Correspondence
In the PRORATA trial (Feb 6, p 463),1 use of a procalcitonin-guided strategy in the intensive-care unit reduced the number of days of antibiotic treatment, compared with a conventional strategy. No effect was seen on mortality and relapses of infection. Lila Bouadma and colleagues conclude that this strategy can be applied to most non-surgical patients, including those who are immunocompromised. However, it is critical to balance the benefits and risks of a procalcitonin-guided strategy in each subgroup of patients. In the procalcitonin group, antibiotics were encouraged for procalcitonin concentrations above 0·5 μg/L. Previous studies showed that increased concentrations of procalcitonin (up to 15 μg/L) may be found in patients treated with T-cell antibodies, alemtuzumab, interleukin 2, granulocyte transfusions, or patients with acute graft-versus-host disease or liver metatstasis.2–4 Therefore, procalcitonin should not be considered as an accurate marker of infection in patients that fall into the above categories, since a procalcitonin-guided strategy could result in treatment excess. In addition, the interferences between infection, malignant diseases, and immune treatment can hamper the interpretation of procalcitonin kinetics. In the PRORATA trial, patients with severe neutropenia were excluded, whereas those with mild-to-moderate neutropenia were included. To the best of our knowledge, guidelines recommend prolonging antibiotic treatment in patients with mildto-moderate neutropenia until neutropenia recovery and fever resolution. When appropriate antibiotics are given, a 5–7 day treatment duration is required before fever resolution.5 Guidelines also recommend that these patients have the option of systemic antibiotics to www.thelancet.com Vol 375 May 8, 2010
be given until 5–7 afebrile days have elapsed, unless the patient is unstable. If antibiotic treatment is stopped during neutropenia, the patient must be carefully monitored and antibiotics restarted immediately on the recurrence of fever or other evidence of bacterial infection.5 According to Bouadma and colleagues,1 few data are available about the 98 patients who were immunocompromised. We would like to know about mortality, the duration of antibiotic therapy, and the relapsing infection rate in these patients with neutropenia. We declare that we have no conflicts of interest.
*Djamel Mokart, Marc Leone
[email protected] Département d’Anesthésie et de Réanimation, Institut Paoli-Calmettes, 13273 Marseille Cedex 9, France (DM); and Service d’Anesthésie et de Réanimation, Hôpital Nord, Assistance PubliqueHôpitaux de Marseille, Université de la Méditerranée, Marseille, France (ML) 1
2
3
4
5
Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patients’ exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet 2010; 375: 463–74. Brodska H, Drabek T, Malickova K, et al. Marked increase of procalcitonin after the administration of anti-thymocyte globulin in patients before hematopoietic stem cell transplantation does not indicate sepsis: a prospective study. Crit Care 2009; 13: R37. Dornbusch HJ, Strenger V, Sovinz P, et al. Noninfectious causes of elevated procalcitonin and C-reactive protein serum levels in pediatric patients with hematologic and oncologic disorders. Support Care Cancer 2008; 16: 1035–40. Matzaraki V, Alexandraki KI, Venetsanou K, et al. Evaluation of serum procalcitonin and interleukin-6 levels as markers of liver metastasis. Clin Biochem 2007; 40: 336–42. Hughes WT, Armstrong D, Bodey GP, et al. 2002 guidelines for the use of antimicrobial agents in neutropenic patients with cancer. Clin Infect Dis 2002; 34: 730–51.
We believe that the PRORATA trial1 is of major concern, and potentially misleading, in suggesting that a strategy guided by procalcitonin could “reduce antibiotic exposure and selective pressure with no apparent adverse outcomes”. Lila Bouadma and colleagues assumed that a true excess mortality, or risk difference, in the procalcitonin
group of up to 9·9% would not be inferior. This equates to a number needed to harm of ten. They conclude that the significant reduction (p<0·0001) of 2·7 days in mean exposure to antibiotics is worth a possible extra death for every ten patients treated by the procalcitonin strategy. If a true excess mortality as great as 2% (one extra death for every 50 patients treated with the procalcitonin strategy) were considered acceptable for a few days fewer antibiotics, this would require a non-inferiority trial with 16 500 patients. A true excess mortality of 4%, similar to that seen in the trial, would require about 4220 patients. The PRORATA trial, which recruited 621 patients, was grossly underpowered to provide reliable evidence for policy, but could be valuable in a synthesis of similar studies. Few patients would accept up to a 9·9% extra risk of death for 3 days fewer antibiotics. Why should your readers?
Science Photo Library
Procalcitonin in intensive care units: the PRORATA trial
We declare that we have no conflicts of interest.
*William Tarnow-Mordi, Val Gebski
[email protected] Westmead International Network for Neonatal Education and Research (WINNER) Centre, Centre for Newborn Care, University of Sydney, Westmead Hospital, Sydney, 2125 NSW, Australia (WT-M, VG); and NHMRC Clinical Trials Centre, University of Sydney, Camperdown, NSW, Australia (VG) 1
Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patients’ exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial. Lancet 2010; 375: 463–74.
Lila Bouadma and colleagues1 nicely showed that a strategy based on the daily measurement of procalcitonin concentration allowed for a reduction of antibiotic consumption in intensive care units (ICUs). Although this held true for most infections, the PRORATA trial mainly included patients with pneumonia and showed that measuring procalcitonin concentrations allowed the clinician to better adhere to the international guidelines on
Submissions should be made via our electronic submission system at http://ees.elsevier.com/ thelancet/
1605