SYSTEMATIC REVIEW PROTOCOL
Safety and effectiveness of pharmacotherapy for depression in adults who have sustained a traumatic brain injury: a systematic review protocol
1,2 1 3 1 2 4 Fiona J. Clay Luke A. Perry Amelia J. Hicks Rachel Batty Catalin Tufanaru Mahesh Jayaram 3,5 6 Jennie Ponsford Malcolm Hopwood
1Department of Psychiatry, Melbourne Neuropsychiatric Centre, The University of Melbourne, Melbourne, Australia, 2The Joanna Briggs Institute, Faculty of Health and Medical Sciences, The University of Adelaide, Adelaide, Australia, 3Monash-Epworth Rehabilitation Research Centre, Monash Institute of Cognitive and Clinical Neurosciences, Monash University, Melbourne, Australia, 4Department of Psychiatry, University of Melbourne, Royal Melbourne Hospital, Melbourne, Australia, 5School of Psychological Sciences, Monash University, Melbourne, Australia, and 6Professorial Psychiatry Unit, Department of Psychiatry, University of Melbourne, Melbourne, Australia
Review objective/question: The objective of this systematic review is to synthesize the current evidence on the effectiveness and harms of pharmacotherapy in the management of depression in adults who have sustained a traumatic brain injury. Keywords Depression; depressive symptomatology; pharmacotherapy; traumatic brain injury JBI Database System Rev Implement Rep 2017; 15(9):2270–2286.
Background of TBI survivors report ongoing neurobehavioral and emotional difficulties which may include depres- raumatic brain injury (TBI) comprises a clini- 6 cally heterogeneous group of disorders caused sive symptomatology. Studies indicate that neuro- T behavioral and emotional disturbances impair by trauma inflicted on the brain by a mechanical 7,8 force of external origin. Traumatic brain injury is quality of life the most for TBI survivors. one of the most common causes of death and dis- Depression is one of the most common and dis- abling psychiatric diagnoses for individuals with ability worldwide. In the United States alone, there is 9 an estimated 1.7 million cases of TBI annually,1 with TBI, with reports suggesting that between 12% an incidence of approximately 558 per 100,000 in and 60% of patients with TBI develop depres- 2 3 sion within 12 months of injury, irrespective of the United States and 235 per 100,000 in Europe. 10,11 People aged 15–24 years and over 64 years, males, severity. Rates of depression in persons with TBI are up to 7.5 times higher than rates reported those from low socioeconomic status backgrounds, 12 and individuals with cognitive impairment are dis- in the general population. Depressive symptoms can proportionately affected by TBI.4,5 As the majority vary between persons post-TBI and may include: of TBI cases occur as a result of motor vehicle fatigue, sleep disturbances, irritability, loss of interest 4 in social activities, poor concentration, memory prob- crashes, falls, assault and violence, the consequen- 13-15 ces of TBI can be severe. Although the majority (up lems and feelings of sadness. Moreover, anxiety symptoms co-morbid to depressive symptoms are to 90%) of TBIs are mild and physically recover 11 within three to six months, a substantial proportion common. Development of depressive symptoms post-TBI has been linked to increased post-concussive symptoms, poorer cognitive functioning, poor adher- ence with treatment plans, increased health service Correspondence: Fiona J. Clay, [email protected] utilization, higher healthcare costs as well as poorer Conflict of interest: The chief investigator (MH) has given talks on this psychosocial outcomes.10,16-19 Depression can under- topic for which travel and accommodation have been paid by the organizers. In addition, he has accepted fees for consulting and mine the survivor’s capacity for close relationships, research from the following pharmaceutical companies: Servier, Bio- social independence, activities of daily living includ- 9,20-22 nomics, Novartis, Eli Lilly and Lundbeck. ing employment and community integration. DOI: 10.11124/JBISRIR-2016-003326 Such symptomatology that leads to disabling
JBI Database of Systematic Reviews and Implementation Reports ß 2017 THE JOANNA BRIGGS INSTITUTE 2270
©2017 Joanna Briggs Institute. Unauthorized reproduction of this article is prohibited. SYSTEMATIC REVIEW PROTOCOL F.J. Clay et al.
