DOI: 10.1097/01.PCC.0000298648.56417.67

,  

,  

PMID: 18185123

Issn Print: 1529-7535

Publication Date: 2008/01/01


Print

Novel trial designs for pediatric traumatic brain injury*

Susan L. Bratton; Brahm Goldstein

    loading  Checking for direct PDF access through Ovid

Excerpt

It is clear that the pediatric community still has a long road ahead to improve outcomes for children who have suffered severe traumatic brain injury (TBI). Despite increased clinical experience with TBI, the mortality rate has not improved over the past 20–30 yrs (1). In a recent publication of evidence-based treatment guidelines, Adelson et al. (2) discussed the acute medical and surgical management of infants, children, and adolescents with severe TBI defined as a Glasgow Coma Scale score ≤8. The topics covered included the following: prehospital management—trauma systems, airway management, resuscitation of blood pressure, and oxygenation; medical management—indications for intracranial pressure (ICP) monitoring threshold, ICP technology, cerebral perfusion pressure, use of sedation and neuromuscular blockade, cerebrospinal fluid drainage, use of hyperosmolar therapy, use of hyperventilation, use of barbiturates, the role of temperature control, use of corticosteroids, nutritional support, role of antiseizure prophylaxis; surgical treatment; development and suggestion for a critical pathway for the treatment of established intracranial hypertension (2).
One of the most striking take-home messages from this exhaustive review was that the current state of severe TBI treatment has many more unanswered questions than evidence-based recommendations. Although some therapies have been used for decades, we still do not have answers to many questions: What is the efficacy of specific, ICP-directed therapies on long-term, age-appropriate, neurologic outcome in infants and children? What are specific threshold values of ICP for institution of therapy in different pediatric age groups and different injury etiologies? What is the relative importance of ICP- and cerebral perfusion pressure-targeted therapies?
In this issue of Pediatric Critical Care Medicine, Dr. Forsyth and colleagues (3) report data from a national database of 501 children <16 yrs old admitted to a pediatric intensive care unit (PICU) for management of severe TBI. The main goal of the study was to develop stratification criteria for future randomized controlled trials for treatment of raised ICP and to estimate recruitment rates and outcomes to inform power calculations. The authors developed decision tree (recursive partitioning) and logistic regression modeling of death before PICU discharge as well as development of raised ICP through analysis of the data set (3). However, they built a logistic model using all factors significantly related to outcome in their univariate model, rather than by building a stepwise model. Given the sample size limitations, this resulted in a model in which only diffuse axonal injury was significantly associated with intracranial hypertension, given the mathematical effects of the 13 other factors. Clearly, the model could be simplified, as the desire was to develop an easy scheme to predict need to monitor.
Dr. Forsyth and colleagues (3) reported that their decision rule predicted development of raised ICP with a sensitivity of 73% and a specificity of 74% (positive predictive value 82% and negative predictive value 63%) (3). It was no surprise that the authors found death was associated both with raised ICP and the nonmeasurement of ICP (e.g., many likely premorbid and terminal cases) (3). The conclusion was that this model, which is based on early clinical data, can predict the development of raised ICP with sufficient accuracy to form the basis of stratification or entry criteria for future randomized controlled trials (3). But of potentially greater impact, the authors estimated that studies designed to detect reductions in ICU mortality by 15% will require >320 children per study arm (3).
We believe that the most important contribution of this study is to call attention to the significant problems involved in studying this patient population if we rely on standard research study designs, such as randomized controlled trials.

Related Topics

    loading  Loading Related Articles