Flaws in the Defense Case that mTBI has no Lasting Effect: The Problem with Meta-analysis
Defense attorneys often cite “meta-analytic” reviews of neurological studies to make the argument that “mild” traumatic brain injuries (mTBIs) cause no lasting effect beyond three months post-injury. A “meta-analysis” involves a statistical study of multiple studies published in the literature.
Meta-analyses in mTBI are often used to show that persistent symptoms are “neurotic” rather than “organic”
Fortunately, it is generally agreed that the majority of people who suffer mTBIs, sometimes referred to as “concussions”, report full recovery from symptoms within three months of the injury – in fact many recover much faster. A great deal of research over the past few years has focused on the minority of people who do not fully recover within three months, described as having a “persistent post-concussion symptoms (PCS).” These patients are sometimes referred to as the “miserable minority.” The “meta-analyses” are often cited as demonstrating that changes in performance after three months have “limited statistical and clinical significance;” in other words, that persistent symptoms must be psychological or “neurotic” rather than “organic” or neurologic.
Recent research demonstrates the flaws with meta-analyses in mTBI studies
A paper published this year by Bigler et. al. in The Clinical Neuropsychologist cogently demonstrates the flaws in the meta-analytic mTBI studies commonly relied upon by the defense. One of the flaws identified is that the meta-analytic studies lump all mTBIs together as though they are homogeneous, or the same, when in fact mTBIs are heterogeneous, not homogeneous. Impaired performance by a few is not detected because it is averaged out in the analysis. For example, mTBIs considered “complicated” because they are accompanied by lesions identified by conventional MRI are lumped in with other mTBIs, even though it is generally agreed that focal demonstrated damage detected in MRIs increases the risk of prolonged symptoms consistent with those commonly seen in moderately severe TBI. [Silver, McAllister and Yudofsky, Textbook of Traumatic Brain Injury, p. 281.] In fact, even the most cited meta-analytic study by Rohling et. al. acknowledges (at pp. 608 and 618) that their conclusions would not necessarily apply to those with complicated mTBIs.
Another flaw in the meta-analyses is that we know from “histopathological” studies (studies of the brain tissue of people who have suffered mTBIs) that chronic lesions can be associated with mTBIs. We also know this from over 100 neuroimaging studies of mTBI using advanced neuroimaging techniques that demonstrate abnormalities at the microstructural level following mTBIs. One obvious question is if a “concussion” has no lasting effect, why is it that multiple concussions can have the profoundly severe consequences we see in athletes (whose brain show severe damage when examined on autopsy.) Are soldiers returning from middle east conflicts with the lasting effects of concussions all malingering? Bigler also cites the “volumetric” studies, discussed in our previous blog post, showing that cerebral atrophy is associated with “persistent post-concussive symptoms.”
Neuropsychological assessment faces its biggest challenge in detecting subtle impairment
Another flaw in the analysis identified by Bigler is accumulating evidence, cited in his study, that regardless of the type of neurological and/or neuropsychiatric disorder, neuropsychological assessment, often used to assess persistent symptoms, “faces its biggest challenge in detecting subtle impairment.” Although these assessments can effectively detect cognitive impairment in the acute recovery stage, they are less effective – the tests are less sensitive – after the acute recovery phase.
Bigler cites a study by B. Johnson et al (2012) which uses functional MRI (fMRI) to show persistent deficits in the brain’s default network in athletes with mTBI. Johnson reports that “Neuropsychological testing and conventional neuroimaging techniques are not sufficiently sensitive to detect neurological changes.” Bigler recommends “the integration of functional neuroimaging with cognitive processing speed tasks that measure performance in milliseconds, a more direct representation of processing speed.” A delay in processing speed may make no difference in outward neuropsychological test performance, in the sterile test environment, but it does represent a genuine delay in neural processing, causing a material change in baseline capabilities.
Needless to say, there is a great deal of incentive in the insurance industry to hang on to the conclusion suggested in meta-analysis that persistent post-concussive symptoms are neurotic and not organic. The scientific evidence, however, has outdated and disproved these conclusions.