Criteria abstracted from The Users' Guide to Medical Literature, from the Health Information Research Unit and Clinical Epidemiology and Biostatistics, McMaster University

Highlighted lines and questions below provide links to the pertinent description of criteria in The EBM User's Guide, now available at the Canadian Centres for Health Evidence

Article Reviewed:

Please visit the new Evidence Based Journal Club Reviews

Efficacy of recombinant human erythropoietin in critically ill patients: a randomized controlled trial.

Corwin HL, Gettinger A, Pearl RG, Fink MP, Levy MM, Shapiro MJ, Corwin MJ, Colton T.

JAMA. 2002;288(22):2827-35. [abstract]

Reviewed by Mehrengise Cooper, MBBS MRCP, Children's Hospital Boston

Review posted March 10, 2003

---

I. What is being studied?:

The study objective:

This group set out to assess the efficacy in critically ill patients of a weekly dosing schedule of recombinant human erythropoietin (rHuEPO) in order to decrease the occurrence of Red Blood Cell transfusion (RBC).

The study design:

This study was designed as a prospective, randomized, double-blind, placebo-controlled, multicenter trial, and was conducted between December 1998 and June 2001.

The patients included:

The patient population compromised all patients admitted to either a medical, surgical or a medical/surgical ICU (65 hospitals participated) who stayed in the ICU for at least two days. These patients were evaluated for study eligibility on ICU day three.

The inclusion criteria were as follows: aged 18 years and above, hematocrit (hct) <38%, provision of signed informed consent.

The patients excluded:

Renal failure with dialysis, uncontrolled hypertension, new-onset or uncontrolled seizures, acute burns, pregnancy or lactation, acute ischemic heart disease, acute gastrointestinal bleeding, prior treatment with rHuEPO, participation in another research protocol, expected ICU discharge within 48 hours of ICU day two.

The interventions compared:

On ICU day three patients received either the study drug (40,000 Units of rHuEPO subcutaneously) or placebo. This was administered on ICU day three and then weekly for those remained in the hospital, for three doses in total. If patients remained in the ICU on day 21, a fourth dose was administered.

The outcomes evaluated:

Outcomes were divided into primary and secondary efficacy end-points. The primary efficacy end-point was transfusion independence - this was measured by comparing the percentage of patients in each treatment group who received any RBC transfusion between study days 1 and 28. Secondary efficacy end-points were (a) cumulative RBC units transfused per patient through study day 28, (b) cumulative mortality through study day 28, (c) change in hemoglobin from baseline, and (d) time to first transfusion or death.

II. Are the results of the study valid?

Primary questions:

1. Was the assignment of patients to treatments randomized?

Yes - 1302 patients were enrolled and 650 were randomized to receive the study drug, with 652 to receive placebo. Randomization was by computer-generated random numbers and stratified by site.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Was followup complete?

Yes - all patients who entered into the trial were accounted for. The study was designed to follow-up patients for 28 days following randomization, which was study day 28. All patients were studied for the planned follow-up time, unless death occurred earlier. Death was one of the secondary efficacy end-points assessed.

Were patients analyzed in the groups to which they were randomized?

All patients were followed for the planned 28 days, unless death occurred and analysis of outcomes was planned on an intent-to-treat basis. (The patients who discontinued participation or who were lost to follow up - 31 in the placebo group and 36 in the treatment group - were all included in the final analysis.) When accounting for those patients who did not receive a transfusion, at either the time of study withdrawal, or were lost to follow-up, this group was considered not transfused for this analysis.

There was also a sub-group analysis, which analyzed patients undergoing transfusion and mortality based upon medical/surgical-trauma/surgical-non-trauma, Apache II score > 20/< 20, and age > 55/< 55.

Secondary questions:

3. Were patients, health workers, and study personnel "blind" to treatment?

This study was a double-blind trial. The study drug and placebo were identical in appearance and administered using identical means.

4. Were the groups similar at the start of the trial?

Yes - the baseline demographic details, laboratory values, admitting diagnoses and comorbidities were generally similar at the start of the study. There were statistically significant differences in a few variables, like serum iron and transferrin saturation, but these were not likely to be clinically significant. There were more patients with "primary hematologic disease" in the treatment group, but the numbers were small (4 in placebo group vs. 13 in the treatment group).

5. Aside from the experimental intervention, was the group treated consistently?

The groups were both treated equally in that all patients received iron supplements and similar transfusion practices were followed (pre-transfusion hemoglobin and hematocrit levels were similar in both groups). No comment is made about other aspects of care, such as nutrition, antacid therapy, etc.

To establish if the patient needed a transfusion, transfusion guidelines were prepared for the study. These were as follows: no RBC transfusion if hct > 27% / Hb > 9g/dl - unless a specific clinical indication deemed it necessary. If hct < 27% / Hb < 9 g/dl, then RBC transfusion was at the physician's discretion. There was no definite hct / Hb level to transfuse RBCs, which in study where the principle outcome measures were related to the number of RBC transfusions used, this seems surprising. On the other hand, this makes the study more realistic, where rigid guidelines to transfuse patients are rarely followed.

