Meta-Analytic Methods and the Importance of Non-TBI Factors Related to Outcome in Mild Traumatic Brain Injury: Response to Bigler et al. (2013)

Abstract

Bigler et al. (2013, The Clinical Neuropsychologist) contend that weak methodology and poor quality of the studies comprising our recent meta-analysis led us to miss detecting a subgroup of mild traumatic brain injury (mTBI) characterized by persisting symptomatic complaint and positive biomarkers for neurological damage. Our computation of non-significant Q, tau² , and I² statistics contradicts the existence of a subgroup of mTBI with poor outcome, or variation in effect size as a function of quality of research design. Consistent with this conclusion, the largest single contributor to our meta-analysis, Dikmen, Machamer, Winn, and Temkin (1995, Neuropsychology, 9, 80) yielded an effect size, –0.02, that was smaller than our overall effect size of –0.07 despite using the most liberal definition of mTBI: loss of consciousness less than 1 hour, with no exclusion of subjects who had positive CT scans. The evidence is weak for biomarkers of mTBI, such as diffusion tensor imaging and for demonstrable neuropathology in uncomplicated mTBI. Postconcussive symptoms, and reduced neuropsychological test scores are not specific to mTBI but can result from pre-existing psychosocial and psychiatric problems, expectancy effects and diagnosis threat. Moreover, neuropsychological impairment is seen in a variety of primary psychiatric disorders, which themselves are predictive of persistent complaints following mTBI. We urge use of prospective studies with orthopedic trauma controls in future investigations of mTBI to control for these confounding factors.

Keywords:

• Mild traumatic brain injury
• Meta-analysis
• Non brain injury factors

Acknowlodgement

Drs. Larrabee, Binder and Rohling provide forensic consultation related to traumatic brain injury. No other potential for conflict of interest is present.

Notes

1. The Heaton et al. (1991) normative data were used to score the Impairment Index since normative data are not reported for this measure in the Heaton, Miller, Taylor, and Grant (2004) publication.