Diane L. Frankenfield

National Profile of Practice Patterns for Hemodialysis Vascular Access in the United States

The Centers for Medicare & Medicaid Services (CMS), national End-Stage Renal Disease (ESRD) Clinical Performance Measures (CPM) Project is a data collection initiative to identify opportunities for improvement of care to adult, Medicare maintenance dialysis beneficiaries. This analysis of 1999 CPM data characterizes the profile of hemodialysis vascular access in the United States and identifies determinants of vascular access type 2 yr after the translation of vascular access clinical practice guideline statements into national CPMs. CPM data were collected during October to December 1999 and stratified by the 18 regional ESRD networks. Univariate and multivariable analyses were conducted to examine associations of access type with demographic, laboratory, and geographic variables. Multivariable logistic regression analyses were performed to identify independent variables associated with access type. A total of 8154 hemodialysis patients were sampled; 17% (n = 1399) were incident. Twenty-eight percent were dialyzed through an autologous arteriovenous fistula (AVF), 49% through a prosthetic graft (AVG), and 23% through a percutaneous catheter. Independent predictors of having a catheter for hemodialysis were female gender, white race, incident to hemodialysis status, and lower hemoglobin and serum albumin. For patients with a fistula or AVG, female gender (odds ration [OR], 2.46 [2.18 to 2.78]) and black race (OR, 1.70 [1.50 to 1.93]) were the strongest predictors of dialysis through an AVG. Other predictors of dialysis through an AVG were older age, increased body mass index (BMI), diabetes mellitus as the cause of ESRD, and lower serum albumin. Even in adjusted analyses, there was significant geographic variability with respect to hemodialysis access type. Despite translation of practice guidelines for hemodialysis vascular access into national CPMs, there is substantial geographic variability and gender and racial disparity in angioaccess allocation in the United States. Quality improvement strategies to improve the prevalence of fistulae should focus on selected regions and include physician education about their practice patterns and potential biases.

Relationship between Clinical Performance Measures and Outcomes among Patients Receiving Long-Term Hemodialysis

