Skip to main content
. Author manuscript; available in PMC: 2015 Mar 1.
Published in final edited form as: Med Care. 2014 Mar;52(0 3):S132–S139. doi: 10.1097/MLR.0b013e3182a53ca8

Table 1. Abbreviated protocol of hypothetical and emulated trials of anemia management strategies.

Component Hypothetical open-labeled, nonblinded randomized clinical trial Emulated trial using USRDS observational data
Aim To study the risks and benefits of epoetin therapy to target hematocrit (Hct) 34.5- 39.0% versus 30.0-34.5%. Same
Study Population Elderly ESRD patients with both diabetes and cardiovascular disease who initiated hemodialysis in US outpatient facilities between January 1, 2006 and December 31, 2008. Same.
Eligibility criteria Inclusion criteria: ≥65 years of age, initiated outpatient dialysis within 90 days of enrolling in the Medicare ESRD Program in 2006-2008, evidence of both diabetes and cardiovascular disease before or at baseline (end of the 3rd month of hemodialysis).
Exclusion criteria: history of cancer before ESRD, no epoetin therapy in the first 30 days of dialysis, in a non-dialysis facility (e.g., hospital) at baseline, kidney transplantation or peritoneal dialysis before baseline, use of darbepoetin before baseline, and hematocrit <24% at baseline.		 Inclusion criteria: Same.
Evidence of diabetes was ascertained as an underlying cause of ESRD, Medical Evidence Form (MEF) reporting of diabetes (on insulin, with oral medications, or without medications) or diabetic retinopathy, or a hospitalization with (primary or secondary) ICD-9 code 252.x during the 3 months before baseline. Evidence of cardiovascular disease was ascertained as MEF reporting of congestive heart failure, atherosclerotic heart disease, cerebrovascular disease, or peripheral vascular disease; ora hospitalization with (primary or secondary reason) ICD-9 codes 428.0 (congestive heart failure), 414.0 (atherosclerotic heart disease), 430-438 (cerebrovascular disease), or 443.9 (peripheral vascular disease) during the 3 months before the start of follow-up. Evidence of cancer was also obtained from the MEF file.
Exclusion criteria: Same.
In addition, patients were excluded if they had incomplete baseline covariates.
Follow-up Start: after completing 3 months of hemodialysis therapy
End: 6 months after baseline, death, switch to darbepoetin, or dropout/loss to follow-up, whichever happens first.		 Start: Same.
End: Same. Dropout/loss to follow-up defined as the earlier of

  • key data become unreliable, e.g., epoetin dose >0 even though Hct level was not reported in the claim

  • 30-day gap in outpatient dialysis or inpatient claims
Treatment assignment Patients are randomly assigned to one of the following two dynamic treatment strategies:

  1. Mid Hct Strategy: intravenous epoetin alfa to achieve and maintain hematocrit values of 34.5-< 39.0% or

  2. Low Hct Strategy: intravenous epoetin alfa to achieve and maintain hematocrit values of 30.0-<34.5%.


Under both strategies, monthly epoetin dose is changed according to the following rules:

  1. If previous hematocrit is below the target range, epoetin dose is increased by ≥10%;

  2. If previous hematocrit is within target range, epoetin dose is decreased by ≤10% times [hematocrit minus lower end of range] or increased by ≤10% times [upper end of range minus hematocrit];

  3. If previous hematocrit is above the target range, epoetin dose is decreased ≥25% or withheld.


The epoetin dose is left to the discretion of the treating physician during the month after the patient undergoes hemodialysis in a facility not participating in the study (e.g., hospital, hospice, nursing home or home health services) and after epoetin dose was withheld. The administration of IV iron is left to the discretion of the treating physician.		 Patients are classified as following one, both, or neither of the Mid Hct/Low Hct strategies.
If a patient's treatment data during the first month of follow-up are consistent with the rules on the left for

  • one strategy: the patient is assigned to that strategy

  • both strategies: we clone the patient's data and assign each clone to one of the two strategies

  • neither strategy: the patient is ineligible for the study.


If a patient has more than one dialysis claim during one month, only the data in first dialysis claim is used.
Endpoints Primary: all-cause mortality.
Secondary: a composite endpoint of mortality and a hospitalization for MI, stroke or congestive heart failure.		 Primary: Same.
Secondary: Same. Cardiovascular events are identified through ICD-9-CM codes for primary reason for hospitalization on Medicare hospital claims using the following International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) codes: MI: codes 410.xx (except 410.x2); CHF: codes 402.x1, 425.xx, 428.xx, 518.4, and 398.91; and Stroke: codes 430.xx, 431.xx, 432.xx, 433.xx, and 434.xx.
Statistical Analysis Intention-to-treat analysis:
Cox model with indicator for strategy (Mid or Low Hct) and with inverse probability (IP) weights to adjust for selection bias due to loss to follow-up. Weighted Survival curves under each strategy.
Per-protocol analysis:
Patients are artificially censored when they deviate from their assigned strategy. IP weights for artificial censoring are estimated as a function of history of epoetin dose, hematocrit, change in hematocrit values, iron treatment, hospitalization, and product terms between these variables.
IP weighted Cox model with indicator for strategy (Mid or Low Hct) and baseline risk factors. Standardized, weighted survival curves under each strategy.		 Intention-to-treat analysis:
Same, except that Cox model also includes baseline risk factors to adjust for baseline confounding: age at ESRD onset, race, gender, US geographic region, dialysis chain membership, patient BMI, Charlson Index score, tobacco use, drug/alcohol dependence, chronic obstructive pulmonary disease, serious conditions including amputation, inability to ambulate, and inability to transfer, predialysis hematocrit, predialysis epoetin use, average baseline hematocrit, average baseline epoetin dose, and average baseline iron dose. Survival curves are standardized to baseline risk factors.
Per-protocol analysis:
Same. Patients are not artificially censored if (a) hemodialysis takes place in a facility other than a Medicare-certified outpatient dialysis facility (e.g., hospital, hospice, nursing home or home health services), or (b) dose was withheld in previous month.