In January 2005, there were 65,121 patients with end-stage renal disease (ESRD) on dialysis treatment in Brazil (1). The high cost of renal replacement therapy (RRT) is well recognized in most countries; however, due to its importance in saving lives, it has been funded by the federal government of Brazil since 1975. In an environment of limited economic resources, the increasing costs to provide this treatment is a cause of concern, as the number of ESRD patients has increased by 10.1 percent each year in Brazil (1). The costs of materials used have also increased, along with a permanent sophistication of the treatment. In 2004 alone, it was estimated that the Federal government spent more than US$ 500 million in the provision of RRT in Brazil.
In 1990 we reported that, in Brazil, renal transplantation was more cost-effective than chronic dialysis treatment (11). Moreover, we also showed that late diagnosis of advanced renal failure incurs greater costs, mainly during the first 6 months of chronic dialysis treatment (12). Recently, economic studies have been published comparing ESRD treatment modalities (3;6;17), type of initial dialysis (4;13), the impact of time of referral (10), and the frequency of dialysis regimens (5;7).
Despite the importance of this issue and the increasing number of patients and costs related to the treatment, there is a very limited number of economic analyses in the literature regarding RRT, particularly in developing countries (11). In this study, our objective was to evaluate the resources and costs incurred by patients on chronic hemodialysis treatment, including an assessment of the costs of all-cause hospital admissions and those related to cardiovascular events.
SUBJECTS AND METHODS
We retrospectively evaluated data on 200 patients with ESRD on chronic hemodialysis treatment at a satellite unit of the Escola Paulista de Medicina, Federal University of São Paulo, from January 2001 to December 2004. Inclusion criteria were patients at least 18 years of age, on chronic dialysis for more than 6 months, admitted to the institution in the mentioned period, and undergoing dialysis sessions for 4 hours, three times per week. Patients were followed from the date of study inclusion until the final date of follow-up (December 2004) or the date of death, transfer to another unit, or the date of transplant, whichever came first.
Patient medical charts from the dialysis unit and from hospital admissions were reviewed. Whenever necessary, patients were personally interviewed to confirm data. The following data were collected: demographics, laboratory and clinical parameters at the time of study entry, ambulatory procedures (dialysis sessions, surgical procedures, medications, medical visits, complications of venous dialysis access), and hospitalizations (length of stay, cardiovascular cause of admission, length of stay in the intensive care unit, surgical interventions, diagnostic tests, medications). The main end points evaluated during the follow-up were all-cause mortality, cause of death, and cardiovascular events. Cardiovascular events were classified based on the International Code of Diseases, Version 10 (ICD-10). Two independent reviewers validated the diagnoses of cardiovascular events and deaths, by blind review of medical charts and interview with patients or relatives (in the case of death).
The economic analysis was conducted on the perspective of both payers in Brazil, which include the Ministry of Health (MoH) or private health insurance (PHI). All costs are expressed in US$, using an exchange rate of 1 US$ = R$ 2.5 (Brazilian reais), which was the average rate from January to September 2005. We computed the direct costs related to ambulatory treatment and hospital admissions. Costs of medications were obtained from market prices (9). The costs of hospitalization from the perspective of MoH, including laboratory tests, medical and surgical procedures, taxes, services, hospital materials, and medical fees were obtained from reimbursement tariffs in September 2005 (15). The costs of hospitalization from the perspective of PHI were obtained through tables of the Brazilian Medical Association and from an average of values used by the three insurance companies with greatest volume of services in São Paulo during the same period.
The reimbursement fees for ambulatory dialysis are based on a value per session, which includes all costs for providing the therapy. In September 2005, the value per session reimbursed by the MoH was US$ 45.5, and on average, US$ 78 by PHI.
RESULTS
Two hundred patients with a mean ± SD age of 59 ± 23 years were included in the study. The majority was male (59 percent) and of white race (60 percent). Twenty-four percent were diabetics, and 16 percent had a previous history of cardiovascular disease. Median time on dialysis was 9.2 months (range, 6 to 231 months). The duration of follow-up in the study was 6,517 patient-months (543.1 patient-years).
One hundred twenty-three patients were hospitalized during the study period, and there was a total of 287 hospitalizations. Mean number of hospitalizations was 0.53 per patient-year. Mean number of days of hospitalization was 12 per patient-year (Table 1).