Context Several intermediate outcomes are used to evaluate care in hemodialysis programs, but the relationship between these measures and patient outcomes is uncertain. Contribution The authors studied 4 intermediate outcome measures (anemia, serum albumin level, functioning vascular access, and dialysis adequacy) in a 5% random sample of all U.S. patients receiving long-term hemodialysis. Annual mortality rates in patients who met 0, 1, 2, 3, or 4 quality measures were 29%, 25%, 21%, 14%, and 7%, respectively. Cautions This study was unable to determine whether failure to meet an outcome measure was due to patient factors, such as severity of illness, or because of poor-quality care. Implications Failure to meet hemodialysis quality measures is associated with increased mortality and hospitalization rates. The Editors Mortality rates among patients with end-stage renal disease (ESRD) in the United States have declined during the past 2 decades despite increases in the age and prevalence of diabetes in this population (1). The improvements in survival are not uniformly distributed across patients with differing durations of long-term dialysis therapy. Mortality rates have declined by 22% during the first 2 years after renal replacement therapy and by 15% during the next 3 years and have actually increased for patients surviving longer than 5 years (1). Despite these improvements in early survival, mortality rates remain quite high among patients with ESRD who are receiving hemodialysis, and 212 of every 1000 patients die annually. It is important to identify potentially modifiable risk factors for the persistently high mortality rates among U.S. patients with ESRD. The quality of care provided after the onset of renal replacement therapy provides one example of a set of modifiable risk factors. In 1997, the National Kidney Foundation developed quality-of-care indicators as part of the Dialysis Outcomes Quality Initiative (DOQI). These indicators were modified by the Centers for Medicare & Medicaid Services (CMS) ESRD Clinical Performance Measures Project (formerly known as the ESRD Core Indicators Project). Increased adequacy of hemodialysis and control of anemia have been shown to be associated with improved survival and quality of life among patients with ESRD (24). During the past decade, there has been a statistically significant improvement in the adequacy of dialysis as measured by Kt/V urea and in anemia management, which was fostered and documented by the CMS Clinical Performance Measures Project (5). Most studies that have examined the association between the quality of care of patients with ESRD and subsequent risk for death have used single measures of the quality of care provided to patients receiving dialysis (613). This approach fails to account for the possibility that achieving certain thresholds for multiple care processes is associated with less hospitalization and improvements in patient survival. We have tested this hypothesis using data from the ESRD Clinical Performance Measures Project to determine the association between hospitalization and mortality rates and meeting multiple guidelines for care delivered to patients receiving chronic hemodialysis. Methods Study Design and Sample Selection Detailed information about CMSs ESRD Clinical Performance Measures Project has previously been published (5). In brief, a random sample of Medicare-eligible adults receiving in-center hemodialysis stratified by the 18 ESRD Networks is selected each year from a census of patients with ESRD. The ESRD Networks are regional organizations contracted by CMS to perform quality oversight activities to ensure the appropriateness of services and quality of services for patients receiving dialysis. Patients were eligible for inclusion in the sample for each year of the study if they were at least 18 years of age as of 1 October of the previous year (for example, 1 October 1998 for study year 1999), were receiving in-center hemodialysis, and were alive on 31 December of the previous year. Clinical information was obtained on the selected patients for October to December of the year before the patient was chosen for the study. If a patient was randomly selected for more than 1 study year, then only the earlier year in which the patient was included in these samples was used for analytic purposes. For this analysis, data were analyzed for study years 1999 and 2000. Data Collection In May of each study year, the ESRD Networks sent a 3-page data collection form to approximately 2700 facilities. These facilities selected 1 or more patients for the sample. The median number of patients sampled per center was 2, and the number of patients sampled at an individual dialysis center ranged from 1 to 27; the 25th and 75th percentiles were 1 and 4 patients, respectively. Available patient information included sex, age, race (white, black, Asian or Pacific Islander, Native American or Alaska native, and other or unknown), Hispanic ethnicity, the primary cause of ESRD, and number of years that the patient received dialysis. In June of each study year, dialysis facility staff abstracted patient data from medical records for October to December of the previous year, including the patients height, first documented blood urea nitrogen concentration before and after dialysis, weight before and after dialysis, dialysis session length, type of vascular access, hemoglobin value, and serum albumin value (bromcresol green or bromcresol purple laboratory methods). Completed forms were returned to the appropriate ESRD Network office where data were reviewed and entered into a computerized database. The Networks forwarded data to CMS for aggregation and analysis. Data regarding comorbid conditions were obtained by merging the ESRD Clinical Performance Measures database with data from CMS Form 2728 (Medical Evidence Form). This form provides additional clinical information on each patient at the initiation of dialysis. Specifically, the presence or absence of each of the following conditions was ascertained at the initiation of dialysis: congestive heart failure; ischemic heart disease; coronary artery disease; myocardial infarction; cardiac arrest; cardiac dysrhythmia; pericarditis; cerebrovascular disease, cerebrovascular accident, or transient ischemic attack; peripheral vascular disease; history of hypertension; diabetes mellitus (primary or contributing); diabetes mellitus and current use of insulin; chronic obstructive pulmonary disease; tobacco use (current smoker); malignant neoplasm or cancer; alcohol dependence; drug dependence; HIV-positive status; AIDS; inability to ambulate; and inability to transfer. Hospitalization information (all-cause hospitalization and 1 hospital admission during the follow-up period) was obtained from the U.S. Renal Data Systems (USRDS) Hospitalization Standard Analytic File; dates of death, censoring information, and withdrawal from dialysis were obtained from the USRDSs Patients Standard Analytic File. Hospitalizations and dates of death were subsequently linked to the clinical information in the ESRD Clinical Performance Measures data file. Statistical Analysis The following intermediate outcomes or clinical measure targets were examined: mean Kt/V urea value of 1.2 or greater, mean hemoglobin value of 110 g/L or greater, use of an arteriovenous fistula for vascular access, and mean serum albumin value of 40 g/L or greater or 37 g/L or greater by the bromcresol green or bromcresol purple laboratory methods, respectively. In calculating mean values for Kt/V urea, hemoglobin, and serum albumin, all available reported data were used. If a value was missing for 1 of the study months, the mean value was calculated by averaging the 2 values over the 3-month study period. If values were missing for 2 of the 3 months, the available value became the mean value for that patient. For calculating rates of arteriovenous fistula use, patients with missing data were excluded from the denominator. Associations of group classification with the clinical measure targets were tested by the chi-square test, the 2-tailed Student t-test, and hierarchical analysis of variance. A 2-tailed P value less than 0.01 was considered statistically significant. Racial categorization was restricted to black and white only because of the small numbers of patients in other racial groups. Comorbid conditions were categorized as cardiovascular or noncardiovascular for subsequent analyses. Cardiovascular comorbid conditions included cardiac arrest, congestive heart failure, cerebrovascular disease, cardiac dysrhythmia, pericarditis, peripheral vascular disease, ischemic heart disease, and myocardial infarction. Noncardiovascular comorbid conditions included alcohol dependence, cancer, drug dependence, HIV infection, AIDS, inability to ambulate, inability to transfer, chronic obstructive pulmonary disease, and tobacco use. If a comorbid condition was not indicated on the Medical Evidence Form, then we assumed it was not present for the purposes of this analysis. To determine whether meeting 1 or more clinical measure targets was associated with hospitalization or death during the 12 months after the end of the study period, separate multivariable Cox proportional hazards modeling was performed. In addition to the number of clinical measure targets that were met, factors that were entered into the models included sex, race (black and white only), Hispanic ethnicity, age, years of dialysis, diabetes mellitus as the cause of ESRD compared with other causes combined, mean body mass index (weight in kilograms divided by the square of the height in meters) after dialysis, and preexisting cardiovascular and noncardiovascular comorbid conditions. Patients were censored at the time of transplantation, when they were switched to peritoneal dialysis, or when they were lost to follow-up for the Cox proportional hazards modeling analyses. All facto