III. What were the results?

1. How large was the treatment effect?

The treatment effect in this study was to look at the efficacy of rHuEPO. Fewer patients who received rHuEPO received RBC transfusions than the group receiving placebo. The relative risk reduction calculates out at 16.5%, with an absolute risk reduction at 10% (50% vs. 60%). The odds ratio was 0.67 (95% CI 0.54-0.83). These results were unchanged when adjusting for baseline characteristics and where patients who were withdrawn or lost to follow-up were considered to have received a transfusion.

The number of RBC transfusions per patient was also lower in those receiving rHuEPO (median of 1 vs. 2 units per patient between groups). Time at risk for transfusion demonstrated a 19% relative risk reduction in RBC units transfused per day alive.

No other differences were demonstrated between these two groups and no differences were noted in subgroup analyses.

Survival analysis demonstrated no difference in transfusion rates between the groups until approximately one week into their hospital course, at which time the curves separate and the treatment group "began" receiving fewer transfusions.

2. How precise was the estimate of the treatment effect?

The odds ratio assessing the number of patients in each group requiring blood transfusion reflected a precise estimate (95% CI 0.54 to 0.83). The 95% CI for the relative risk reduction of 16.5% is 8% to 24%. So we can conclude, with 95% confidence, that the true relative risk reduction of transfusion with rHuEPO therapy is in this range.

IV. Will the results help me in caring for my patients?

1. Can the results be applied to my patient care?

As a pediatric intensivist, the patient population studied in this paper is similar in that the patients are critically ill, but they were all adults and the applicability to pediatric patients is uncertain. Those patients who require a stay on the ICU of > 72 hours include those with disease processes including acute respiratory distress syndrome, respiratory illness and postoperative patients. It is common practice to use a restrictive transfusion strategy in the majority of patients. In the patients with chronic disease, or a lengthy ICU course, then rHuEPO may be an alternative to RBC transfusion.

2. Were all clinically important outcomes considered?

The outcomes were divided into primary and secondary. It was necessary to assess transfusion independence as a primary outcome. Secondary outcomes looked at factors associated with critically ill patients Ð for example death. The other important and necessary outcomes also analyzed included length of hospital stay and factors related to mechanical ventilation, notably, ICU-free days and ventilator-free days, which are essential when studying any group of critically ill patients. A number of complications were assessed as well, including organ dysfunction, bleeding, arrhythmias, etc., and were no different between the two groups.

3. Are the likely treatment benefits worth the potential harms and costs?

As noted by the authors, the cost of rHuEPO is ~$400 per 40,000 U dose, in contrast to a unit of RBCs, which is ~$300-$400. The average treated patient had one less transfusion. Therefore, during an ICU admission, the cost for a critically ill patient who is treated with two or three doses of rHuEPO is $800-$1200, saving $300-$400. Therefore looking at costs, there is no benefit.

That said, the reduction in number of units of blood transfused might be of benefit. The treated group overall used 19% fewer units of blood. This can also be accomplished by using a restrictive transfusion strategy, and Hebert et al (1) have shown that restricting RBC transfusions to an Hb < 7 g/dl is safe unless the diagnosis is severe ischemic heart disease. In the Hebert et al. study, the restricted transfusion group received about the same amount of blood as the placebo group in the current study, suggesting rHuEPO can further reduce RBC use. However, in this study the decision to transfuse was at the discretion of the physician and there is no information in the study discussing how those decisions were made.

Avoiding blood transfusions will directly lead to a reduction in the morbidity and mortality caused by RBCs; it also reduces any potential medical errors associated with the transfusion process. RBCs - themselves a precious resource - could be saved.

Concerns have been raised regarding complications of the use of erythropoietin. Neutralizing antierythropoietin antibodies and pure-red cell aplasia have been reported in patients who have chronic renal disease, receiving epoetin for the treatment of anemia of chronic renal disease (2). These patients all developed severe transfusion dependent anemia following an initial response to epoetin. All patients had been receiving epoetin for at least 3 months (3-67 months), compared to critically ill patients who receive up to four doses of rHuEPO. It would be interesting to follow the patients in the current study to see if any developed antierythropoietin antibodies.

If rHuEPO were used in conjunction with a more restrictive transfusion guideline, the potential exists to further reduce the use of allogeneic blood transfusions in critically ill patients.

References

1. Hebert PC, Wells G, Blajchman MA, et al. A Multicenter, Randomized, Controlled Clinical Trial of Transfusion Requirements in Critical Care. New Engl J Med 1999; 340:409-417. [abstract] ; PedsCCM EB Journal Club Review
2. Casadevall N, Nataf J, Viron B, Kolta A, Kiladjian J, Martin-Dupont P, Michaud P, Papo T, Ugo V, Teyssandier I, Varet B, Mayeux P. Pure Red-cell aplasia and antierythropoietin antibodies in patients treated with recombinant erythropoietin. N Engl J Med 2002;346:469-475 [abstract]

Comments

Back to the EB Journal Club Index