This article is deliberately divided in to subsections to discuss the topic thoroughly and systematically.
In this article the author details how a researcher should conduct a systematic review of bipolar disorder. This will help future reviewers to conduct their reviews according to the most scientific standards of the day.
In contrast to a narrative review, the author provides a summary of randomized controlled trials on the specific and focused clinical question of the review, using explicit methods to search, critically appraise, and synthesized the literature systematically. He brings together a number of separately conducted studies, regardless of their findings, and synthesis their results.
The reviewer is well aware of the needed rigor in the preparation of a systematic review and conducts a formal process for this purpose. This includes a comprehensive and systematic search for primary studies on the focused question, followed by selection of studies using clear and reproducible eligibility criteria, critical quality appraisal of primary studies, and finally synthesis of results according to predetermined and explicit methods.
Aims and objectives
The author endeavors to examine the effectiveness of various types of adjunct psychosocial interventions in management of bipolar affective disorder, using an explicit a priori methodology according to a standard review protocol.
A protocol was developed to minimize bias, by taking all key methodological decisions clearly and systematically, before going to the literature. The protocol aimed to set out the tasks and clear and explicit methods to be followed in this systematic review and to ensure that results are reproducible.
Focused question: Are different methods of adjunct psychosocial management for people with bipolar disorder useful, effective and superior to standard medical treatment solo, in relapse prevention, functional improvement, and reduction of severity and duration of bipolar episodes?
Definition of psychosocial management for the purpose of conducting a systematic review: Psychosocial management is an umbrella term used to cover various forms of psychological therapies used in the management of bipolar disorder. This includes cognitive therapy, family focused therapy, interpersonal and social rhythm therapy, psycho-education, and relapse prevention. The author includes studies where some of the following elements were considered:
1) Education about the illness: All forms of studies included should have offered general education to the subjects of the study intervention arm, in order to improve the individual’s awareness and understanding of bipolar disorder.
2) Monitoring and self regulation: Monitoring, vigilance, identification and management of acute symptoms and relapse prevention should have been part of
the agenda for the intervention arm. Individual’s ability to recognize and manage the relapse prodromes or the internal and external stressors that may increase their vulnerability to future relapse should have been discussed.
3) Enhancement of adherence to pharmacotherapy: Among the included studies, forming a therapeutic alliance with the psychiatrist and importance of adherence to pharmacotherapy should have been discussed to the therapy group. Management of side effects, and pros and cons of medical treatment and dangers of abrupt treatment withdrawal should have been discussed.
4) Cognitive processes in bipolar disorder should have been discussed with the intervention arm of included studies. This might include education on techniques to monitor, examine and change dysfunctional habits and behavior associated with undesirable mood consequences.
5) Studies included should have described the content and duration of the psychological therapy for the intervention arm, and have a follow up period of at least two years.
6) A minimum total of 6 sessions should have been delivered to the study participants included in each study.
The above mentioned elements are considered to be integral parts of psychosocial management of bipolar disorder. Different treatment regimes, might give more weight and emphasis on one or the other, but it is deemed necessary for some of the above elements to be introduced, no matter how briefly through the course of therapy provided.
Eligibility criteria with rationales
Type of studies
Randomized controlled trials, Nonrandomised and quasi-randomized trials should not included.
Rationale: The reviewers only includes RCTs, for the reason that randomized trials are the gold standard of examination of effectiveness, they ensure random allocation to intervention and control arms of the studies, help eliminate selection bias, and ensure the similarity in characteristics and treatments of both groups in the long run, except for the intervention under study.
1) In all included studies, all patients had a diagnosis of bipolar disorder I or bipolar disorder II, according to explicit diagnostic criteria, identified by structured clinical interviews.
Rationale: To avoid bias resulting from different definitions between studies the author restricts the review to studies using DSM IV criteria as the reference standard for mental disorders.
2) Studies should had not solely recruited patients who were suffering from acute mania or patients who were hospitalized in acute wards at the point of recruitment. Studies should have not recruited patients with only depressive or manic episodes.
However, the studies might include those attending day centers. Studies with patients with rapid cycling or mixed affective episodes can not be included.
Rationale: Reviewers should try to include studies where the recruited patients, present with similar clinical picture and need similar sort of support and treatment. The groups that fall under exclusion criteria stated above have different needs, severity of illness and compliance to the intervention provided.
3) Studies might include patients with mild levels of depression (defined as a Beck depression inventory of <15) can be included.
Rationale:This group of patients can benefit from therapies provided and be able to comply with the treatment.
4) Patients on both arms of the included studies should be on regular prophylactic medication.
Rationale: Standard prophylactic pharmacotherapy is the mainstay of treatment of bipolar disorder and it is considered unethical to interrupt medical treatment for experimental purposes. Non-compliance with the medical treatment will significantly change the clinical outcomes of either arm of the study.
