Abstract
Background and objectives: In hemodialysis patients, both hemoglobin variability and targeting normalization of hemoglobin may have adverse consequences. There are few data on epoetin management in patients achieving high hemoglobin levels.
Design, setting, participants, & measurements: Maintenance hemodialysis patients within Dialysis Clinic Inc. (DCI) who were treated with a 20 to 30% epoetin dose reduction versus epoetin discontinuation following achievement of a hemoglobin level of ≥13 g/dl were evaluated. The primary outcome was nadir hemoglobin within 2 months of epoetin reduction or discontinuation.
Results: There were 2789 patients in whom epoetin was discontinued and 754 patients in whom epoetin was reduced after hemoglobin ≥13 g/dl. Patients treated with reduction received significantly more epoetin over the subsequent 2 months. More patients dropped below 11 (21.5 versus 10.1%) and 10 g/dl (7.2 versus 3.6%) within 2 months of discontinuing epoetin, whereas reduction was associated with more frequent nadir hemoglobin levels >12 g/dl (31.1 versus 62.8%). In multivariable models including age, ferritin, albumin, and baseline epoetin dose, discontinuation was associated with nadir hemoglobin below 10 g/dl [OR = 1.91 (CI: 1.22, 2.99)], whereas reduction was associated with a hemoglobin nadir above 12 g/dl [OR = 4.41 (CI: 3.63, 5.37)].
Conclusions: In hemodialysis patients with baseline hemoglobin >13 g/dl, epoetin discontinuation is associated with lower epoetin use and a higher probability of reaching a hemoglobin target range of 10 to 12 g/dl within 2 months; discontinuation is also associated with a higher likelihood nadir hemoglobin <10 g/dl.
Erythropoietin deficiency is an important contributor to anemia in individuals with chronic kidney disease (CKD) (1). Epoetin therapy was approved for the treatment of anemia of CKD based on data showing decreased transfusion requirements and improved health-related quality of life in dialysis patients with severe anemia (2); subsequent small studies showed similar benefits (3–8). Notably, many of these benefits were not seen in more recent trials targeting normalization of hemoglobin in late-stage CKD and dialysis patients (9,10), with several larger trials raising substantial concerns regarding either a lack of benefit or potential harms associated with targeting hemoglobin “normalization” (to >13 g/dl) (10–13). These concerns, when balanced against cohort data showing potential risks associated with marked hemoglobin variability (14,15), result in substantial uncertainty regarding the optimal anemia management strategy in dialysis (16).
In April 2006, perhaps reflecting a desire to minimize hemoglobin variability, the Center for Medicare and Medicaid Services (CMS) implemented a new EPO Monitoring Policy allowing reimbursement of continued epoetin dosing at hemoglobin levels >13 g/dl with a mandatory 25% reduction in total epoetin administered for patients exceeding this level (17). This policy change mirrored concurrent recommendations from the National Kidney Foundation's Kidney Disease Outcome Quality Initiative (KDOQI), which had been updated to eliminate an upper hemoglobin target level. After publication of the Correction of Hemoglobin Outcomes in Renal Insufficiency (CHOIR) study and the Cardiovascular Risk Reduction by Early Anemia Treatment with Epoetin Beta (CREATE) study, KDOQI revisited the anemia guideline, recommending a target hemoglobin level of 11 to 12 g/dl (18). This recommendation differs from the U.S. Food and Drug Adminstration (FDA) and CMS, both of which now recommend a target of 10 to 12 g/dl (19).
A prior cross-sectional analysis by our group using Dialysis Clinic Inc (DCI) data showed that reduction versus discontinuation of epoetin at a hemoglobin level of 13 g/dl resulted in significantly more hemoglobin levels >13 g/dl, fewer levels between 11 and 12.9 g/dl, and no difference in the proportion of levels <11 g/dl each month (20). This study builds on this observation by evaluating longitudinal changes in hemoglobin levels in individual hemodialysis patients with hemoglobin ≥ 13 g/dl treated with reduction or discontinuation of epoetin to help inform best management of high hemoglobin levels in these individuals. Using DCI data and taking advantage of a change in epoetin dosing patterns in response to the CMS policy change in 2006, we evaluated 2-month nadir hemoglobin levels after reduction or discontinuation of epoetin after achieving hemoglobin ≥13 g/dl and explored individual patient factors associated with more precipitous drops in hemoglobin level.