outcomes highlights the importance of treating most recent search (October 2014) to identify rele- depression post-TBI. vant literature for both guidelines and systematic The management of depressive symptomatology reviews, additional relevant studies have been pub- in a person’s post-TBI requires a multimodal and lished.34,35 multidisciplinary approach incorporating social, While recent systematic reviews and guidelines psychological and medical interventions.23 The conclude that prescribing sertraline for depressive aim of pharmacotherapy is to provide targeted treat- symptoms post-TBI may have merit, these conclu- ment for discrete neurobehavioral symptoms and sions are based on a small number of participants emotional disturbances on an individualized basis (n ¼ 348), the majority of whom have sustained a while managing the adverse effects of medications. mild TBI. Barker-Collo et al.28 conducted a meta- To date, standard antidepressant pharmacological analysis of treatment for depression following mild agents for major depressive disorder in the general TBI, reporting that the effectiveness of sertraline population have been used to treat depression post- differed across the included studies. The authors’ TBI.24 Although depression is common post-TBI and overall finding was one of limited effective treat- pharmacotherapies are often prescribed, there is a ments for depression post-TBI. In a second meta- general lack of clinical consensus on their effective- analysis, Salter et al.30 reported that based only on ness and harms, and their place in multimodal man- within group comparisons, pharmacotherapy post- agement of depression post-TBI remains unclear.25 TBI might be associated with a reduction in depres- The extent to which the effectiveness and harms of sive symptoms but that these results should be available pharmacotherapies differs between people treated with caution. The three RCTs included in with TBI and the general population is unclear, with the Salter et al.30 meta-analysis examined three concerns that the TBI population may be more different pharmacotherapies and were based on susceptible to toxicity and refractory to therapeutic a very small number of participants (interventions effects.26,27 Given this, there is an urgent need to n ¼ 48, comparator n ¼ 33). The consensus across review the safety and effectiveness of pharmacother- systematic reviews is that more research is necessary apy for the management of depressive symptoms to develop evidence-based practice guidelines. In the across the TBI severity spectrum. absence of available RCT data, carefully evaluating Previous reviews in this area have focused on mild other study designs with an explicit acknowledge- TBI,28 pharmacological and non-pharmacological ment of inherent biases may provide some data interventions,28,29 clinical trials30 and neurological including data on safety that could inform practice. disorders, including TBI.31 A meta-analysis by Salter The current review forms part of a larger project et al.30 included literature published up to October which aims to synthesize the evidence for pharma- 2014, while a recently published clinical guideline on cotherapy in the treatment of neurobehavioral symp- this topic included literature to 2012 but did not toms post-TBI, as a prelude to the development of a report a risk of bias assessment.32 Of concern is that clinical practice guideline. The ultimate goal of none of the recent reviews have formally considered the guideline is to increase the number of skilled the adverse effects of pharmacotherapy for depres- prescribers with current knowledge of pharmaco- sion post-TBI.28-31 The Canadian clinical practice therapy for adults experiencing neurobehavioral guidelines for concussion/mild TBI and persistent symptoms post-TBI. symptoms graded all evidence on pharmacotherapy As we are focusing on a range of study designs, the as Grade C (expert opinion) except for the optimal systematic review will include a large descriptive duration of pharmacotherapy remaining as incon- component. Risk of bias will be considered when clusive which was graded A (based on at least interpreting the results, therefore, we have incorpo- one randomized controlled trial [RCT], systematic rated validated tools that will assess the methodo- review or meta-analysis).33 The authors of a system- logical quality of a range of study designs including atic review of drugs for behavior disorders after TBI RCTs, controlled non-randomized clinical trials, leading to French recommendation for good practice analytical observational studies and case series. published in 2016 concluded that there was insuffi- Our review will inform health providers, drug cient evidence to standardize drug treatments for regulators, policy makers, researchers and consum- these disorders, (including depression).32 Since the ers as end-users on the safety and efficacy of
JBI Database of Systematic Reviews and Implementation Reports ß 2017 THE JOANNA BRIGGS INSTITUTE 2271
©2017 Joanna Briggs Institute. Unauthorized reproduction of this article is prohibited. SYSTEMATIC REVIEW PROTOCOL F.J. Clay et al.