Racial/ethnic analysis of selected intermediate outcomes for hemodialysis patients: results from the 1997 ESRD Core Indicators Project.

Principal goals of the End-Stage Renal Disease (ESRD) Core Indicators Project are to improve the care provided to ESRD patients and to identify categorical variability in intermediate outcomes of dialysis care. The purpose of the current analysis is to extend our observations about the variability of intermediate outcomes of ESRD care among different racial and gender groups to a previously unreported group, Hispanic Americans. This group is a significant and growing minority segment of the ESRD population. A random sample of Medicare-eligible adult, in-center, hemodialysis patients was selected and stratified from an end-of-year ESRD patient census for 1996. Of the 6,858 patients in the final sample, 45% were non-Hispanic whites, 36% were non-Hispanic blacks, and 11% were Hispanic. Whites were older than blacks or Hispanics (P < 0.001). Hispanics were more likely to have diabetes mellitus as a primary diagnosis than either blacks or whites (P < 0.001). Even though they received longer hemodialysis times and were treated with high-flux hemodialyzers, blacks had significantly lower hemodialysis doses than white or Hispanic patients (P < 0.001). The intradialytic weight losses were greater for blacks (P < 0.05). The delivered hemodialysis dose was lower for blacks than for whites or Hispanics whether measured as a urea reduction ratio (URR) or as the Kt/V calculated by the second generation formula of Daugirdas (median 1. 32, 1.36, and 1.37, respectively, P < 0.001). Hispanics and whites had modestly higher hematocrits than blacks (33.2, 33.2, and 33.0%, respectively, P < 0.01). There was no significant difference among groups in the weekly prescribed epoetin alfa dose ( approximately 172 units/kg/week). A significantly greater proportion of Hispanic patients had transferrin saturations >/=20% compared with the other two groups (P < 0.001). Logistic regression modeling revealed that whites were significantly more likely to have serum albumin <3. 5(BCG)/3.2(BCP) gm/dL (OR 1.4, p < 0.01); blacks were significantly more likely to have a delivered Kt/V < 1.2 (OR 1.4, P < 0.001) and hematocrit <30%, (OR 1.2; P < 0.05) and both blacks and Hispanics were significantly more likely to have a delivered URR < 65% (OR 1.5, P < 0.001 and 1.2, P < 0.05, respectively).