Materials and Methods
Subjects
Patients receiving hemodialysis in DCI units between October 2005 and September 2006 with hemoglobin levels ≥13 g/dl were eligible; each patient could be included only once. To maximize the likelihood that changes in hemoglobin were driven by anemia management strategies rather than acute illness or changes in epoetin dose motivated by missed sessions, we excluded individuals who did not have three in-center hemodialysis treatments the week before measuring the hemoglobin levels ≥13 g/dl and individuals who did not have at least 20 in-center hemodialysis treatments over the subsequent 2 months after the dose change. Because epoetin reduction was the favored strategy for only 7 months at DCI, we chose a follow-up of 2 months to balance maximizing our available data with the knowledge that red blood cell survival in hemodialysis patients may be only 20 to 35 days (1,21).
Epoetin Dosing
In 2004, DCI introduced a computer-assisted EPO dosing algorithm (22). Use of this algorithm was optional and, through 2007, was used by a minority of DCI units. Units not using the algorithm determined epoetin dose manually or pursuant to local paper-based protocols. Through April 30, 2006, the algorithm mandated discontinuation of epoetin at hemoglobin ≥13 g/dl. A revised algorithm, implemented on May 1, 2006, recommended a dose reduction starting at a minimum of 25% that could increase to as high as 75% over the course of a month for sustained hemoglobin values >13 g/dl. Iron was dosed at the discretion of the treating physician. All patients received erythropoietin alfa (Epogen; Amgen Inc, Thousand Oaks, CA) by intravenous administration during dialysis; the dose administered was recorded at each treatment. For patients to be classified as “discontinuation,” they had to receive epoetin the week before reaching hemoglobin levels ≥13 g/dl and could not receive any epoetin for at least a week after a measurement of ≥13 g/dl; similarly, to be classified as “reduction,” patients needed to be receiving epoetin in the week leading up to the measurement with a reduction in epoetin dose of 20 to 30% in the ensuing week (to account for variability introduced by rounding).
Covariates
Demographic variables were obtained from the DCI database. Frequency of hemoglobin measurement was at the discretion of the individual dialysis unit, typically occurring once to twice monthly; in patients where epoetin was discontinued by protocol, hemoglobin was assessed weekly or biweekly to facilitate timely reinitiation. Nearly 97% of hemoglobin levels were measured at DCI's central laboratory in Nashville, TN. The remainder were measured at local laboratories. Most other measures occurred at DCI's central laboratory, including ferritin, transferrin saturation, serum albumin (bromocresol green), intact parathyroid hormone (PTH, Diasorin assay), and pre- and postdialysis urea nitrogen levels to determine single pool Kt/V. To minimize heterogeneity associated with laboratory variability, baseline values for covariates were determined using the 2-month average of ferritin, transferrin saturation, albumin, and Kt/V. PTH was generally measured quarterly, with the most proximate measure used.
Outcomes
The primary outcome was nadir hemoglobin within 2 months of a dose change for hemoglobin ≥13 g/dl, with multivariable analyses assessing the impact of baseline epoetin dose (before dose reduction or discontinuation) on subsequent nadir hemoglobin levels in each patient, as well as other variables that were associated with nadir hemoglobin <10 g/dl. We also examined the proportion of individuals maintaining hemoglobin levels >12 g/dl after 2 months.
Statistical Analyses
Two-sample t tests and the Wilcoxon rank test were used for comparisons of continuous variables. Pearson χ2 tests were used for categorical variables. Logistic regression was used to determine variables associated with nadir hemoglobin levels <10 and >12 g/dl. Candidate variables for multivariable regression models included age, gender, race, transferrin saturation, ferritin, intact PTH, time on dialysis, total number of hemoglobin measurements, weight-adjusted dose of epoetin administered the week before reaching hemoglobin levels ≥13 g/dl, serum albumin level, spKt/V, and hemoglobin level at the time of epoetin dose change. Stepwise regression was used, with variables retained at P < 0.05. A sensitivity analysis limited the study population to patients receiving hemodialysis in units using a computer-assisted protocol. SAS version 9.1 (SAS Institute, Cary, NC) was used for all analyses.
Results
Baseline Characteristics
There were 2789 patients in whom epoetin was discontinued and 754 patients in whom epoetin was reduced in response to hemoglobin ≥13 g/dl. Individuals in whom epoetin was discontinued were younger, were more likely to be African American, had higher hemoglobin levels at the time of dose change, and had higher ferritin and iron saturation during the ensuing 2 months. The total dose of epoetin given during the 2 months after dose change was significantly higher in individuals treated with a reduction strategy (Table 1).
Table 1.
