All you need to read in the other general journalsBMJ 2009; 338 doi: https://doi.org/10.1136/bmj.b2511 (Published 22 June 2009) Cite this as: BMJ 2009;338:b2511
Researchers criticised for stopping sepsis trial too early
Endotoxins, a feature of Gram negative sepsis, can be filtered out of the blood using a haemoperfusion column impregnated with polymyxin B. There’s some evidence that patients with the right kind of severe sepsis can benefit from this treatment, including a preliminary trial from Italy that recently reported a significant reduction in 28 day mortality in patients who underwent haemoperfusion with polymyxin B (11/34 (32%) v 16/30 (53%) in controls given conventional therapy alone; adjusted hazard ratio 0.36, 95% confidence interval 0.16 to 0.80)⇑. The 64 patients in this trial had abdominal sepsis, mostly from perforated bowel. Those given two sessions of haemoperfusion with polymyxin B had greater improvements in haemodynamics and organ function than controls given conventional therapy alone. The fall in mortality prompted an oversight board to stop the trial early.
Stopping the trial was a mistake, says a linked editorial (pp 2496-7). These data are preliminary and cannot establish benefit with any certainty, or rule out harm. The mortality results were a surprise and are insecure. Even the authors say their promising results must be confirmed in bigger trials.
Recruiting for such trials will be very difficult, says the editorial, now an oversight board has been persuaded that randomising is unethical in this setting. The available evidence simply isn’t enough to convince most regulatory authorities, nor should it. A definitive answer was possible and is now further away than ever. Evaluation could now stall, denying patients with a lethal condition access to a potentially life saving treatment. “The position taken by the oversight board . . . should be rescinded,” conclude the editorial’s authors.