Diagnostic accuracy studies in mild traumatic brain injury: a systematic review and descriptive analysis of published
Theresa L-B Pape1, Walter M High, Justin St Andre
1Center for Innovation in Complex Chronic Healthcare & Research, The Department of Veterans Affairs (VA), Edward Hines Jr. VA Hospital, PO Box 5000 (M/C 151H), Hines, IL 60141; Department of Physical Medicine and Rehabilitation, Northwestern University Feinberg School of Medicine, Chicago, IL∗.
Objectives:
To synthesize evidence and report findings from a systematic search and descriptive analysis of peer-reviewed published evidence of the accuracy of tests used for diagnosing mild traumatic brain injury (mTBI). The article also summarizes points of concurrence and divergence regarding case definitions of mTBI identified during the review. TYPE: Systematic review and descriptive analysis of published evidence.
Literature Survey:
A search of PubMed, PsychInfo, and the Cochrane Library for peer-reviewed publications between 1990 and July 6, 2011, identified 1218 abstracts; 277 articles were identified for full review, and 13 articles met the criteria for evaluation.
Methodology:
Manuscript inclusion criteria were (1) reported sensitivity (Se) and specificity (Sp), or reported data were sufficient to compute Se and Sp; (2) >1 participant in the study; (3) at least 80% of the study cohort was ≥18 years of age; and (4) written in English. Articles describing clinical practice guidelines, opinions, theories, or clinical protocols were excluded. Seven investigators independently evaluated each article according to the Standards for Reporting of Diagnostic Accuracy (STARD) criteria.
Synthesis:
Findings indicate that all 13 studies involved civilian noncombat populations. In 7 studies, authors examined acute mTBI, and in 4 studies, historical remote mTBI was examined. In the 13 studies, Se ranged from 13%-92% and Sp ranged from 72%-99%, but confidence in these findings is problematic because the STARD review indicates opportunities for bias in each study.
Conclusions:
Findings indicate that no well-defined definition or clinical diagnostic criteria exist for mTBI and that diagnostic accuracy is currently insufficient for discriminating between mTBI and co-occurring mental health conditions for acute and historic mTBI. Findings highlight the need for research examining the diagnostic accuracy for acute and historic mTBI.