Baseline and follow-up data for dialysis patients with either discontinuation or reduction of epoetin at hemoglobin ≥13 g/dl
| Discontinuation (n = 2789) | Reduction (n = 754) | P | |
|---|---|---|---|
| Age (years) | 62.2 ± 15.2 | 63.4 ± 15.3 | 0.05 |
| Women | 1301 (46.65%) | 1337 (44.69%) | 0.34 |
| Race | |||
| White | 1332 (47.8%) | 434 (57.6%) | <0.001 |
| Black | 1307 (46.9%) | 253 (33.6%) | |
| Other | 150 (5.4%) | 67 (8.9%) | |
| Hemoglobin | 13.5 (13.2 to 13.9) | 13.2 (13.1 to 13.5) | <0.001 |
| Transferrin saturation | 25 (20 to 32) | 22 (17 to 29) | <0.001 |
| Ferritin | 639 (406 to 879) | 547 (301 to 769) | <0.001 |
| Albumin | 3.7 (0.4) | 3.8 (0.3) | 0.11 |
| Intact PTH | 189 (105 to 315) | 195 (107 to 323) | 0.29 |
| Months of dialysis | 27 (8 to 59) | 24 (4 to 56) | 0.01 |
| spKt/V | 1.59 ± 0.31 | 1.57 ± 0.31 | 0.05 |
| Cause of ESRD | |||
| Glomerulonephritis | 328 (11.8%) | 80 (10.6%) | 0.07 |
| Hypertension | 835 (29.9%) | 194 (25.7%) | |
| Diabetes | 1181 (42.3%) | 354 (47.0%) | |
| Cystic/hereditary | 328 (11.8%) | 89 (11.8%) | |
| Other | 117 (4.2%) | 37 (4.9%) | |
| During follow-up | |||
| Sessions without Epo | 9 (6 to 15) | n/a | n/a |
| Total Epo administered | 100 (54 to 174) | 144 (82 to 240) | <0.001 |
| Hemoglobin tests | 5 (4 to 6) | 4 (3 to 5) | <0.001 |
Continuous data are mean ± SD or median (25th to 75th percentile). Transferrin saturation is in %, ferritin in mg/dl, albumin and hemoglobin in g/dl, and intact parathyroid hormone (PTH) in pg/ml. Total Epo administered is per 1000 IU.
Comparison of Treatment Strategies
Discontinuation of epoetin was associated with more patients achieving nadir hemoglobin within the range specified by either the FDA (10 to 12 g/dl) or KDOQI guidelines (11 to 12 g/dl), whereas the reduction strategy was associated with more frequent 2-month nadir levels >12 g/dl (Table 2 and Figure 1). Similar results were seen after stratifying by the dose of epoetin received the week before reaching hemoglobin levels ≥13 g/dl (Table 3).
Table 2.
Frequency of nadir hemoglobin levels achieved within 2 months after epoetin dose change for hemoglobin ≥13 g/dl
| Discontinuation (n = 2789) | Reduction (n = 754) | P | |
|---|---|---|---|
| <10 g/dl | 201 (7.2%) | 27 (3.6%) | <0.001 |
| 10 to 10.9 g/dl | 399 (14.3%) | 49 (6.5%) | |
| 11 to 12 g/dl | 1321 (47.4%) | 205 (27.2%) | |
| >12 g/dl | 868 (31.1%) | 473 (62.8%) |
Figure 1.
Nadir hemoglobin levels within 2 months after either discontinuation (dark bars/line) or reduction (light bars/line) of epoetin. Trend lines represent a running average of the four hemoglobin values before and the four values after each specific hemoglobin point.
Table 3.
Nadir hemoglobin level as a function of the dose of epoetin administered per dialysis session in the week before achieving hemoglobin ≥13 g/dl
| <5000 Units |
5000 to 10,000 Units |
>10,000 Units |
||||
|---|---|---|---|---|---|---|
| Discontinuation (n = 956) | Reduction (n = 223) | Discontinuation (n = 918) | Reduction (n = 237) | Discontinuation (n = 915) | Reduction (n = 294) | |
| <10 g/dl | 60 (6.3%) | 9 (4%) | 60 (6.5%) | 2 (0.8%) | 81 (8.9%) | 16 (5.4%) |
| 10 to 10.9 g/dl | 131 (13.7%) | 16 (7.2%) | 123 (13.4%) | 15 (6.3%) | 145 (15.9%) | 18 (6.1%) |
| 11 to 12 g/dl | 468 (49%) | 83 (37.2%) | 432 (47.1%) | 59 (24.9%) | 421 (46%) | 63 (21.4%) |
| >12 g/dl | 297 (31.1%) | 115 (51.6%) | 303 (33%) | 161 (67.9%) | 268 (29.3%) | 197 (67%) |
P value for all within-group comparisons between discontinuation and reduction <0.001.