Trends in Anemia Management among US Hemodialysis Patients

This study was undertaken to describe the relationship between hematocrit (Hct) and changes in the prescribed dose of erythropoietin (EPO) as well as selected patient and process care measures across annual national samples of hemodialysis patients from 1994 to 1998. This study uses the cohorts identified in the ESRD Core Indicators Project, random samples of 6181, 6241, 6364, 6634, and 7660 patients, stratified by ESRD Networks drawn for each year from 1994 to 1998. Patient demographic and clinical information was collected from October to December for each year. Surrogates of iron stores and patterns of iron and EPO administration were profiled from 1996 to 1998. Multivariable stepwise linear regression analyses were performed to adjust for potential confounding variables and to identify independent variables associated with Hct and EPO dose. Mean Hct and EPO dose increased each year from 31.1 +/- 5.2% to 34.1 +/- 3.7% and from 58.2 +/- 41.8 U/kg to 68.2 +/- 55.0 U/kg, respectively (P = 0.0001). Increasing Hct was positively associated with male gender, more years on dialysis, older age, higher urea reduction ratio and transferrin saturation, prescription of intravenous iron, and lower ferritin and EPO dose in multivariable models (all P = 0.0001). Male gender, older age, diabetes, higher Hct, and increasing weight, urea reduction ration, and transferrin saturation were associated with lower EPO doses (all P < 0.01). Conversely, intravenous EPO and iron were associated with higher prescribed EPO doses (all P = 0.0001). Although increasing Hct is associated with decreasing EPO dose at the patient level, the increase in Hct seen across years among the cohorts of hemodialysis patients in the United States has been associated with increasing doses of EPO at the population level.

Hypertension in Pediatric Long-term Hemodialysis Patients in the United States

BACKGROUND AND OBJECTIVES Data are limited regarding BP distribution and the prevalence of hypertension in pediatric long-term dialysis patients. This study aimed to examine BP distribution in U.S. pediatric long-term hemodialysis patients. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS This cross-sectional study of all U.S. pediatric (aged 0-< 18 yr, n = 624) long-term hemodialysis patients was performed as part of the Centers for Medicare & Medicaid Services End-Stage Renal Disease (ESRD) Clinical Performance Measures Project. BP and clinical information were collected monthly in October, November, and December 2001. Hypertension was defined as the mean of pre- and postdialysis systolic or diastolic BP above the 95th percentile for age, height, and sex, or antihypertensive medication use. Results were calculated by age, sex, race, ethnicity, ESRD duration, body mass index percentile, primary cause of ESRD, and laboratory data. RESULTS Hypertension was present in 79% of patients; 62% used antihypertensive medication. Five percent of patients were prehypertensive (mean BP at 90th to 95th percentile). Hypertension was uncontrolled in 74% of treated patients. Characteristics associated with hypertension included acquired kidney disease, shorter duration of ESRD, and lower mean hemoglobin and calcium values. Characteristics associated with uncontrolled hypertension were younger age and shorter duration of ESRD. CONCLUSIONS Hypertension is common in U.S. pediatric long-term hemodialysis patients, uncontrolled in 74% of treated patients, and untreated in 21% of hypertensive patients. It is concluded that a more aggressive approach to treatment of hypertension is warranted in pediatric long-term hemodialysis patients.

Relationship between urea reduction ratio, demographic characteristics, and body weight for patients in the 1996 National ESRD Core Indicators Project

The 1996 Health Care Financing Administrations (HCFA) Core Indicators Project for in-center, hemodialysis patients collects information on the quality of care delivered in four clinical areas that were anticipated to predict patient outcomes. Included among these clinical performance measurements is the delivered dose of hemodialysis, measured by the fractional reduction of urea achieved during a single hemodialysis session (urea reduction ratio [URR]). A random sample (N = 7,310) of adult (aged > or =18 years), in-center hemodialysis patients was selected, and a one-page data collection form for each patient was sent to the dialysis facility in which care was provided during the last quarter of 1995. The dialysis facilities provided information for 6,861 (94%) patients, and at least one paired predialysis and postdialysis blood urea nitrogen (BUN) concentration was reported for 6,655 (97%) of these patients. The URR of this cohort was 65.5% +/- 8.0% (mean +/- SD), and 41% of patients had a URR less than 65%. The mean dialysis session length was 203 minutes, and more than half of the patients received dialysis with a dialyzer membrane with a KUf less than 10 mL/mm Hg/h. The patients with a URR less than 65% had a mean body weight approximately 10 kg greater than patients with a URR of 65% or greater. This relationship was present for all demographic characteristics studied, including age, gender, race, and primary cause of end-stage renal disease (ESRD). Patients receiving dialysis for less than 6 months were more likely to have a URR less than 65% than patients on dialysis for longer periods. By multivariate analysis, variables significantly associated with a delivered URR less than 65% were body weight in the heaviest quartile (odds ratio [OR] = 6.1), male gender (OR = 2.6), on dialysis therapy less than 6 months (OR = 2.5), youngest quartile of age (<49 years) (OR = 2.0), lowest quartile of serum albumin values less than 3.6 g/dL (bromcresol green method) or less than 3.3 g/dL (bromcresol purple method) (OR = 1.6), black (OR = 1.5), dialyzed with a dialyzer KUf less than 20 mL/mm Hg/h (OR = 1.8), lowest quartile hematocrit (<29.7%) (OR = 1.2), and shorter dialysis session length (OR = 1.02/min). In conclusion, both patient-specific demographic variables and treatment-specific parameters are significantly associated with ESRD patients receiving a URR less than 65%. Furthermore, these data suggest statistically significant linkages between the delivered dose of hemodialysis and other independent outcome predictors such as serum albumin concentration. Prospective study is required to determine whether intervention strategies to improve the delivered dose of hemodialysis will affect this outcome predictor or whether serum albumin and dialysis dose share a common cause not amenable to increasing the URR.