5) The included studies only should have trialed adults (between 18 and 65).
Rationale: Studies should include examine the adult age group. The clinical picture, diagnosis and management of childhood bipolar disorder vary significantly from the adult conditions. Older groups commonly have co-morbid physical, mental and cognitive conditions that might introduce confounding to the results and would be very difficult to account for.
6) Studies that mainly focus on patients with other psychiatric co-morbidities or bipolar disorder secondary to organic causes should not be included.
Rationale: The clinical picture, diagnosis, management and complications vary in the above groups.
7) Only studies with patients with a history of at least 2 bipolar episodes and at least one episode within last two years should be included (not in full remission for more than 2 years).
Rationale: A minimum number of 2 episodes ensures diagnostic certainty, and helps avoid first time diagnostic errors to include other mental health conditions such as PTSD and schizophrenia. One episode should be within last 2 years, to ensure that the disease was ongoing at the time of recruitment and the patient was not in full long-term recovery or burnt out phase.
The included studies should examine some of the following as their outcome measures:
1) Mean number of bipolar episodes and mean number of bipolar related hospitalisation.
2) Time to next episodes (as defined by DSM IV criteria for manic, depressive and mixed episodes).
3) Changes in global functioning and/or duration or severity of bipolar symptoms, using validation instruments.
4) Mean number of days fulfilling the DSM IV diagnosis a bipolar episode.
5) Mean number of episode free days.
6) Mean number of bipolar related days in hospital.
7) Rate of suicide in intervention and control groups.
As the rigour of systematic search methods is an important determinant of unbiased systematic reviews, extended systematic search methods including hand-searching, reference lists, personal communication searching of specialised databases and registries; is used by the reviewer to carry out this review.
The search strategy aims at increasing sensitivity of our search, by minimising non-retrieval of the documents that were relevant to the review question and to maximise retrieval of the documents that are relevant to the review respectively.
Every effort should be made for the search to be as extensive as possible. This means that the reviewer may err on the side of retrieval of too many items and subsequently excluding those that are not relevant after direct examination of the papers.
The search terms used in a systematic review are constructed using the following strategy:
1) The reviewer derives major terms from the questions by identifying the population, interventions and outcomes.
2) Alternative spellings and synonyms are identified for major terms. The reviewer also includes terms identified through discussions with experts in the field and subject librarians of mental health trusts.
3) The keywords are checked in any relevant papers available to the reviewer at the outset.
4) The Boolean operator OR is used to incorporate alternative spellings and synonyms.
5) The Boolean operator AND is used to link the major terms from the population, interventions and outcomes.
6) Brackets are used for grouping of terms.
7) Each stage is double checked with a specialist librarian based at mental health library.
The following specified electronic databases have to be searched from inception with the following Mesh terms (or their equivalents in different databases):
(“bipolar disorder” OR “manic depressive psychosis” OR “bipolar depression” OR “manic depression”) Combined with the following subject headings using the Boolean connector AND (Cognitive therapy OR social rhythm therapy OR psycho-education OR family therapy OR family focused therapy OR psychosocial management OR psychosocial intervention OR psychological therapy).
The following free text searches are combined using the Boolean connectors accordingly: (“bipolar disorder*” OR “bipolar depress*” OR “manic depress*” ) AND ( Cognitive therap* OR cognitive behavio* OR social rhythm therap* OR psycho-education OR psychosocial intervention* OR psychosocial management* OR psychosocial treatment OR relapse prevention OR psychological therap* OR psychological management OR psycho-education OR family therap* OR family focus*)
1.The Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register (CCDANCTR) will be checked.
2. The Cochrane Central Register of Controlled Trials (CENTRAL) will be checked meticulously according to existing database.
The following additional databases are searched to check the completeness of the review:
5.CCDANCTR and CENTRAL
The reference lists of all identified randomised controlled trials, other relevant papers and major textbooks of bipolar disorder written in English should be checked. This process will be repeated until no further reports or papers seem relevant, and until no new studies are found that are not already identified electronically.
The journal Bipolar Disorder, will be hand-searched. No further studies should be found though this method which were not already identified among the electronic hits.
The authors of significant papers are identified from authorship lists over the last two decades. They, and other experts in the field, are contacted and asked of their knowledge of other published or unpublished studies, relevant to this review. No further papers are identified through this process.
Inclusion and exclusion process
Studies scoped by the search strategies elaborated earlier above were checked to ensure satisfaction of both inclusion and exclusion criteria.
Abstracts of all cited studies should be obtained. Studies will be excluded at this stage only if unequivocal evidence is found in the abstracts. When this is not possible full texts of studies are obtained to take a decision regarding exclusion. Excluded studies are recorded with details of the author´s reasons for exclusion.