A total of 228 patients (6.4%) had a hemoglobin nadir <10 g/dl, including 201 (7.2%) who had epoetin discontinued and 27 (3.6%) who had the dose reduced. In univariate analyses, individuals with nadir hemoglobin <10 g/dl were younger, more often women, and more likely to have discontinued epoetin (Table 4). In multivariable models, discontinuation of epoetin remained associated with increased likelihood of nadir hemoglobin <10 g/dl (Table 5). In analyses examining high nadir hemoglobin levels within 2 mo of an epoetin prescription change, reduction rather than discontinuation of epoetin, older age, higher serum albumin, more frequent hemoglobin measurement during this time period, lower Kt/V, lower ferritin, and higher baseline epoetin dose per kilogram body weight were associated with nadir hemoglobin level >12 g/dl (Table 5).
Table 4.
Patient characteristics stratified by nadir hemoglobin within 2 months after epoetin dose change
| Hgb < 10(n = 228) | Hgb 10 to 12(n = 1974) | Hgb > 12(n = 1341) | P | |
|---|---|---|---|---|
| Age (years) | 60.1 (16.5) | 62.3 (15) | 63.1 (15.4) | 0.04 |
| Women | 122 (53.5%) | 954 (48.3%) | 562 (41.9%) | <0.001 |
| Race | ||||
| White | 106 (46.5%) | 960 (48.6%) | 700 (52.2%) | 0.01 |
| Black | 108 (47.4%) | 908 (46%) | 545 (40.6%) | |
| Other | 14 (6.1%) | 106 (5.4%) | 97 (7.2%) | |
| Treatment strategy | ||||
| Discontinuation | 201 (88.16%) | 1720 (87.1%) | 868 (64.7%) | <0.001 |
| Reduction | 27 (11.8%) | 254 (12.9%) | 473 (35.3%) | |
| Hemoglobin | 13.4 (13.2 to 13.8) | 13.4 (13.1 to 13.7) | 13.5 (13.2 to 14) | <0.001 |
| Transferrin saturation | 24 (19 to 31) | 25 (19 to 32) | 24 (20 to 31) | 0.77 |
| Ferritin | 693 (400 to 967) | 651 (416 to 871) | 563 (332 to 794) | <0.001 |
| Albumin | 3.6 (0.4) | 3.7 (0.4) | 3.8 (0.4) | 0.003 |
| Intact PTH | 190 (98 to 294) | 192 (105 to 326) | 188 (107 to 305) | 0.41 |
| Months of dialysis | 32 (9 to 64) | 28 (9 to 59) | 22 (4 to 56) | <0.001 |
| spKt/V | 1.61 (0.36) | 1.60 (0.30) | 1.56 (0.30) | <0.001 |
| Cause of ESRD | ||||
| Glomerulonephritis | 26 (11.4%) | 237 (12.1%) | 145 (10.8%) | 0.01 |
| Hypertension | 63 (27.6%) | 590 (29.9%) | 376 (28%) | |
| Diabetes | 89 (39%) | 824 (41.7%) | 622 (46.4%) | |
| Cystic/hereditary | 40 (17.5%) | 244 (12.4%) | 133 (9.9%) | |
| Other | 10 (4.4%) | 79 (4%) | 65 (4.9%) | |
| During follow-up | ||||
| Sessions without Epo | 9 (6 to 12) | 9 (6 to 14) | 12 (7 to 20) | <0.001 |
| Total Epo administered | 176 (99 to 274) | 106 (59 to 184) | 104 (54 to 180) | <0.001 |
| Hemoglobin tests | 5 (4 to 6) | 4 (3 to 5) | 5 (3 to 6) | <0.001 |
Continuous data are mean ± SD or median (25th to 75th percentile). Transferrin saturation is in %, ferritin in mg/dl, albumin and hemoglobin in g/dl, and intact parathyroid hormone (PTH) in pg/ml. Sessions without Epo reflects only individuals in whom Epoetin was discontinued. Total Epo administered is per 1000 IU.
Table 5.
Factors associated with nadir hemoglobin <10 and >12 g/dl within 2 months [odds ratio (95% confidence interval)]
| Primary Analyses |
Computer to Assisted Protocol Subgroup |
|||
|---|---|---|---|---|
| Hemoglobin <10 g/dl | Hemoglobin >12 g/dl | Hemoglobin <10 g/dl | Hemoglobin >12 g/dl | |
| Reduction versus discontinuation | 0.53 (0.34, 0.82) | 4.41 (3.63, 5.37) | 0.55 (0.24, 1.25) | 6.31 (4.19, 9.51) |
| Age (per 10 years) | 0.87 (0.79, 0.96) | 1.10 (1.04, 1.16) | 0.98 (0.80, 1.19) | 1.10 (0.96, 1.26) |
| Ferritin (per 100 mg/dl) | 1.07 (1.03, 1.11) | 0.96 (0.94, 0.99) | 1.04 (0.94, 1.15) | 0.91 (0.86, 0.97) |
| Albumin (per g/dl) | 0.56 (0.38, 0.82) | 1.40 (1.10, 1.77) | 0.63 (0.25, 1.38) | 1.71 (0.98, 2.99) |
| spKt/V | — | 0.64 (0.49, 0.84) | — | 0.70 (0.37, 1.35) |
| Epoetin dose by weight in kga | 1.04 (1.01, 1.07) | 0.98 (0.96, 1.00) | 1.04 (0.96, 1.12) | 0.98 (0.94, 1.04) |
| Hemoglobin tests | 1.15 (1.05, 1.25) | 2.03 (1.76, 2.35) | 1.04 (0.84, 1.28) | 1.16 (1.04, 1.30) |
Results for the primary and subgroup analyses each reflect two separate models. Odds ratio represents each 20 IU/kg increase in epoetin dose. spKt/V was not included in models examining hemoglobin <10 g/dl. All variables significant in the primary analyses were included in the subgroup analyses.
