Dr. Jagannathan and colleagues provide a detailed review of the management of severe pediatric head injury in 96 patients in whom intracranial pressure (ICP) was elevated when first measured and who were followed-up for a minimum of 2 years in almost every case. The present paper follows a previous study from the same investigative group on decompressive craniectomy1 but adds considerable new information. This new study appropriately excludes patients with nonaccidental trauma and includes only those patients in whom the initial CT scan result was abnormal. These investigators used a definition of ICP as . 20 mm Hg as their cutoff point for elevated ICP. Although it is a retrospective review, this study is one of the largest patient series of pediatric head injury specifically focusing on intracranial hypertension and its influence on outcome, and is particularly important given the long-term follow-up aspect of the study. Using a standardized protocol with a bias toward using early decompressive craniectomy, ICP could be controlled in 82 of the patients (85%) in this study. Medical therapy was effective in 35% of the patients, which was defined as using sedation, intermittent hyperventilation, and muscle relaxation. When ventriculostomy was used, 24% of the patients also attained ICP control. Fourteen patients underwent either an attempt or successful evacuation of a mass lesion; ICP could still not be controlled in 2 patients after this procedure, and both patients died. Decompressive craniectomy was performed in 25 patients and ICP was controlled in 16 of these patients (64%). In the 9 cases in which ICP control could not be achieved after decompressive craniectomy, all the patients died. This paper provides a number of very important observations, but also raises some questions. First, it appears that the means to ICP control is less important than the method in which ICP can be brought below 20 mm Hg. The authors’ bias toward using early decompressive craniectomy appears, in part, to reflect dissatisfaction with the rate of infection following placement of a ventricular drain. Clearly, decompressive craniectomy is a much more invasive procedure, and although the infection rate after ventriculostomy in this study (~ 8%) is consistent with rates in other studies, it is possible that changes in the technique of ventricular drain placement might reduce the frequency of infection. The recent focus on reducing catheter-related sepsis has resulted in dramatic reductions in such infections and perhaps a much more vigorous approach to sterility when a ventricular catheter is placed might yield similar benefits. Nevertheless, the authors should not be criticized for attempting decompressive craniectomy at a relative early time period because it is fairly clear that a late decompressive craniectomy, when all other measures have failed, is likely to result in very unsatisfactory outcomes. One particularly attractive feature of the study is the long-term follow-up aspect. As the authors note in the paper, this feature is probably particularly important in the pediatric population, in which substantial plasticity and microcellular reshaping of the brain is occurring at a time in these patients when the brain is severely traumatized. The brain is literally “pruning itself” in its day-to-day evolution to optimize neuronal connections, and therefore improved outcomes in children and young adults are not terribly surprising. Nevertheless, permitting adequate cerebral perfusion by reducing ICP as Juul and colleagues2 have demonstrated appears to be the appropriate approach in a critical care environment. Of interest is the observation that vascular injuries were present in ~ 4% of the patients. This indicates that these impairments of cerebral perfusion may be more common than previously believed, and given modern interventional techniques these injuries might be amenable to treatment if they could be recognized early. It would be of interest to know the authors’ view on this subject. It would also be helpful for the authors to clarify what they did in those patients who underwent operative intervention for hematoma without a decompressive craniectomy, who developed significant brain swelling in the operating room following evacuation of the mass lesion. Under those circumstances, many neurosurgeons would have already turned a large flap or would enlarge the intracranial opening to provide a large decompression. In Table 6 one notes that in patients with diffuse injury
Read full abstract