Developing clinical performance measures based on the Dialysis Outcomes Quality Initiative Clinical Practice Guidelines: process, outcomes, and implications.

BACKGROUND The National Kidney Foundation-Dialysis Outcomes Quality Initiative (NKF-DOQI) Clinical Practice Guidelines established a widely accepted set of recommendations for high-quality dialysis care. To enhance the End-Stage Renal Disease Core Indicators Project, an ongoing effort to assess and improve dialysis care in the United States, the Centers for Medicare and Medicaid Services (CMS) commissioned a project to develop clinical performance measures (CPMs) based on the NKF-DOQI guidelines. METHODS The CMS contracted with Qualis Health, a private nonprofit organization serving as a Medicare Quality Improvement Organization, to facilitate a 9-month project to develop dialysis CPMs with the participation of a broad range of stakeholders from the renal community. Work groups were established to develop CPMs addressing 4 areas: hemodialysis adequacy, peritoneal dialysis adequacy, vascular access management, and anemia management. The NKF-DOQI guidelines were prioritized based on the strength of the evidence supporting the guidelines, the feasibility of developing performance measures, and the significance of the areas addressed to the quality of care delivered to dialysis patients. Expert panels developed data specifications, sampling approaches, data-collection tools, and analytic strategies. RESULTS Sixteen CPMs were developed based on 22 of 114 NKF-DOQI guidelines. After establishing reliability through field-testing of data-collection instruments, the CPMs were applied to a sample of 8,838 randomly selected hemodialysis patients and 1,650 randomly selected adult peritoneal dialysis patients in summer 1999. CONCLUSION The development of CPMs based on the NKF-DOQI Clinical Practice Guidelines for dialysis care was accomplished in a timely and effective manner by engaging a broad range of stakeholders and technical experts. The CPMs are important tools to assess and improve the quality of dialysis care in the United States. Few comparable efforts exist in other fields of medicine.

Variability in Reasons for Hemodialysis Catheter Use by Race, Sex, and Geography: Findings From the ESRD Clinical Performance Measures Project

BACKGROUND Race, sex, and geographic differences in hemodialysis vascular access use have been reported, but differences in reasons for catheter use have not been assessed. STUDY DESIGN Cross-sectional. SETTING & PARTICIPANTS Data obtained from the 2005 Centers for Medicare & Medicaid Services End-Stage Renal Disease Clinical Performance Measures Project for adult hemodialysis patients. PREDICTORS Race, sex, and geographic region. OUTCOMES & MEASUREMENTS Reasons for catheter use were categorized as short term and long term. Race, sex, and geographic associations with reasons were assessed by using bivariate analyses and multivariate logistic regression. RESULTS Of 8,479 hemodialysis patients, 3,302 (39%) used a fistula, 2,725 (32%) used a graft, and 2,299 (27%) used a catheter. We placed 857 patients with a catheter (37%) in the short-term-reason cohort and 1,404 (61%) in the long-term-reason cohort, and 38 (2%) lacked information to be placed. Reasons for catheter use were independently associated with race, sex, and geographic region. Whites were 43%, 49%, and 34% less likely than African Americans to use a catheter because of graft maturation, graft interruption, and all vascular access sites exhausted and 70% and 40% more likely because of fistula maturation and no fistula or graft surgically planned, respectively. Men were 50% less likely than women to use a catheter because of graft interruption and 80% more likely because of fistula maturation. Geographic end-stage renal disease network was associated with catheter use because of fistula maturation (P = 0.03), no fistula or graft surgically created (P < 0.001), and no fistula or graft surgically planned (P = 0.05). LIMITATIONS The cross-sectional study design precludes our ability to assess trends over time in reasons for catheter use. Associations were assessed for a limited set of variables. CONCLUSION Race, sex, and geographic differences in reasons for hemodialysis catheter use exist. Understanding these differences may aid in developing strategies to decrease catheter initiation rates.

