. Given a free choice, as many as 50% of patients choose the modality.19 Furthermore, PD is cost-effective and consistent with modern paradigms of patient empowerment and autonomy. There is a serious risk that its use will fade away to insignificance in the United States if present trends continue and more innovative approaches are not made. LDOs in particular need to take on board the findings of the study by Mehrotra et al.3
FULL TEXT