During the follow-up time, 105 cardiovascular events were detected; the most frequent causes were coronary disease (n = 59, 56 percent), cardiac insufficiency (n = 26, 25 percent) and cerebrovascular disease (n = 12, 12 percent). Mean number of cardiovascular events was 0.51 per patient, and the rate of cardiovascular events was 193 per 1,000 patient-years of follow-up. There were 43 deaths (21.5 percent) during the follow-up, and their main causes were cardiovascular (n = 15, 35 percent), infectious (n = 12, 28 percent), other (n = 4, 9 percent), and unknown (n = 12, 28 percent). The death rate was 79 per 1,000 patient-years of follow-up.
The costs per episode of hospitalization are shown in Table 2. Median cost per admission (all causes) was US$ 675 and US$ 932, from the perspective of the MoH and PHI, respectively. For admissions of cardiovascular cause, the corresponding median costs were US$ 1,639 and US$ 4,499, respectively (Table 2). Medication use was evaluated among patients during the follow-up in the ambulatory setting. Forty-three percent of the patients were treated with vitamin D (mean daily dose, 0.4 ± 0.25 μg), 31 percent with sevelamer (mean daily dose, 4200 ± 400 mg), 72 percent with calcium carbonate (mean daily dose, 2.4 ± 2.2 g), 13 percent with statins (mean daily dose, 12 ± 4 mg), and 95 percent with erythropoietin (mean weekly dose, 6,000 ± 3,000 U).
Overall costs and their main components are shown in Table 3 for both types of settings (ambulatory or hospital) and from both payers' perspectives. Costs were classified as medications, diagnostic laboratory/radiology tests and professional fees, dialysis sessions, and hospital services/taxes. The largest cost component was dialysis therapy, followed by hospitalization and medications. The mean global cost was US$ 7,980 per patient-year and US$ 13,428 per patient-year, from the perspective of the MoH and PHI, respectively.
DISCUSSION
In this study, we estimated that the mean overall annual cost, including ambulatory treatment and hospital admissions, to maintain a patient with ESRD in chronic hemodialysis was US$ 7,980 and US$ 13,428 from the perspective of the government or private insurance, respectively. Data from the Brazilian Society of Nephrology census demonstrated that the costs of 90 percent of the patients are supported by the government and only 10 percent by private health insurance companies (1). Furthermore, 89 percent of the ESRD patients undergo hemodialysis and 11 percent peritoneal dialysis (1).
In this study, the relative contribution of hospitalizations to the overall costs may have been underestimated due to the retrospective fashion of the data collection and the inherent difficulties in obtaining these costs. Furthermore, it is recognized that the costs reimbursed by the MoH for procedures, hospitalizations, and professional fees are relatively low. This finding reinforces the importance of our analysis of costs from the perspective of private insurance, which reflects values that are closer to the actual costs of care. However, the latter also confirms how low hospitalization costs are in relation to dialysis costs in Brazil.
By contrast, in 2004 it was reported that the annual cost to maintain a patient on chronic hemodialysis in the United States was around US$ 70,000 per year (8;13) and that 40–50 percent of this cost was due to dialysis therapy and the remaining 50–60 percent due to hospitalization. Moreover, the value of professional honorary was much higher in the United States than in our setting. St. Peter et al. (14) showed that, in the United States, hospital costs were an important component of overall costs, especially during the 6 months before dialysis commencement and during the 6 first months after commencing therapy. This finding could not be verified in our study because the selected patients had more than 6 months in dialysis treatment. A systematic review of 13 cost-effectiveness studies of RRT in the past three decades showed that the cost-effectiveness ratio of dialysis was US$ 55,000–80,000 per year of life gained, and US$ 10,000 per life year gained for renal transplant (6).
In a previous study, we observed that the mean monthly cost during the first 6 months of dialysis treatment in Brazil was significantly greater for patients with late diagnosis of chronic renal failure compared with those with an earlier diagnosis and referral (US$ 3,360 vs. US$ 2,080 per month, respectively) (12). This finding probably reflects the fact that patients who are referred later experience a greater number of hospitalizations, clinical complications, and more frequent use of temporary dialysis access (6;10).
In a cost-effectiveness analysis in 1990, we estimated that the cost per life year gained on continuous ambulatory peritoneal dialysis and hemodialysis in Brazil was US$ 12,134 and US$ 10,065, respectively. The costs per life year gained for living and cadaveric donor transplantation were lower than for dialysis, US$ 3,022 and US$ 6,978, respectively (11).
