Description of the condition
There are several definitions of traumatic brain injury in the academic literature. For the purpose of this review, we define traumatic brain injury as an alteration in brain function, or other evidence of brain pathology, caused by an external force (Menon 2010). Data from the United States suggest that an estimated 1.7 million traumatic brain injuries (TBIs) occur annually. Nearly 80% of TBI patients are treated and discharged from emergency departments. TBI is a contributing factor to a third of all injury-related deaths (Faul 2010). The incidence rate of TBI in Europe is estimated at 235 per 100,000 population (Tagliaferri 2006). Worldwide, TBI is a major cause of disability, particularly with declining mortality rates. With improvements in retrieval, and better neurosurgical and intensive care treatments, many more people survive injuries that would have been fatal just a few decades ago. However, survivors may be left with long-term disabilities (Fann 2009). Much of the disability following TBI is hidden as many survivors may not have physical evidence of injury, but may be left with cognitive, psychological and behavioural difficulties (Khan 2003).
Major depression appears to be the most prevalent psychiatric disorder following TBI, with a reported period prevalence of 33% to 42% within the first year and 61% in the first 7 years following injury (Fann 2009). TBI increases the risk of major depression. A retrospective study of 100 participants within 0.5 to 5.5 years after TBI showed that the prevalence of depression increased from 15% pre-TBI to 45% post-TBI (Whelan-Goodinson 2009b). Another study of 164 people by Deb 1999 reported that the rate of depression at one year post-TBI was 13.9%, compared to 2.1% in the general population.
There are issues regarding the reliability and validity of diagnosis of depression following TBI. Many features of depression overlap with symptoms of TBI (Whelan-Goodinson 2009a). For example, changes to sleep, libido and appetite are seen in TBI. This may lead to over-diagnosis of depression. On the other hand, people may deny the presence of low mood and depressive symptoms as part of a general awareness deficit. This may lead to under-diagnosis of depression following TBI.
Depression may be a result of direct or secondary injury to brain tissue. Studies of patients with TBI have shown that focal and diffuse injuries preferentially affect frontal and temporal poles of the brain. Injury to these areas of the brain, and to related circuits such as frontal lobe-basal ganglia circuit, have been associated with the onset of depression following TBI (Fann 2009). Several studies have investigated the role of psychological and social factors in the development of depression following TBI (Gomez-Hernandez 1997; Hoofien 2001; Morton 1995). Disruption to close relationships with family members, social isolation, loneliness, change or loss of employment, loss of income and financial worries have been identified as risk factors for development of post-TBI depression (Jorge 1993; Thomsen 1984). However, the exact mechanism of interaction of these various factors is poorly understood (Fann 2009).
Description of the intervention
Since the introduction of monoamine oxidase inhibitors and tricyclic antidepressants in the 1950s, many drug therapies have been introduced for treating depression. Currently, antidepressants are the most commonly used drugs in depression and the first line of treatment for depression with at least moderate severity (Kupfer 2012; NICE 2009). Selective serotonin reuptake inhibitors (SSRIs) were introduced in the 1980s and soon became the most commonly prescribed antidepressant medication (Stafford 2001). Among people with depression, approximately 20% will respond with no treatment at all, 30% will respond to placebo and 50% will respond to an antidepressant drug (NICE 2009). This gives a number needed to treat for an additional beneficial outcome (NNTB) of three for antidepressants over waiting list controls and five for antidepressants over placebo (NICE 2009).
How the intervention might work
The three main neurotransmitters implicated in depression are serotonin, noradrenaline and dopamine. Pharmacological agents work in depression by acting on neurotransmitter systems in the brain (Anderson 2004). The biological actions of antidepressant drugs are still not fully understood. It is thought that most antidepressants exert their initial effects by increasing the levels of neurotransmitters such as serotonin and noradrenaline in the intrasynaptic connections. However, clinically-observable antidepressant effects occur when the drug is administered for days to weeks. This suggests downstream effects of the drugs on signalling and plasticity pathways in the brain (Machado-Vieira 2008). Recent research has shown that alterations in the brain structure may occur in depression, and antidepressant agents have cellular effects leading to enhanced expression of neuroprotective factors that increase neuronal survival and synaptic connectivity (Anderson 2004).