The average per session dose of epoetin given the week before achieving hemoglobin ≥13 g/dl.
In the subgroup analysis of 807 patients whose epoetin dose was managed using the computer-assisted protocol, 546 discontinued epoetin and 261 had a reduced dose. The discontinuation strategy was associated with more patients with nadir hemoglobin level <10 g/dl [38 (7.0%) versus 11 (4.2%)] and more patients within contemporary target ranges proposed by the FDA (10 to 12 g/dl) and KDOQI (11 to 12 g/dl). The proportion of patients with nadir hemoglobin level >12 g/dl was higher with the reduction strategy [204 (37.4%) versus 186 (71.3%)]. Results of multivariable models examining this subgroup were similar to primary analyses (Table 5).
Discussion
In this study, we explored the effects of discontinuing versus reducing epoetin dose in dialysis patients who achieve a high hemoglobin level, and we showed that discontinuation results in less use of epoetin and more frequent hemoglobin levels within the KDOQI and FDA recommended range; however, discontinuation of epoetin is also associated with more frequent nadir hemoglobin levels below the current target of 10 g/dl and the 2007 KDOQI target of 11 g/dl. Notably, although this is a statistically significant difference, discontinuation of epoetin resulted in only a 3.6% greater incidence of very low hemoglobin levels within 2 months of this therapeutic change.
There are several important findings arising from this study. First, we provide information about the balance required between epoetin administration and achievement of target hemoglobin levels. This may have multiple consequences, including (1) increased health benefits for patients if specific hemoglobin levels rather than targets are truly associated with improved patient outcomes (23,24); (2) facilitating achievement of CMS quality performance guidelines, which, at the current time, evaluate dialysis units on their ability to maintain hemoglobin levels between 10 and 12 g/dl (19); and (3) generation of information necessary to determine cost-effective epoetin dosing strategies. For dialysis providers, maximizing results while minimizing costs will assume greater importance after the implementation of an expanded prospective payment system that will include erythropoiesis-stimulating agents (25). Assuming acquisition costs of epoetin of approximately $9 dollars per 1000 units (26), the potential additional cost associated with use of a reduction versus discontinuation strategy when an individual's hemoglobin level reaches 13 g/dl is $39,735 over a 2-month period for every hundred patients with high hemoglobin level at baseline. The benefit associated with this expenditure is that 3.6 individuals who would otherwise have had a nadir hemoglobin <10 g/dl will not reach this threshold within 2 months. Further study is required to elucidate whether potential harms associated with hemoglobin cycling make this expenditure cost effective.
Second, we explored factors contributing to a more precipitous drop in hemoglobin after reaching levels of 13 g/dl, because knowledge of these factors may facilitate development of individualized epoetin dosing strategies. In multivariable analyses, both lower serum albumin level (likely identifying either poor nutrition or an inflammatory state) and elevated serum ferritin level (likely identifying an inflammatory state) were associated with a higher likelihood of achieving a nadir hemoglobin level <10 g/dl. Additionally, individuals with a higher baseline dose of epoetin were more likely to have a precipitous drop, again likely consistent with pre-existing inflammation and/or bone marrow resistance to epoetin. Finally age had an inverse correlation with a decrease in hemoglobin in our study, such that older individuals were less likely to have a precipitous decline. Although the reason for this result is uncertain, it has been noted elsewhere (27).
Concerns over potential adverse effects associated with abrupt discontinuation of epoetin are rooted in red blood cell physiology. Erythropoietin maintains hemoglobin levels through two mechanisms. First, it prevents the apoptosis of early bone marrow red blood cell progenitors while stimulating the proliferation of these cells (28). Second, it prevents the selective hemolysis of the youngest circulating red blood cells, a process called neocytolysis, whereby the body adapts to excessive red blood cell mass by decreasing the circulating number of red blood cells (29). If neocytolysis is an important factor in dialysis patients, abrupt discontinuation of epoetin could produce an acute drop in hemoglobin levels. However, the significance of neocytolysis in dialysis patients remains uncertain, because erythropoietin deficiency is relative rather than absolute in this population (30–32); accordingly, other factors may influence whether neocytolysis occurs in individual dialysis patients (33,34).