Shih et al. (13) reported that the mean annual cost of starting treatment with peritoneal dialysis was US$ 53,277 and with hemodialysis was US$ 72,189 in the United States, and this difference has been confirmed by a systematic literature review (4). As only 11 percent of chronic dialysis patients in Brazil undergo peritoneal dialysis, it seems to be timely and reasonable to consider the expansion of the use of this dialysis modality as the initial one of choice. Other economic studies verifying the comparison of these modalities in our setting would be useful before recommending this shift, however.
In the present study, we evaluated patients on “conventional” hemodialysis in a satellite center, undergoing three, 4-hour sessions per week, because this is the modality performed by the majority of patients on hemodialysis in Brazil. In Canada, it has been suggested that daily hemodialysis incurs lower annual costs than nocturnal or “conventional” hemodialysis (Can$ 67,330, Can$ 74,400, and Can$ 72,700, respectively) (5). Other U.S. studies have also suggested this tendency (7).
Finally, we highlight the relative economic significance of cardiovascular events in these patients (193 per 1,000 patient-years), as it has been demonstrated that these events are more prevalent in patients with chronic renal failure than in the general population (2). Cardiovascular disease accounts for at least 50 percent of the mortality in this population (16). Moreover, we found that cardiovascular disease made up a considerable proportion of hospitalization costs. The opportunity afforded by therapies that can reduce this burden is great indeed. Although our study has a limited generalizability to other countries in terms of costs, taking into account the lack of economic analyses on the issue in most countries, it adds important information about dialysis care and the resources consumed to provide this high cost technology in a developing country. Moreover, the methodology used could be transferable and applied in further studies.
In conclusion, this study demonstrates the significant economic impact of chronic hemodialysis treatment in Brazil, including the economic burden of cardiovascular hospitalization. The rate of cardiovascular events accounted for approximately 40 percent of all hospitalizations, which represents a significantly higher proportion than in the general population. Given the significant resources provided for ESRD patients as we have shown, this study demonstrates that interventions that may alleviate cardiovascular morbidity will help to ensure that ESRD treatment programs are sustainable.
POLICY IMPLICATIONS
Chronic in-center hemodialysis is a high cost treatment modality for ESRD. New policies to improve cardiovascular disease management through innovative interventions are needed, particular those that promote the prevention, early diagnosis, and treatment of cardiovascular diseases in patients at initial stages of chronic renal failure. Alternative strategies of RRT, such as an increasing renal transplant activity and the greater use of peritoneal dialysis, if they confirm to be more cost-effective than “conventional” hemodialysis, should be explored in developing countries.
CONTACT INFORMATION
Ricardo Sesso, MD (rsesso@nefro.epm.br), Associate Professor, Chief, Camilla Barbosa da Silva, RN (dialisefor@uol.com.br), Postgraduate student, Nurse, Department of Medicine, Division of Nephrology, Federal University of São Paulo, Escola Paulista de Medicina, Rua Botucatu 740, São Paulo, SP, Brazil, 04023-900
Sérgio C. Kowalski, MD, PhD (sergio.kowalski@hcnet.usp.br), Assistant Physician, Department of Internal Medicine, Hospital das Clínicas, Faculty of Medicine the University of São Paulo, 455 Dr. Arnaldo 3rd floor, room 3133, São Paulo, SP, Brazil, 01246-903
Silvia R. Manfredi, RN (dialisefor@uol.com.br), Postgraduate student, Nurse, Maria E. Canziani, MD (dialisefor@uol.com.br), Assistant Professor, Assistant Physician, Sergio A. Draibe, MD (dialisefor@uol.com.br), Full Professor, Physician, Department of Medicine, Division of Nephrology, Federal University of São Paulo, Escola Paulista de Medicina, Rua Botucatu 740, São Paulo, SP, Brazil, 04023-900
Heba A. Elgazzar, MSc (heba.elgazzar@genzyme.com), Associate Director, Global Health Outcomes and Strategic Pricing, Genzyme Corporation, Oxford Business Park South, Oxford OX4 2SU, UK
Marcos B. Ferraz, MD, PhD (marcos.ferraz@cpes.org.br), Associate Professor, Department of Medicine, Federal University of São Paulo, Rua Botucatu 740; Director, São Paulo Center for Health Economics, Federal University of São Paulo, Rua Botucatu 685, São Paulo, São Paulo, 04023-062
This work was conducted with support from the Genzyme Corporation. Dr. Ricardo Sesso, Dr. Sergio Draibe, and Dr. Marcos B. Ferraz received a research grant from the Brazilian Research Council (CNPq).