Why it is important to do this review
There is substantial evidence for the use of pharmacological agents in depressive disorder (NICE 2009). The evidence for effectiveness of pharmacological agents for depressive disorder in people with traumatic brain injury is scarce (Warden 2006). A systematic review considered all treatments (pharmacological, psychological, electroconvulsive therapy, transcranial magnetic stimulation etc) for depression after traumatic injury (Fann 2009). This review was very broad in its scope and did not focus on pharmacological interventions. It also included lower quality studies such as non-randomised single blind trials, and open label trials. This review aims to focus on evidence for the use of pharmacological agents for people with traumatic brain injury and depression. We will only include randomised controlled trials.
To assess the effects of pharmacological interventions for depressive disorder in people with traumatic brain injury.
Criteria for considering studies for this review
Types of studies
Randomised controlled trials. We will only include placebo-controlled trials.
Types of participants
People with depressive disorder following traumatic brain injury.
We will only include participants fulfilling the diagnostic criteria according to DSM-IV (Diagnostic and Statistical Manual of Mental Disorders) for minor and major depression, or alternatively those who have a diagnosis according to ICD-10 (International Classification of Diseases) criteria for mild, moderate or severe depression.
Types of interventions
Any drug acting on the central nervous system (chapter 4 of the BNF 2011):
4.1 hypnotics and anxiolytics;
4.2 drugs used in psychoses and related disorders;
4.3 antidepressant drugs;
4.4 central nervous system stimulants and drugs used for attention deficit hyperactivity disorder;
4.5 drugs used in the treatment of obesity;
4.6 drugs used in nausea and vertigo;
4.8 antiepileptics drugs;
4.9 drugs used in parkinsonism and related disorders;
4.10 drugs used in substance dependence;
4.11 drugs for dementia.
Types of outcome measures
We will use final follow up measures as outcome measures. If there are wide time range variations of the final measures, we will consider groups of studies separately for short-, medium-, and long-term measures. The short, medium and long ranges will be decided based on internationally accepted clinical guidance.
Reduction in depressive symptom scores on validated tools such as the Hamilton Depression Rating Scale (HAM-D) (Hamilton 1960), Beck Depression Inventory (BDI) (Beck 1961), Hospital Anxiety and Depression Scale (HADS) (Zigmond 1983), and Montgomery-Aperg Depression Rating Scale (MADRS) (Montgomery 1979).
- Adverse effects from pharmacological interventions, such as gastrointestinal disturbances, weight change, insomnia, worsening of anxiety, agitation.
- Improvement in quality of life measured using validated scales such as Quality of Life after Brain Injury (QOLIBRI) (von Steinbüchel 2010).
Search methods for identification of studies
The search for studies will not be restricted by language or publication status (published or unpublished).
The Cochrane Injuries Group Trials Search Co-ordinator will search the following electronic databases:
- Cochrane Injuries Group Specialised Register (present version);
- Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library) (most recent issue);
- MEDLINE (Ovid SP) (1946 to present);
- Embase Classic + Embase (Ovid SP) (1947 to present);
- PsycINFO (1806 to present);
- CINAHL (EBSCO) (1982 to present);
- LILACS (BIREME) (1982 to present);
- ISI WOS Science Citation Index Expanded (SCI-EXPANDED) (1970 to present);
- ISI WOS Conference Proceedings Citation Index- Science (CPCI-S) (1990 to present).
The search strategy in Appendix 1 was formulated in MEDLINE and will be adapted, where necessary, for use in each of the other databases.
Searching other resources
The bibliographies of all identified trials and previously published reviews will be scanned for additional studies. We will also search www.clinicaltrials.gov and www.controlled-trials.com for ongoing and unpublished studies.
Data collection and analysis
Selection of studies
We will collate search results and remove duplicates. Two authors (JJV and NL) will independently assess the search results and select studies for further investigation based on the following classification:
- unclear eligibility
- not eligible
For those falling into either category 1 or 2 we will retrieve the full article.
Two authors (JJV and NL) will independently assess full text articles against the selection criteria. In cases of disagreement, we will reach a decision by open discussion and consensus. We will record reasons for exclusion.
