Background

Depression is common after traumatic brain injury (TBI), with at least 1 in 5 patients meeting criteria for a Major Depressive Episode within the first six months [1,2,3]. This rate is similar across the spectrum of TBI severity. Depression may magnify the burden of physical and cognitive symptoms as well as functional disability after TBI [2, 4], making it an important treatment target.

Main text

Prior meta-analyses have concluded that antidepressant medications are effective for depression in a variety of neurological disorders [5], including TBI [6]. Correspondingly, selective serotonin reuptake inhibitors are recommended as first-line treatment for depression in recently published expert consensus clinical practice guidelines for management of TBI [7, 8]. However, a recently published systematic review [9] raised doubt about this evidence base. Kreitzer et al. [9] screened 1020 articles published before September 20, 2017 and found 11 eligible pharmacological intervention studies in TBI samples. Their meta-analysis of the five placebo-controlled trials revealed “no benefit of antidepressant over placebo in the treatment of [Major Depressive Disorder] following TBI” (standardized mean difference = − 0.3; 95% confidence interval = − 0.6 to 0.0; I [2] = 17%) [9]. This conclusion might lead some clinicians to not offer antidepressant therapy to their patients who present with depression after TBI, which in our view, would be unfortunate. Even if this meta-analytic finding is “real” (not a Type II error), we argue here that proactive treatment should be considered, especially in mild TBI (concussion), for the following reasons.

First, when non-randomized and open-label studies were included in the Kreitzer et al. [9] study as well as prior meta-analyses on the same topic [6, 10], the treatment effect was significant. That is, patients who received antidepressant therapy got better. These gains may be attributable to placebo effects. Even so, placebos may be a powerfully effective treatment for various TBI-related problems [11]. For a patient who presents with depression after TBI, prescribers may have an opportunity to harness placebo effects with on-label use of an active medication.

Second, patients whose depression does not respond to an initial antidepressant trial often benefit from augmentation or switching to an alternative selective serotonin reuptake inhibitor or a selective serotonin-epinephrine reuptake inhibitor. This stepwise approach is recommended in guidelines for TBI management [7]. Placebo-controlled trials such as those synthesized in Kreitzer et al. [9] measure the change in depressive symptoms on the first attempted antidepressant agent in the “average” patient, not the potential for depression remission with stepwise medication trials, as in real-world clinical practice.

Third, traditional depression outcome measures may lose responsiveness when applied in TBI studies, resulting in underestimation of treatment benefit. For example, 3 out of the 5 placebo-controlled trials pooled by Kreitzer et al. [9] measured depressive symptoms with the Hamilton Depression Rating Scale (HAM-D). The authors acknowledge that "instead of demonstrating antidepressants do not work over time, it could be possible that this outcome measure may not change sufficiently with time due to factors related to the underlying TBI [9]. The HAM-D, like most depression measures, includes non-specific symptoms that could reflect structural brain injury. There is external evidence from non-pharmacological treatment trials that the HAM-D and its subscales are suboptimally responsive to improvements in depression after TBI [12].

The rationale for using antidepressants after mild TBI is most compelling. Patients with mild TBI patients may more closely resemble patients with no known TBI history enrolled in typical primary Major Depressive Disorder clinical trials than patients enrolled in TBI trials in placebo-controlled trials published to date. Consider for example the “only study to influence the results of the RCT meta-analysis” in Kreitzer et al’s sensitivity analyses (that is, the only study that, if excluded, would have resulted in a statistically significant meta-analytic finding favoring antidepressants): the recent trial by Fann et al. [13]. These authors only enrolled patients with mild TBI if they had radiological evidence of brain injury and were admitted to a Level 1 trauma center. Many patients with mild TBI do not present to an acute care hospital. Of those who do, a minority will have a trauma-related intracranial abnormality on computed tomography and/or be admitted to hospital. Therefore, the mild TBI patients in Fann et al’s study are not representative of the mild TBI population, but rather reflect a subset at the “severe” end of mild TBI spectrum.

Conclusions

We argue that TBI, and especially mild TBI, is not a contraindication for antidepressant therapy. Proactive detection and treatment of psychiatric problems following TBI of any severity has the potential to improve not only mental health outcomes [14, 15] but also cognition [16,17,18], somatic symptoms [19, 20], and daily functioning [2, 4, 21]. We recognize that a systematic review of randomized placebo-controlled trials is the highest level of evidence, but caution clinicians against changing their practice on the basis of the Kreitzer et al. [9] study. Clinicians who continue to prescribe will be compliant with practice guidelines [7, 8] and the best available evidence [6, 9, 10]. Selective serotonin reuptake inhibitors (e.g., sertraline or citalopram) are generally recommended as first-line [7, 18, 22, 23]. Efficacy and tolerability of selective serotonin-epinephrine reuptake inhibitors are less well-established [23], but expert consensus guidelines endorse their use [7]. Tricyclic antidepressants are thought to have a less favorable benefit-risk profile, with lowering of the seizure threshold in patients with moderate-severe TBI being a potential concern [8, 23, 24].

Psychotherapy should be offered as an alternative or adjunctive treatment, where accessible [7]. The majority of patients with TBI prefer psychotherapy to pharmacotherapy treatment options [25]. There is insufficient evidence to recommend one modality of psychotherapy over another, but cognitive-behavioural therapies have been most well-studied [23, 26]. Remote delivery, such as by telephone, appears feasible and as effective as in-person therapy [27]. The minimum effective “dose” of psychotherapy is unknown.

Routine screening by rehabilitation and primary care providers can facilitate timely detection of depression after TBI. For example, the Personal Health Questionnaire-9 is a brief self-report depression case-finding tool with strong psychometric properties [12] and diagnostic accuracy in TBI [28]. A positive screening result should trigger a diagnostic clinical assessment with a qualified provider [7, 8].