pharmacological management of depressive symp- resonance imaging), Glasgow Coma Scale score, tomatology in adults who have sustained a TBI. post-traumatic amnesia, loss of consciousness. Gaps in the evidence together with weaknesses in Self-report of TBI from either the individual or an study design will be highlighted, offering directions informant in the absence of other medical evidence for future research. Recommendations for pharma- regarding the head injury as outlined above will not covigilance will be outlined. be deemed sufficient. A preliminary search (performed in December Studies of acquired brain injury populations will 2016) of PubMed, Cochrane Library, JBI Database only be included if data for the TBI group are of Systematic Reviews and Implementation Reports, reported separately. The literature indicates variance PROSPERO and Epistemonikos found that there are in the criteria used to diagnose traumatic brain no recent systematic reviews that consider research injuries and depressive symptoms.36 There will be published since October 2014 or systematic review no exclusion of studies based on their use of different protocols exploring the precise review objective and diagnostic and assessment criteria. questions, with the inclusion criteria of interest for The current review will exclude: this review. Studies that include self-reported TBI in the ab- sence of other medical evidence regarding the Inclusion criteria head injury. Types of participants Studies that rely on self reports of depressive The review will include studies of persons post-TBI symptoms in the absence of a standardized diag- who present with depressive symptoms. The pres- nostic interview procedure, validated scale or ence or absence of depressive symptoms must be DSM criteria. determined using a standardized diagnostic inter- Studies that include mixed brain injury popula- view procedure, a validated assessment tool or tions where the data on TBI are not able to Diagnostic and StatisticalManualofMentalDis- be disaggregated. orders criteria (DSM). Studies that rely on self- Studies that report on clusters of individuals reported depressive symptoms that have not used rather than individuals themselves, for example, either an interview or valid rating scale will not clinic attendees, families, people living in a be included. defined geographical area. The review will consider studies that include adults, both males and females, who have sustained Types of intervention(s) a TBI of any severity. Studies where 80% of the The current review will consider studies that evalu- baseline population are 16 years and over will ate the use of pharmacotherapy interventions. All be eligible for inclusion. Traumatic brain injury is pharmacotherapy will be included. This will primar- defined as an alteration in brain function, or other ily include antidepressant medications including evidence of brain pathology, caused by an external selective serotonin re-uptake inhibitors (SSRI), sero- force. Studies will be included regardless of the tonin-norepinephrine (noradrenaline) reuptake mechanism of injury. Post-TBI patients in both the inhibitors (SNRI), tricyclic antidepressants and early stages of recovery and in rehabilitation will monoamine oxidase inhibitors. It may also include be included. Studies of participants with closed and benzodiazepines and neurostimulants. No restric- open head injury, penetrating head injury and non- tion on dose, duration, frequency, timing of delivery penetrating head injury will be eligible for inclusion. or combination of pharmacotherapies drugs will Participants must have definite medical evidence of be made. TBI. For the purpose of this review, this means that Studies reporting mixed interventions (e.g. phar- TBI has been documented in medical records macotherapy and psychological therapy) will be or other health-medical reports sighted by the included if the data for the pharmacotherapeutic research team associated with the published article. intervention is reported separately. These record reports must provide unequivocal The current review will exclude studies of com- evidence of injury. Examples of unequivocal evi- plementary medicines and over the counter medicine dence include findings from brain imaging unless these are part of the intervention, as a (e.g. computed tomography [CT] scan, magnetic co-intervention to pharmacotherapy.