Data extraction and management
Data will be extracted independently by two authors (JJV and NL) using a data extraction form (Appendix 2). We will enter data into Review Manager software. Where data are not available in the published trial reports, we will attempt to contact the authors and request the missing information.
Assessment of risk of bias in included studies
Two authors (JJV and NL) independently will assess the risk of bias in included studies using the Cochrane 'Risk of bias' tool (Higgins 2011). We will assess the risk of bias in the following domains: sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other sources of bias. We will assess studies as being at low, high or unclear risk of bias. We will resolve disagreements by discussion.
Measures of treatment effect
We will calculate the risk ratio (RR) and its 95% confidence interval (CI). Risk ratios are better understood by patients and healthcare professionals compared to odds ratios (ORs) (Higgins 2011). For outcomes where an event is rare, we may use Peto's OR, if appropriate. Where there would be a clinically-relevant summary of the effects of treatment for the available set of included studies, we will calculate the number needed to treat to provide benefit (NNTB) and the number needed to treat to induce harm (NNTH).
As we expect the studies to use different scales to measure the same outcome (for e.g. BDI and MADRS to measure depressive symptoms), we will use the standardised mean difference (SMD) to measure the treatment effect in continuous data. Should studies use the same scale, we will use the mean difference (MD). If data from cluster or cross-over trials are included in the review, we may use the generic inverse variance method, in which case we would transform the pooled OR into an RR.
Unit of analysis issues
A major concern of cross-over trials is the carry-over effect; that is, the effect of one intervention period carried over to a subsequent intervention period. As a consequence, on entry to the second phase, the participants may differ systematically from their initial state despite a washout phase (Higgins 2011). Where carry-over effects are thought to exist (i.e. they are clinically possible), we will only include data from the first period in the analysis. Two authors will judge the possibility of a carry-over effect.
Studies with multiple intervention groups
Where the study involved more than two intervention groups, we will include the intervention groups relevant to the systematic review (for example if a study compares antidepressants versus psychological treatments versus placebo, we will only consider antidepressant treatments versus placebo). If more than one intervention group is relevant, we will decide which intervention group is relevant for a particular meta-analysis. We will record all interventions of a multi-intervention study in the table 'Characteristics of included studies', either in the 'intervention' cell or 'notes' cell. We will not double-count data. The shared group will be split into two or more smaller groups and include two or more comparisons.
We will assess whether the trials avoided unit of analysis errors by conducting the analysis at the same level as the allocation, using a summary measurement from each cluster. If this is not the case, we will assess whether the trials used statistical methods that allow analysis at the level of individuals.
If the cluster trials did not report an appropriate analysis or analysed the data as if the randomisation was performed on individuals and was not clustered, we will attempt to perform an approximately correct analysis as recommended by Cochrane Handbook (Higgins 2011) if the following data can be extracted:
- number of clusters randomised to each intervention group or the average (mean) size of each cluster.
- outcome data ignoring the cluster design for the total number of individuals.
- an estimate of the intra-cluster correlation coefficient (ICC).
We will identify any cluster trials in the review and state how we handled the data. We will conduct a sensitivity analysis and seek statistical advice from the Cochrane Injuries Group's statistician.
Dealing with missing data
We will attempt to contact the original investigators for information regarding missing data and whether or not the data missing can be assumed to be 'missing at random'. If the missing data are assumed to be missing at random, then only the available data will be analysed. If the data missing are not assumed to be missing at random, then we will follow the principles outlined in the Cochrane Handbook (Higgins 2011). We will input the missing data with replacement values and treat them as if they were observed. We will use methods such as last observation carried forward (LOCF). We will then perform sensitivity analyses to assess how sensitive the results are to reasonable changes to the assumptions made, and we will address the potential impact of missing data on the findings of the review in the Discussion section of the review.
Assessment of heterogeneity
We will consider the interventions and participants in the included studies when deciding if it is appropriate to pool data, and whether to use a fixed-effect or random-effects model. We will examine statistical heterogeneity using the I
Assessment of reporting biases
We will use a funnel plot to assess reporting bias. However, we will only use a funnel plot if at least 10 studies are included in the meta-analysis. We will consider using a funnel plot even if the studies are of similar sizes. Tests for funnel plot asymmetry will be interpreted in light of visual inspection of the funnel plot.
