Systematic search and review procedures: results of the International Collaboration on Mild Traumatic Brain Injury Prognosis.Arch Phys Med Rehabil. 2014 Mar; 95(3 Suppl):S101-31.AP
To update the last best-evidence synthesis conducted by the World Health Organization Collaborating Centre for Neurotrauma, Prevention, Management and Rehabilitation in 2002; and to describe the course, identify prognostic factors, determine long-term sequelae, identify effects of interventions for mild traumatic brain injury (MTBI), identify knowledge gaps in the literature, and make recommendations for future research.
MEDLINE, Embase, PsycINFO, Cumulative Index to Nursing and Allied Health, and SPORTDiscus were searched between 2001 and 2012. Inclusion criteria included published peer-reviewed articles in English and 5 other languages. References were also identified from relevant reviews and meta-analyses and the bibliographies of eligible articles.
Controlled trials and cohort and case-control studies were selected according to predefined inclusion/exclusion criteria. Studies had to have at least 30 MTBI cases and assess outcomes relevant to prognosis after MTBI.
Eligible studies were critically appraised using modified Scottish Intercollegiate Guidelines Network (SIGN) criteria. Two reviewers independently reviewed each study and extracted data from accepted articles (ie, with a low risk of bias) into evidence tables.
The evidence was synthesized qualitatively according to modified SIGN criteria and prioritized according to design as exploratory or confirmatory. The evidence was organized into separate articles according to population (eg, adults, children, and athletes) and outcomes (eg, risk of dementia after MTBI).
After 77,914 records were screened, 299 articles were eligible and reviewed. Of these, 101 (34%) were accepted as scientifically admissible and form the basis of our findings, which are organized into 10 articles in this supplement. These reviews present the best available evidence on MTBI prognosis, but more research is needed.