Can dialysis therapy be improved? A report from the ESRD Core Indicators Project

We assessed the association between quality improvement interventions conducted during the End-Stage Renal Disease (ESRD) Core Indicators Project and changes in the adequacy of hemodialysis between 1993 and 1996. Improvement of hemodialysis adequacy was measured by baseline and annual urea reduction ratios (URRs) in representative samples of ESRD Network patients. Random samples of in-center hemodialysis patients aged 18 years and older who had received hemodialysis during the fourth quarters of 1993, 1994, 1995, and 1996 were used to calculate Network-specific outcomes. A mean URR was calculated for each patient using the first pretreatment and posttreatment blood urea nitrogen for October, November, and December of each study year. Both national and Network-specific interventions were used to provide feedback reports and technical assistance to treatment centers to foster improvement in hemodialysis adequacy. All Networks distributed reports on the patterns of treatment center URR levels and physician and patient educational materials to each center in the Network. Each Network selected an annual 10% sample of treatment centers in 1994 and 1995 and conducted quality improvement activities to assist the selected centers to improve dialysis adequacy. We defined Network-specific interventions by a survey of the 18 Networks conducted during 1995 to determine the characteristics of Network-specific activities used to improve adequacy of hemodialysis. The outcome of interest was the change over time in Network-specific URR value. Sustained improvement in the URR occurred within all 18 Networks between 1993 and 1996. The mean national URR increased from 62.7% in 1993 to 66. 8% in 1996. The proportion of patients with URR >/= 65% increased from 43% in 1993 to 68% in 1996. Networks reported implementing a variety of intervention strategies that included educational activities, continuous quality improvement workshops, on-site assistance, and supervision of selected treatment facilities until care improved. Network-specific interventions independently associated with an increased rate of improvement in URR included prolonged supervision of the selected facilities. We concluded that the sustained improvement in hemodialysis care that occurred after the inception of the ESRD Core Indicators Project was associated with specific ESRD Network interventions.

Subcutaneous erythropoietin results in lower dose and equivalent hematocrit levels among adult hemodialysis patients: Results from the 1998 End-Stage Renal Disease Core Indicators Project

The National Kidney Foundations Dialysis Outcome Quality Initiative (NKF-DOQI) guidelines recommend that epoetin alfa should be administered by the subcutaneous route in hemodialysis patients. We determined whether hematocrit levels in hemodialysis patients differed by route of epoetin alfa administration after controlling for demographic factors and iron status. Data were available for 7,092 of the 7,658 patients randomly chosen for inclusion in the 1997 Health Care Financing Administration Core Indicators sample. Epoetin alfa was administered to 96% of the study cohort and was administered subcutaneously in 10% of patients. After controlling for hematocrit, patient characteristics, adequacy of dialysis, iron status, serum albumin, postdialysis weight, and duration of dialysis, the epoetin alfa dose by the intravenous route was 193.6 units/kg/wk (95% confidence interval, 189.5 to 197.8 units/kg/wk) compared with 167.4 units/kg/wk (95% confidence interval, 153.9 to 180.8 units/kg/wk) for the subcutaneous route (P < 0.001). The mean hematocrit for the subcutaneous route was 32.7% +/- 3.4% and for the intravenous route was 33.0% +/- 3.2% (P < 0.05). Factors independently associated with increased hematocrit included male gender, white race, older patient age, greater number of years on dialysis, higher serum albumin concentration, higher urea reduction ratio, and percent transferrin saturation (all P < 0.001). After controlling for patient factors and weekly epoetin alfa dose, there was no association between route of epoetin alfa administration and hematocrit level (P = 0.144). Patients receiving epoetin alfa by the subcutaneous route had comparable hematocrit values using a lower epoetin alfa dose than patients receiving epoetin alfa intravenously. These data support the NKF-DOQI recommendation that epoetin alfa be administered subcutaneously in long-term hemodialysis patients.