Early Death From Pediatric Severe Sepsis: Still a Problem and a Mandate for Future Clinical Research*
In this issue of Pediatric Critical Care Medicine, Weiss et al (5) report findings of their well-done study of the timing, etiology, and mechanism of in-hospital death of 79 patients with severe sepsis at two institutions. They hypothesized that in the current era of sepsis screening and organ support, early death due to refractory shock would be rare and that most deaths would be due to persistent multiple organ dysfunction and subsequent limitation of life-sustaining therapies later in the course of illness, consistent with data from adults (6). Contrary to their hypothesis, however, Weiss et al (5) found that refractory shock was in fact the most common cause of death among pediatric patients and that 25% of deaths occurred very early, within 1 day of sepsis recognition.
Particularly notable about these findings is that the patients who died soon after sepsis recognition seem to be fundamentally different than those who died later. Those who died very early were predominately previously healthy; 63% of previously healthy children died on the first day and 75% died within the first 3 days. Despite representing only 20% of all sepsis deaths, previously healthy children accounted for half of deaths on the first day. In contrast, nearly all deaths after 7 days occurred in children with comorbidities. The very early deaths were also much more likely to be in children with community-acquired sepsis and be recognized in the emergency department rather than the inpatient setting. Not surprisingly, most early deaths were due to refractory shock, and 60% occurred after unsuccessful cardiopulmonary resuscitation. Most of the deaths occurring later were due to withdrawal of life-sustaining therapies.
The findings from this study come at a time when the proportion of previously healthy children in the PICU, and their corresponding mortality, is declining steadily (1, 4, 7). Over three quarters of children with severe sepsis now have underlying comorbidities, with a mortality rate 50% higher than previously healthy children (4). The current findings by Weiss et al (5) suggest that despite these trends, previously healthy children still comprise a substantial portion of the most severely ill sepsis patients. This incongruity exposes how critical it is that we find a way to include children who die soon after presentation to the hospital in rigorous clinical research, something that to date has proven extraordinarily challenging.
Previous pediatric investigations demonstrating high rates of early sepsis death (8, 9) largely reflected the mortality pattern of fulminant meningococcemia, a clinical entity that has become much less common since these studies were performed. Over 50% of deaths in children with severe sepsis in the United Kingdom occurred prior to or within 24 hours of PICU admission; many of the early deaths were due to meningococcemia (8). Among patients screened for enrollment in the recombinant bactericidal/permeability-increasing protein drug trial, 63% of deaths occurred before the drug could even be given (9, 10). High early mortality in these studies undoubtedly resulted in missed opportunities for clinical trial enrollment.