JBI Database of Systematic Reviews and Implementation Reports ß 2017 THE JOANNA BRIGGS INSTITUTE 2272
©2017 Joanna Briggs Institute. Unauthorized reproduction of this article is prohibited. SYSTEMATIC REVIEW PROTOCOL F.J. Clay et al.
Comparator(s) Functional Independence Measure, Glasgow The current review will include studies that compare Outcome Scale-Extended, Community Integra- the intervention to all types of comparators. There tion Questionnaire.36 will be no restrictions on the type of comparator. Cognitive functioning using validated measure- Placebo (dummy or active), supportive or standard ment instruments including but not limited to the care or a non-pharmacological intervention will be Wechsler Adult intelligence scale and the Wechs- accepted. Studies comparing drugs within a pharma- ler Memory Scale.36 cological class will also be included. Health-related quality of life using validated measurement instruments such as the Medical Outcomes Outcomes Short Form 36. The current review will preferentially report on all Change in co-morbid anxiety symptoms mea- and any validated disease specific and generic mea- sured using validated measurement scales such surement tools used in each of the included studies as the Hospital Anxiety and Depression Scale.36 for the measurement of the primary and secondary outcomes of interest. For some of the outcomes of Types of studies interest, validated measurement scales may not be The current review will consider the following study available. In this situation, information on outcomes types: will be reported as described in the included study. Randomized controlled trials The primary outcomes of interest for the review Controlled non-randomized clinical trials will be: Quasi-RCTs including, but not limited, to con- Treatment efficacy as measured by an improve- trolled before and after studies, interrupted time ment in depressive symptoms and diagnoses. series with a control group, and interrupted time Severity of depression, response to treatment, series without a parallel concurrent control remission following treatment and loss of depres- group sion diagnosis will be reported according to the Analytical observational studies (including co- criteria specified in individual papers. This will hort and case control studies) be measured using validated measurement instru- Case series with pre-test/post-test outcomes ments such as but not limited to the Hamilton Single arm studies. Depression Rating Scale, Centre for Epidemio- The literature search will be limited to the English logic Studies Depression Inventory Hospital Anx- language and human subjects. If we identify poten- iety and Depression Scale and the Beck tially relevant titles in any other language, they will Depression inventory.36 be provided as an appendix to the review. No date Harms including adverse effects resulting from restriction will apply to the search for literature. pharmacotherapy and study dropouts: - Adverse effects: General and specific (qualita- Search strategy tively documented or documented using Med- Published and unpublished studies will be identified ical Dictionary for Regulatory Activities, using a seven-step approach. Initially the Cochrane World Health Organization Adverse Reaction Library and PubMed will be searched to identify key Terminology or other classification systems) words used in the titles and abstracts and in the search associated with pharmacotherapy. The need strategies of relevant published systematic reviews. for medication to manage adverse effects as Second, a more extensive search in international reported. electronic bibliographic databases will be undertaken - Study dropout: Loss to follow-up and leaving to further identify articles. We will search the follow- the study early due to adverse events, ineffi- ing databases with no date limitation: cacy of treatment, death, any reason. MEDLINE, OVID SP interface The secondary outcomes of interest for the review PubMed excluding MEDLINE will be: CINAHL: Cumulative Index to Nursing and Functional independence in activities of daily Allied Health Literature living measured using validated measurement Embase: Excerpta Medica Database excluding instruments such as but not limited to the MEDLINE, OVID SP interface
JBI Database of Systematic Reviews and Implementation Reports ß 2017 THE JOANNA BRIGGS INSTITUTE 2273
©2017 Joanna Briggs Institute. Unauthorized reproduction of this article is prohibited. SYSTEMATIC REVIEW PROTOCOL F.J. Clay et al.