We will use the random-effects model or the fixed-effect model depending on our assessment of heterogeneity. When carrying out meta-analysis, the weight given to each study will be inverse of variance so that studies that have more precise estimates will be given more weight (inverse-variance method).
Subgroup analysis and investigation of heterogeneity
If two or more studies can be included in a subgroup analysis, we will undertake a subgroup analysis based on:
- Different types of drugs within a class
- Severity of depression
- Severity of traumatic brain injury (based on widely-accepted classification systems such as the Glasgow Coma Scale (Saatman 2008))
- Setting (community, inpatient sample)
- Age (children and adolescents, adults)
A priori sensitivity analysis is planned for studies that have an unclear or high risk of bias based on the following criteria:
- concealment of allocation
- blinding of assessors
- extent of drop-outs
Depending on the nature and peculiarities of studies identified in the review process, we may identify more issues that are appropriate for sensitivity analysis.
We thank Karen Blackhall (Trial Search Coordinator) for helping us to develop the search strategy.
Appendix 1. Search strategy
1. exp Craniocerebral Trauma/
2. exp Brain Edema/
3. exp Glasgow Coma Scale/
4. exp Glasgow Outcome Scale/
5. exp Unconsciousness/
6. exp Cerebrovascular Trauma/
7. exp Pneumocephalus/
8. exp Epilepsy, post traumatic/
9. exp Cerebral hemorrhage, traumatic/
10. ((head or crani* or cerebr* or capitis or brain* or forebrain* or skull* or hemispher* or intra?cran* or inter?cran* or intracran* or intercran*) adj3 (injur* or trauma* or damag* or lesion* or wound* or destruction* or oedema* or edema* or contusion* or concus* or fracture*)).ab,ti.
11. ((head or crani* or cerebr* or brain* or intra?cran* or inter?cran* or intracran* or intercran*) adj3 (haematoma* or hematoma* or haemorrhag* or hemorrhag* or bleed* or pressur*)).ti,ab.
12. (Glasgow adj (coma or outcome) adj (scale* or score*)).ab,ti.
13. "rancho los amigos scale".ti,ab.
14. ("diffuse axonal injury" or "diffuse axonal injuries").ti,ab.
15. ((brain or cerebral or intracranial) adj3 (oedema or edema or swell*)).ab,ti.
16. ((unconscious* or coma* or concuss* or 'persistent vegetative state') adj3 (injur* or trauma* or damag* or wound* or fracture* or contusion* or haematoma* or hematoma* or haemorrhag* or hemorrhag* or pressur*)).ti,ab.
17. exp coma/
18. (injur* or trauma* or damag* or wound* or fractur* or contusion* or haematoma* or hematoma* or haemorrhag* or hemorrhag* or pressur* or lesion* or destruction* or oedema* or edema* or contusion* or concus*).ti,ab.
19. 17 and 18
20. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 19
22. randomized controlled trial.pt.
23. controlled clinical trial.pt.
25. clinical trials as topic.sh.
28. 21 or 22 or 23 or 24 or 25 or 26 or 27
29. (animals not (humans and animals)).sh.
30. 28 not 29
31. (rat* or rodent* or mouse or mice or murin* or dog* or canine* or cat* or feline* or rabbit* or pig* or porcine or swine or sheep or ovine* or guinea pig* or horse* or hamster* or goat* or chick or cattle or bovine).ti.
32. 30 not 31
33. 20 and 32
35. depressive disorder/ or depressive disorder, major/ or dysthymic disorder/
36. (depress* or melancholia).ab,ti.
37. 34 or 35 or 36
38. 33 and 37
39. exp Antidepressive Agents/
40. exp Antipsychotic Agents/
41. (Anti?psychotic* or Anti?depress* or Antipsychotic* or Antidepress*).mp.
