Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 September 15; 335(7619): 532.
PMCID: PMC1976521

Off-label use of erythropoietin may be harmful, doctors are told

Doctors should be vigilant about the off-label use of erythropoietin to treat anaemia in critically ill patients, says an early release editorial in the journal of the Canadian Medical Association.

The editorial (CMAJ 2007;177:697 doi: 10.1503/cmaj.071223) is available on the journal's website ( and is based on a meta-analysis in the journal.

Recombinant erythropoietin, a complex glycoprotein hormone, is approved for the treatment of anaemia in patients on dialysis, patients who have had major surgery, and those undergoing treatment for cancer.

When used off label to treat critically ill patients the drug, which costs about $400 (£200; €290) a dose, will save, on average, less than one unit of blood, will not improve clinical outcomes, and may increase the likelihood of thrombotic complications, says the editorial, by Paul Hébert, editor in chief of CMAJ, and Matthew Stanbrook, the journal's deputy editor (scientific).

The editorial's conclusion is based on findings from a meta-analysis of nine randomised controlled trials (CMAJ 2007;177:725-34 doi: 10.1503/cmaj.071055) and a commentary on the use of erythropoietin in critically ill patients (CMAJ 2007;177:747-9 doi: 10.1503/cmaj.071150), also available on the journal's website.

It praises the drug firms Johnson & Johnson and its subsidiary Janssen-Ortho for investing in clinical trials to find an indication for the use of erythropoietin in critically ill patients, but it says that in the United States the hormone's manufacturers have promoted it aggressively through direct to consumer advertising and incentive payments to doctors, prompting an investigation by the US Congress.

It also gives other examples of other drugs being promoted for off-label use without sufficient evidence of efficacy, saying that drug manufacturers used a loophole to bypass drug regulation.

The meta-analysis, which compared erythropoietin with placebo or no intervention, showed that the drug had no significant effect on overall mortality (odds ratio 0.86 (95% confidence interval 0.71 to 1.05); I2=0). The treatment and control groups did not differ in length of stay in hospital or intensive care unit or in the duration of mechanical ventilation in the three studies that reported these outcomes.

When compared with placebo erythropoietin reduced the chance of a patient receiving at least one transfusion (odds ratio 0.73 (0.64 to 0.84); I2=0.5%). The mean number of units of blood transfused per patient was 0.41 (0.1 to 0.74; I2=79%) units less in the erythropoietin group. Only one study provided detailed reports of adverse events, and none systematically evaluated all patients for venous thromboembolism.

In his commentary Howard Corwin, of the critical care medicine department, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire, who was involved in four of the studies included in the meta-analysis, presents an overview of the potential role of erythropoietin in critically ill patients.

He says that the drug seemed to strongly decrease mortality among patients with trauma who were in intensive care for more than 48 hours, although it did not do so among either surgical patients without trauma or non-surgical patients without trauma. Among critically ill patients, however, use of erythropoietin seems to be associated with a higher risk of thrombotic events.

The commentary concludes that the evidence doesn't support use of erythropoietin in surgical or non-surgical patients without trauma who are admitted to intensive care unless they have an approved indication for use of the drug, and that administering it would expose them to potential risks with no identifiable benefits in terms of lessened risk of transfusions or mortality.

The publication of the editorial and meta-analysis in the CMAJ coincides with the publication of two articles on the same subject in the New England Journal of Medicine (2007;357:965-76; 2007;357;1037-9).

(See Short Cuts doi: 10.1136/bmj.39332.399259.BE.)

Articles from The BMJ are provided here courtesy of BMJ Publishing Group