Our findings are concordant with other studies. Berns et al. (27) described a direct correlation among mean corpuscular volume, ferritin level, and hemoglobin variability, whereas albumin and older age had an inverse correlation with hemoglobin variability. Similar to our findings, other laboratory parameters in the study by Berns et al., including iPTH and Kt/V, were not significantly correlated with hemoglobin variability in their study. In a follow-up study, Fishbane and Berns (35) noted that factors associated with drops in hemoglobin included both holding and reducing the baseline dose of epoetin, intercurrent infection, and discontinuation of iron. Similarly, Ebben et al. (14) found an association between hemoglobin variability and the number of comorbid conditions.
This study expands on a prior analysis by our group that showed a benefit on a dialysis provider level associated with discontinuation of epoetin at higher levels (20). In that study, there was no significant difference in the number of patients with a hemoglobin level <10 g/dl in any given month, whereas in this study, the discontinuation protocol was associated with significantly more patients with low nadir hemoglobin levels. This apparent discrepancy reflects the fact that this study population was limited to the small subset of patients in whom hemoglobin of 13 g/dl was reached, whereas the prior study included the entire dialysis population potentially affected by the protocol at any given time. Accordingly, although the twofold rise in individuals with nadir hemoglobin level <10 g/dl seems substantial, the total number of patients with this drop in hemoglobin level represents a very small proportion of the entire dialysis population.
There are several limitations to this study. First, detailed information about other anemia management strategies, including iron administration, was not available and may vary among dialysis units. We addressed this limitation by performing analyses isolated to the subset of units that were managed with the computer-assisted protocol throughout the study period, with the supposition that, within unit, iron strategies were not likely to substantially change over the course of the 14-month study period; these subgroup analyses showed similar results. Second, several baseline characteristics differed across management strategies. We adjusted for these differences in multivariable models and noted that similar covariates as described by others were associated with a decline in hemoglobin levels. Third, comorbidities that may affect hemoglobin level, including hospitalization, bleeding episodes, and infection, were not available for analysis. We minimized this effect by excluding all individuals who were missing more than three monthly dialysis sessions in the 2-month follow-up period. Additionally, we limited hemoglobin ascertainment to 2 months after epoetin dose change; whereas this may result in missing a more gradual achievement of low nadir hemoglobin, we felt that extant data on neocytolysis supported the proposition that a more rapid decline in the setting of abrupt epoetin discontinuation would be expected.
Our study also has substantial strengths. We analyzed a fairly large and representative dataset that is generalizable to real-world practices. Additionally, reflecting conservative epoetin utilization practices within DCI (36), we describe a population with relatively low utilization of epoetin; this is important, because it is likely that, moving forward, industry-wide epoetin utilization will decrease. Finally, we are able to incorporate multiple demographic characteristics and baseline iron status into multivariable models assessing associates of hemoglobin decline.
Conclusions
Once a maintenance hemodialysis patient achieves a hemoglobin level ≥13 g/dl, an epoetin discontinuation strategy is associated with a higher probability of reaching the hemoglobin target range stipulated by KDOQI or FDA guidelines within 2 months compared with a reduction strategy. However, a discontinuation strategy is associated with a higher likelihood of a low nadir hemoglobin level, whereas a reduction strategy is associated with significantly increased time at higher than desired hemoglobin levels. Reflecting current limitations in knowledge, it is possible that both strategies may be associated with adverse outcomes. This study provides important data on factors that are associated with more precipitous decline in hemoglobin levels and may enable cost–benefit analyses that will facilitate dialysis practices in years to come.
Disclosures
None.
Supplementary Material
Acknowledgments
We thank Jason Nelson, Marcia Landa, and Hocine Tighiouart for assistance with statistical analyses. We also thank DCI as a whole, and specifically, H. Keith Johnson, MD, Douglas Johnson, MD, Philip Zager, MD, and Karen Majchrzak, MS, for thoughtful comments and feedback. J.A.C. is supported by NIH Grant T32DK007777. D.E.W. is supported by NIH Grant K23DK071636.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
Supplemental information for this article is available online at http://www.cjasn.org/.
