Microsoft word - 3cee3ff2-085d-0c9c.doc

Pem-Database.Org
Editorial Comment on:
“Efficacy of oral dexamethasone in outpatients with acute bronchiolitis”
S Schuh, AL Coates, R Binnie, et al. J Pediatr 2002;140:27-32
David W. Johnson,
Associate Professor,
Departments of Pediatrics and Pharmacology & Therapeutics,
University of Calgary, CANADA
February 2002

Numerous clinical trials – somewhere in the neighborhood of 16 - have been published
which examine whether corticosteroids are effective in the treatment of bronchiolitis. Until recently, the results of these trials - when examined individually - pointed strongly towards a lack of benefit. However, the results of a recent meta-analysis of these trials (Pediatrics 2000;105:e44) fly in the face of conventional wisdom, indicating that corticosteroids reduce the duration of hospitalization by a modest amount – on average Now Schuh et al. - in line with the findings of this meta-analysis - report that corticosteroids reduce the rate of hospitalization in children with bronchiolitis by approximately one half. In this trial, children less than two years of age with their first episode of wheeze (approximately 50% were positive for RSV) who presented to an emergency department with moderately severe respiratory distress were enrolled, randomized, and treated orally with 1 mg/kg of dexamethasone or placebo. Treatment with nebulized albuterol was standardized. Study investigators assessed clinical severity hourly after treatment for four hours using a standardized and validated clinical score (the Respiratory Assessment Change Score). Staff physicians, masked to treatment group and assessment of clinical severity by study investigators, determined - based on their own clinical judgement - whether or not the child was admitted or discharged home. These investigators enrolled a total of 70 patients. Children treated with placebo improved less (the mean clinical score improved by 3.2±3.7 points versus 5.0±3.1 points, p = 0.29), and were significantly more likely to be hospitalized (44% (15/34) versus 44% (7/36), p = 0.039) than those children treated with dexamethasone. None of the 48 patients discharged home were subsequently hospitalized. Schuh’s results suggest significantly greater benefit for the use of corticosteroids than any other trial reported to date. Why is that? We should start by asking in what ways Schuh’s study design differs from previous studies. First, her study – to my knowledge – is the first to have focused on change clinical outcomes in the emergency department, and its impact on hospitalization rates. Virtually all studies have focused on outcomes over the course of several days, and, in some cases, up to months later. Second, the dose of corticosteroids administered was substantially larger (1 mg/kg of dexamethasone, the equivalent of approximately 7 mg/kg of prednisone) than most studies administered. And last, they enrolled patients fairly early in the course of their disease (on average slightly Do the differences in study design outlined above make a logical case for why Schuh found such a striking difference, whereas previous investigators did not? Its possible but, in my opinion, not likely. Our experience in the treatment of asthma and croup suggests that demonstrable clinical benefit occurs as early as three to five hours after treatment. But it seems counter-intuitive that if corticosteroids yield such a striking benefit within a matter of hours, that it would attenuate so sharply over the following days. Regarding the dose of corticosteroid used, Roosevelt et al (Lancet 1996;348:292) also used a dose of 1 mg/kg/day of dexamethasone, but found no benefit. Furthermore, several studies in adults and children with asthma have not shown supratherapeutic dosing of corticosteroids to be more effective than doses in the range of 1 to 2 mg/kg/day prednsione (0.15-0.30 mg/kg/day dexamethasone). And last, though almost all published studies have focused on hospitalized children, as opposed to those evaluated in an emergency department, several studies have enrolled children at a comparable stage in their disease. For example, Klassen et al. enrolled patients who had had symptoms, on average, for slightly less than 72 hours (J Pediatr 1997:130:191), in contrast to Schuh whose patients had been symptomatic for slightly less than 48 hours. Schuh’s study was well-designed and reported, but – because it involved only a single center – was relatively small. Given the striking difference between Schuh’s results, and the previous literature, I think it is critical that a large multicenter randomized controlled trial be completed with a similar study design similar before we accept that steroids are

Source: http://www.pemdatabase.org/files/Editorial_Comment__David_Johnson.pdf