42. (Benzodiazepin* or nitrazepam or Flurazepam or Loprazolam or Lormetazepam or lorazepam or temazepam or diazepam or alprozolam or chlordiazepoxide hydrochloride or oxazepam or Buspirone hydrochloride or meprobamate or Zaleplon or Zolpidem Tartrate or Zopiclone or Chloral hydrate or clomethiazole or Promethazine hydrochloride or Sodium Oxybate or melatonin or benperidol or chlorpromazine hydrochloride or flupentixol or haloperidol or levomepromazine or pericyazine or perphenazine or pimozide or Prochlorperazine or promazine hydrochloride or sulpiride or trifluoperazine or zuclopenthixol acetate or zuclopenthixol or amisulpride or aripiprazole or clozapine or olanzapine or paliperidone or quetiapine or risperidone or flupentixol decanoate or fluphenazine decanoate or haloperidol or olanzapine embonate or pipotiazine palmitate or risperidone or zuclopenthixol decanoate or Carbamazepine or valproic acid or sodium valproate or lithium carbonate or lithium citrate).mp.
43. (Tricyclic* or amitriptyline hydrochloride or clomipramine hydrochloride or dosulepin hydrochloride or doxepin or imipramine hydrochloride or lofepramine or nortriptyline or trimipramine or mianserin hydrochloride or trazodone hydrochloride or monoamine oxidase inhibitors or phenelzine or isocarboxazid or tranylcypromine or moclobemide or selective serotonin re-uptake inhibitors or citalopram or escitalopram or fluoxetine or fluvoxamine maleate or paroxetine or sertraline or agomelatine or duloxetine or flupentixol or mirtazpine or reboxetine or tryptophan or venlafaxine or Atomoxetine or dexamfetamine sulphate or methylphenidate hydrochloride or modafinil or Orlistat or Cinnarazine or cyclizine or promethazine hydrochloride or promethazine teoclate or droperidol or perphenazine or prochlorperazine or trifluperazine or Domperidone metoclopramide hydrochloride or Granisetron or ondansetron or palonosetron or Aprepitant or fosaprepitant or Cannabinoids or nabilone or hyoscine hydrobromide or betahistine dihydrochloride or Aspirin or paracetamol or nefopam hydrochloride).mp.
44. (Opiod* or uprenorphine or codeine phosphate or diamorphine hydrochloride or dihydrocodiene tartrate or dipipanone hydrochloride or fentanyl or hydromorphon hydrochloride or meptazinol or methadone hydrochloride or morphine salt* or oxycodone hydrochloride or papaveretum or pentazocine or pethidine hydrochloride or tramadol hydrochloride or Tolfenamic acid or Almotriptan or eletriptan or frovatriptan or naratriptan or rizatriptan or sumatriptan or zolmitriptan or ergotamine tartrate or Pizotifen or clonidine hydrochloride or methysergide or Carbamazepine or eslicarbamazapine acetate or oxcarbamazepine or ethosuximide or gabapentin or pregabalin or lacosamide or lamotrigine or levetiracetam or Phenobarbital or primidone or phenytoin or rufinamide or tiagabine or topiramate or vigabatrin or zonisamide or clobazam or clonazepam or fosphenytoin sodium or paraldehyde or phenobarbital sodium or phenytoin sodium).mp.
45. (Apomorphine hydrochloride or bromocriptine or cabergoline or pergolide or pramipexole or ropinirole or rotigotine or levodopa or co-beneldopa or co-careldopa or rasagiline or selegiline hydrochloride or entacapone or tolcapone or amantadine hydrochloride or orphenadrine hydrochloride or procyclidine hydrochloride or trihexyphenidyl hydrochloride or Piracetam or riluzole or tetrabenazine or botulinum toxin type A or botulinum toxin type B or Acamprosate or disulfiram or bupropion hydrochloride or nicotine or verencline or buprenorphine or methadone hydrochloride or lofexidine hydrochloride or naltrexone hydrochloride or Donepezil hydrochloride or galantamine or memantine hydrochloride or rivastigmine).mp.
46. 39 or 40 or 41 or 42 or 43 or 44 or 45
47. 38 and 46
Appendix 2. Data extraction form
Trial ID: E.g. Jones et al 2000
Outcome measures used
Mean baseline score of all outcome measures
Mean endpoint score of all outcome measures
Outcomes able to use and why
Outcomes unable to use and why
ANY ACTIONS: e.g. write to the authors
Contributions of authors
Joe John Vattakatuchery, Nismen Lathif and Jaya Joy: protocol development.
Hugh Rickards and Andrea Cavanna: supervision of protocol development.
Declarations of interest
Joe Vattakatuchery: I have attended educational events organised and/or sponsored by pharmaceutical companies.
All other authors: None known.