References
- 1.Eschbach JW: The anemia of chronic renal failure: Pathophysiology and the effects of recombinant erythropoietin. Kidney Int 35: 134–148, 1989 [DOI] [PubMed] [Google Scholar]
- 2.Eschbach JW, Abdulhadi MH, Browne JK, Delano BG, Downing MR, Egrie JC, Evans RW, Friedman EA, Graber SE, Haley NR, Korbet S, Krantz SB, Lundin AP, Nissenson AR, Ogden DA, Paganini EP, Rader B, Rutsky EA, Stivelman J, Stone WJ, Teschan P, Van Stone JC, Van Wyck DB, Zuckerman K, Adamson JW: Recombinant human erythropoietin in anemic patients with end-stage renal disease. Results of a phase III multicenter clinical trial. Ann Intern Med 111: 992–1000, 1989 [DOI] [PubMed] [Google Scholar]
- 3.Sikole A, Polenakovic M, Spirovska V, Polenakovic B, Masin G: Analysis of heart morphology and function following erythropoietin treatment of anemic dialysis patients. Artif Organs 17: 977–984, 1993 [DOI] [PubMed] [Google Scholar]
- 4.Wizemann V, Schafer R, Kramer W: Follow-up of cardiac changes induced by anemia compensation in normotensive hemodialysis patients with left-ventricular hypertrophy. Nephron 64: 202–206, 1993 [DOI] [PubMed] [Google Scholar]
- 5.Parfrey PS, Lauve M, Latremouille-Viau D, Lefebvre P: Erythropoietin therapy and left ventricular mass index in CKD and ESRD patients: A meta-analysis. Clin J Am Soc Nephrol 4: 755–762, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Canadian Erythropoietin Study Group: Association between recombinant human erythropoietin and quality of life and exercise capacity of patients receiving haemodialysis. BMJ 300: 573–578, 1990 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Macdougall IC, Lewis NP, Saunders MJ, Cochlin DL, Davies ME, Hutton RD, Fox KA, Coles GA, Williams JD: Long-term cardiorespiratory effects of amelioration of renal anaemia by erythropoietin. Lancet 335: 489–493, 1990 [DOI] [PubMed] [Google Scholar]
- 8.Robertson HT, Haley NR, Guthrie M, Cardenas D, Eschbach JW, Adamson JW: Recombinant erythropoietin improves exercise capacity in anemic hemodialysis patients. Am J Kidney Dis 15: 325–332, 1990 [DOI] [PubMed] [Google Scholar]
- 9.Pfeffer MA: Critical missing data on erythropoiesis-stimulating agents in CKD: First beat placebo. Am J Kidney Dis 51: 366–369, 2008 [DOI] [PubMed] [Google Scholar]
- 10.Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR, Okamoto DM, Schwab SJ, Goodkin DA: The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med 339: 584–590, 1998 [DOI] [PubMed] [Google Scholar]
- 11.Singh AK, Szczech L, Tang KL, Barnhart H, Sapp S, Wolfson M, Reddan D: Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med 355: 2085–2098, 2006 [DOI] [PubMed] [Google Scholar]
- 12.Pfeffer MA, Burdmann EA, Chen CY, Cooper ME, de Zeeuw D, Eckardt KU, Feyzi JM, Ivanovich P, Kewalramani R, Levey AS, Lewis EF, McGill JB, McMurray JJ, Parfrey P, Parving HH, Remuzzi G, Singh AK, Solomon SD, Toto R: A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. N Engl J Med 361: 2019–2032, 2009 [DOI] [PubMed] [Google Scholar]
- 13.Drueke TB, Locatelli F, Clyne N, Eckardt KU, Macdougall IC, Tsakiris D, Burger HU, Scherhag A: Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med 355: 2071–2084, 2006 [DOI] [PubMed] [Google Scholar]
- 14.Ebben JP, Gilbertson DT, Foley RN, Collins AJ: Hemoglobin level variability: Associations with comorbidity, intercurrent events, and hospitalizations. Clin J Am Soc Nephrol 1: 1205–1210, 2006 [DOI] [PubMed] [Google Scholar]
- 15.Gilbertson DT, Ebben JP, Foley RN, Weinhandl ED, Bradbury BD, Collins AJ: Hemoglobin level variability: associations with mortality. Clin J Am Soc Nephrol 3: 133–138, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Department of Health and Human Services Centers for Medicare & Medicaid Services: End-stage renal disease prospective payment system proposed rules. Fed Reg 74: FR49922–FR50102, 2009 [Google Scholar]
- 17.Department of Health & Human Services Center for Medicare and Medicaid Services: Pub 100–04 Medicare claims processing. National Monitoring Policy for EPO and Aranesp for end stage renal disease (ESRD) patients treated in renal dialysis facilities. Available at www.cms.hhs.gov/CoverageGenInfo/Downloads/ESA04012006.pdf Accessed March 22, 2010
