Page 32 - DECISION-MAKING IN SEVERE TRAUMATIC BRAIN INJURY PATIENT OUTCOME, HOSPITAL COSTS, AND RESEARCH PRACTICE
P. 32

Chapter 2
EVIDENCE ACQUISITION
The literature review was conducted according to a predefined search protocol. A systematic review attempt was abandoned as randomized studies and methodological sound prospective studies were lacking. Keywords were “brain injury”, “traumatic”, “surgery”, “neurosurgical procedures”, “operative” and “severe” (Appendix I). The sections discussing penetrating brain injury (PBI) are separately informed by the literature search used for the Guidelines for the Management of Penetrating Brain Injury,14 which was expanded by an additional literature search in Medline. Search terms included “penetrating head or brain injuries”, “brain”, “head”, “wounds” and “gunshot” (Appendix I).
Two reviewers independently selected relevant studies, extracted data and discussed disagreements until consensus was reached. If consensus was not reached one of the senior authors was capable to take the final decision.
Two stages of study selection were needed (Figure 1). First, studies were selected on title and abstract at least containing: (1) s-TBI patients, (2) (neuro) surgical treatment and (3) clinical outcome. Secondly, during full-text screening, only original data studies with patient cohorts (N>10) consisting of vs-TBI patients (early GCS Score 3-5) were included if data on (neuro) surgical treatment and outcome were presented. Studies were excluded when published before 2000 and non-English. Authors excluded series without a detailed initial GCS and only mentioning mean or median scores for obvious clinimetric reasons.
Manuscripts containing information on outcome in vs-TBI in adult populations were subsequently divided based on surgical treatment; ICP monitoring, decompressive craniectomy and other surgical interventions. Studies discussing elderly and pediatric patients were discussed separately. Authors used various synonyms for good or favorable outcome (GOS 4 or 5), representing “moderate disability” and “good outcome” respectively. The same classification and denomination was used in the specific references.
30




























































































   30   31   32   33   34