Thomas Jefferson, Steroids, and Sepsis*
As it turns out, it may be just as hard to abandon routine clinical practices in the face of clinical trials evidence suggesting that they are no longer beneficial, when the evidence is less than convincing or when the evidence just does not exist (4). As a case in point, the use of corticosteroids in critically ill patients with septic shock has been hotly debated since the 1950s (i.e., three generations, as defined by Jefferson above). In fact, one of the first clinical trials of corticosteroids performed in adults and children with sepsis in 1963 showed no benefit (5). Regardless, the most recent 2016 Surviving Sepsis Campaign consensus guidelines for the management of critically ill patients with septic shock continue to recommend corticosteroids, though only for patients who are hemodynamically unstable despite adequate fluid resuscitation and vasopressor therapy (6). The evidence for this particular recommendation is admittedly weak and of low quality, based upon a number of systematic reviews and meta-analyses that suggest that even if corticosteroids reverse shock in critically ill adults with septic shock, outcomes are not significantly improved (7).
There are no large multicenter, randomized, controlled clinical studies on the administration of corticosteroids in pediatric septic shock (8). However, there are still a number of clinicians who continue to administer corticosteroids in the setting of pediatric septic shock (9, 10), which is surprising given the adult studies mentioned above combined with a number of observational studies in children showing either no benefit (11–14) or even harm (15–17). In this issue of Pediatric Critical Care Medicine, Nichols et al (18) report the findings of a pragmatic (i.e., “real world” use) analysis of corticosteroids in pediatric septic shock. Two points are noteworthy. First, the investigators followed the 2012 Surviving Sepsis Campaign pediatric guidelines that recommended obtaining a random serum total cortisol (rSTC) level and then initiating stress-dose hydrocortisone for those children with fluid-refractory and catecholamine-dependent shock and rSTC level below 18 µg/dL (19). The most recent guidelines have abandoned the use of rSTC as criteria for initiating corticosteroids in this setting (6), so one could certainly argue that the study design may not have reflected “real world” clinical practice.