NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Guillamondegui OD, Montgomery SA, Phibbs FT, et al. Traumatic Brain Injury and Depression [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2011 Apr. (Comparative Effectiveness Reviews, No. 25.)
This publication is provided for historical reference only and the information may be out of date.
This chapter summarizes the strength of the evidence to address our key questions (KQs) and presents methodologic considerations and a discussion of the findings for each of our six key questions. We conclude with a discussion of the status of research, limitations of the current literature, and our recommendations for future research priorities.
Strength of the Evidence
Strength of evidence is typically assigned to reviews of medical treatments after assessing four domains: risk of bias, consistency, directness, and precision.29 Although these categories were developed for assessing the strength of treatment studies, the domains also apply to studies of prevalence and screening. Available evidence for each key question was assessed for each of these four domains; the domains were combined qualitatively to develop the strength of evidence for each key question.
We graded the body of literature for each key question and present those ratings as part of the discussion in Chapter 4. The possible grades were:
- High: High confidence that the evidence reflects the true effect. Further research is unlikely to change estimates.
- Moderate: Moderate confidence that the evidence reflects the true effect. Further research may change our confidence in the estimate of effect and may change the estimate.
- Low: Low confidence that the evidence reflects the true effect. Further research is likely to change confidence in the estimate of effect and is also likely to change the estimate.
- Insufficient: Evidence is either unavailable or does not permit a conclusion.
As a global assessment of this literature, the evidence about prevalence and elevated risk of depression is moderate.
The strength of the evidence to inform when, with what tools, and in what settings screening for depression should occur after traumatic brain injury (TBI) is low. This assessment is based on lack of studies designed to specifically address these concerns that make explicit comparisons. The risk for bias is high since the data used was often cross-sectional, full populations were rarely delimited, and settings and populations may represent a biased portion of the spectrum of both the severity of TBI and of depression. As for prevalence, data are inconsistent across studies. Summary evidence is to a degree indirect. Though we included only studies for which the timing of administration of the screening or diagnostic tool was known and for which the tool was described, the studies overall were not designed for this use, and estimates are imprecise.
Likewise, understanding of psychiatric comorbidities is based on a small number of studies that do not provide consistency in conditions measured and are at risk of bias, resulting in low strength of evidence. Intervention studies included one randomized controlled trial (RCT) of good quality and one case series; therefore, risk of bias was high, consistency was unknown, and results lacked precision. The strength of evidence is insufficient for pharmacologic treatment of depression after TBI.
Principal Findings and Considerations
KQ1. Prevalence and Incidence of TBI and Depression
We identified 112 publications30,13–15,31–138 from 79 distinct study populations in which prevalence could be estimated.
We considered the Structured Clinical Interview for DSM-IV (SCID) and other formal structured clinical interview protocols that map to the DSM and/or International Classification of Diseases (ICD) codes to be the measures of depression that are most relevant to clinical care. Among studies that used an SCID or other structured protocol to reach a formal diagnosis of depression, the prevalence of depression after TBI ranged from 12.2 percent66 to 76.7 percent.13 If we focus on the subset of studies with both clearly operationalized criteria for TBI and use of the SCID, the range was 12.2 percent66 to 54.0 percent.109
Data are sparse to assess whether severity of injury influences risk of depression. Using structured interviews among those with mild or mild/moderate TBI populations, the overall prevalence of depression was 20.3 percent compared to 32.5 percent in studies that enrolled or followed up populations of all severities. Too few studies isolate a sufficient number of those with mild TBI compared to those with moderate and/or severe injuries to make valid severity-based estimates. Some multivariate models suggest severity of injury is not a strong predictor of risk.131,134 Likewise, stratification of prevalence by explanatory factors such as age, gender, area of brain injured, or mechanism of injury is not possible within the current body of literature.
No strong predictors of risk are available to tailor guidance or screening of those with TBI.
KQ2. Screening for Depression After TBI
We identified 11330,13–15,31–137,139 publications in 79 distinct populations that provide information about timing of screening or comparison of tools.
Across all timeframes and using all depression measures, in studies with clear TBI definitions, the weighted average for prevalence of depression was 31.8 percent. Among those studies with repeated assessments and/or longer term followup, no clear pattern of expected natural history or peak prevalence emerged. Depression was more common among those with TBI than among normal comparison groups.
Five publications compared SCID to candidate tools for assessment of depression, the Beck Depression Inventory (BDI), Patient Health Questionnaire-9 (PHQ-9), and Hospital Anxiety and Depression Scale (HADS).41,103,128,130,139 None of the tools reported simultaneous sensitivity and specificity above 90 percent. One study identified different optimal cutoffs of the BDI-II; maximum sensitivity of 87 percent and specificity of 79 percent were obtained with cutoffs of 19 for participants with mild TBI and 35 for those with moderate or severe TBI. With modification of the scoring algorithm as proposed by the authors, the PHQ-9 achieved a sensitivity of 93 percent, specificity of 89 percent, positive predictive value of 63 percent, and negative predictive value of 99 percent. The BDI had poor sensitivity of 48 percent and 32 percent at specificities of 80 and 90 percent, respectively. The HADS provided 54 percent sensitivity and 76 percent specificity.
Therefore, no evidence provides a basis for targeting screening to one timeframe over another. Likewise, the literature is insufficient to determine whether tools validated in other populations for detecting depression appropriately identify individuals with depression after a TBI, or to choose among available tools.
KQ3. Prevalence of Concomitant Psychiatric Conditions
In general, the few papers that do specifically examine the prevalence of concomitant psychiatric conditions within the population of depressed TBI patients report high rates of such conditions within this population. When conditions were reported individually, anxiety disorders were most prevalent and affected 31 percent to 61 percent of study participants in four papers.14,72,109,138 Post-traumatic stress disorder (PTSD), a major anxiety disorder, was observed in 37 percent of depressed patients and in no patients without depression,123 and panic disorder was seen in 15 percent of patients with major depression but not measured in those without depression.109 Consideration of potential for coexisting psychiatric conditions is warranted.
KQ4. Outcomes of Treatment for Depression After TBI
Only two publications141–142 addressed a treatment for individuals diagnosed with depression after TBI. One of the treatment studies was conducted in the United States,141 and the second was in Canada.142 Both were studies of medication, the first being an RCT of sertraline; the second, an open-label case series of the effects of citalopram. Of those that completed the sertraline study, 59 percent of the treated group and 32 percent of the control group had a positive response; the difference in response rates between the two groups was not statistically significant (p = 0.08). In the 6-week data on citalopram (n = 54), 27.7 percent were classified as responders and 24.1 percent were in remission. Among participants with data at 10 weeks, 46.2 percent were responders and 26.9 were in remission. Results at both time points were significant (p < .0001). Of the 11 individuals who dropped out of this case series, 10 had experienced an adverse event.
KQ5. Comparisons of Treatments
No studies were available to answer this key question.
KQ6. Modifiers of Outcomes of Treatment
No studies were available to answer this key question.
- Discussion - Traumatic Brain Injury and DepressionDiscussion - Traumatic Brain Injury and Depression
Your browsing activity is empty.
Activity recording is turned off.
See more...