- 18.KDOQI: KDOQI clinical practice guideline and clinical practice recommendations for anemia in chronic kidney disease: 2007 update of hemoglobin target. Am J Kidney Dis 50: 471–530, 2007 [DOI] [PubMed] [Google Scholar]
- 19.Epoetin alfa (marketed as Epoetin, Procrit) Label, 2/16/10. Available at http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/103234s5199lbl.pdf Accessed March 1, 2010
- 20.Weiner DE, Miskulin DC, Seefeld K, Ladik V, Zager PG, Singh AK, Johnson HK, Meyer KB: Reducing versus discontinuing erythropoietin at high hemoglobin levels. J Am Soc Nephrol 18: 3184–3191, 2007 [DOI] [PubMed] [Google Scholar]
- 21.Polenakovic M, Sikole A: Is erythropoietin a survival factor for red blood cells? J Am Soc Nephrol 7: 1178–1182, 1996 [DOI] [PubMed] [Google Scholar]
- 22.Miskulin DC, Weiner DE, Tighiouart H, Ladik V, Servilla K, Zager PG, Martin A, Johnson HK, Meyer KB: Computerized decision support for EPO dosing in hemodialysis patients. Am J Kidney Dis 54: 1081–1088, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Regidor DL, Kopple JD, Kovesdy CP, Kilpatrick RD, McAllister CJ, Aronovitz J, Greenland S, Kalantar-Zadeh K: Associations between changes in hemoglobin and administered erythropoiesis-stimulating agent and survival in hemodialysis patients. J Am Soc Nephrol 17: 1181–1191, 2006 [DOI] [PubMed] [Google Scholar]
- 24.Messana JM, Chuang CC, Turenne M, Wheeler J, Turner J, Sleeman K, Tedeschi P, Hirth R: Association of quarterly average achieved hematocrit with mortality in dialysis patients: A time-dependent comorbidity-adjusted model. Am J Kidney Dis 53: 503–512, 2009 [DOI] [PubMed] [Google Scholar]
- 25.Weiner DE, Watnick SG: The 2009 proposed rule for prospective ESRD payment: Historical perspectives and public policies—bundle up! Am J Kidney Dis 55: 217–222, 2010 [DOI] [PubMed] [Google Scholar]
- 26.Levinson DR: Department of Health and Human Services Office of the Inspector General. Medicare reimbursement for end stage renal disease drugs: Third quarter 2006. Available at: http://www.oig.hhs.gov/oei/reports/oei-03-06-00590.pdf Accessed June 1, 2010 [Google Scholar]
- 27.Berns JS, Elzein H, Lynn RI, Fishbane S, Meisels IS, Deoreo PB: Hemoglobin variability in epoetin-treated hemodialysis patients. Kidney Int 64: 1514–1521, 2003 [DOI] [PubMed] [Google Scholar]
- 28.Fried W: Erythropoietin and erythropoiesis. Exp Hematol 37: 1007–1015, 2009 [DOI] [PubMed] [Google Scholar]
- 29.Rice L, Alfrey CP: The negative regulation of red cell mass by neocytolysis: Physiologic and pathophysiologic manifestations. Cell Physiol Biochem 15: 245–250, 2005 [DOI] [PubMed] [Google Scholar]
- 30.Ross RP, McCrea JB, Besarab A: Erythropoietin response to blood loss in hemodialysis patients in blunted but preserved. ASAIO J 40: M880–885, 1994 [DOI] [PubMed] [Google Scholar]
- 31.Kato A, Hishida A, Kumagai H, Furuya R, Nakajima T, Honda N: Erythropoietin production in patients with chronic renal failure. Renal Failure 16: 645–651, 1994 [DOI] [PubMed] [Google Scholar]
- 32.Chandra M, Clemons GK, McVicar MI: Relation of serum erythropoietin levels to renal excretory function: Evidence for lowered set point for erythropoietin production in chronic renal failure. J Pediatr 113: 1015–1021, 1988 [DOI] [PubMed] [Google Scholar]
- 33.Rice L, Alfrey CP, Driscoll T, Whitley CE, Hachey DL, Suki W: Neocytolysis contributes to the anemia of renal disease. Am J Kidney Dis 33: 59–62, 1999 [DOI] [PubMed] [Google Scholar]
- 34.Means RT, Jr: Neocytolysis: from outer space to the dialysis unit. Am J Kidney Dis 33: 140–141, 1999 [DOI] [PubMed] [Google Scholar]
- 35.Fishbane S, Berns JS: Hemoglobin cycling in hemodialysis patients treated with recombinant human erythropoietin. Kidney Int 68: 1337–1343, 2005 [DOI] [PubMed] [Google Scholar]
- 36.Collins AJ, Ebben JP, Gilbertson DT: EPO adjustments in patients with elevated hemoglobin levels: Provider practice patterns compared with recommended practice guidelines. Am J Kidney Dis 49: 135–142, 2007 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.