Systematic and narrative reviews
There are two principal types of review papers: systematic and narrative reviews, each with a specific research question and rationale. In both types of reviews the objective of the review must be precisely defined and stated.
A systematic review should be carried out according to a specified methodological plan in order to minimise bias and omission of relevant studies. There would be a survey of databases or other types of register for the relevant literature using specified key words and subheadings and for a specific period. There may or may not be a meta-analysis included, which means a statistical combination of data coming from independent studies with the aim of producing an overall estimation of the effect of an intervention. Note that combining data from studies, which are too different in design or methods used may be inappropriate. For useful information and advices concerning systematic reviews a book (Advances in Rehabilitation) is available free on the internet (1).
A narrative review is a review of what is considered relevant for the topic and the aim of the review, but without a specified methodological plan as for a systematic review. Journal of Rehabilitation Medicine has a restrictive policy regarding publication of narrative reviews.
In reporting a systematic review or meta-analysis support will be given by guidelines of PRISMA (Preferred Items for Reporting of Systematic reviews and Meta-analysis) (2, 3). The authors are advised to follow the checklist in PRISMA. A systematic review paper should have a structured Abstract of no more than 200 words using headlines as Objective, Data Sources, Study Selection, Data Extraction, Data Synthesis and Conclusions and with at least 3 key words for indexing.
- Objective: Give precise statement of the primary objective for the review. Define if the review emphasises cause and diagnosis, prognosis, therapy and intervention, or prevention. Define if the review would be highly selective as including only randomized controlled trials (RCT) or have wider inclusion criteria.
- Data Sources: Present data sources used, including any time restriction. - Study Selection: Describe criteria to select studies for detailed review. Specify methods used, as blinded review, consensus, multiple reviewers.
- Data Extraction: Describe how extraction was made, including assessment of quality and validity.
- Data Synthesis: Present the main results of the review and state major identified sources of variation between studies.
- Conclusion: Give a clear statement of the conclusions made, its generalisability and limitations.
The Introduction of the paper could be similar to an original report, but without any longer literature survey, only reviewing shortly previous structural reviews and stating the reason and aim of the present review.
The Method section may have subheadings corresponding to the Abstract (Data Sources, Study Selection, Data Extraction) and should include clearly defined and reported inclusion and exclusion criteria, and specification of databases and other formal register, conference proceedings, reference lists and trial authors, which are used as sources.
The full search strategy should be given so that it is easy to reproduce. If it is considered too long to be published in the article, an electronic document as an Appendix may be alternative. The stages of selection usually include several steps, each undertaken by at least two independent researchers (identified in the Methods). There will be an initial selection from titles/abstracts to select the articles to be examined in full.
The full articles should be re-screened against the selection criteria. The articles fulfilling the criteria should be subjected to quality assessment. Summarize in a flow chart with the number of articles selected and reasons for rejection at each stage. The quality of the methodology should be assessed having an appropriate tool and also for outcome measures and blinding of outcome assessors. The tool that is most appropriate will depend on the extent and nature of the anticipated research evidence. Options include:
- the Scottish Intercollegiate Guidelines network (SIGN) methodology (4) or those developed within the Cochrane collaboration (5) which are primarily suitable for evidence that leans substantially on randomised controlled clinical trials.
- The GRADE system (6, 7, 8) allows somewhat greater flexibility to upgrading high quality non-randomised studies.
- The typology that was developed for the UK National Service Framework for Long Term Conditions (9), is designed to be applied across all types of research and evidence, including both quantitative and qualitative designs (10). It emphasises the quality of research, and the appropriateness of the method to the research question, regardless of study design. It may be particularly relevant in areas where qualitative and mixed methods research are to be included alongside quantitative methods - which may also include cohort and longitudinal studies, and other non-experimental designs.
The Result section corresponds to Data synthesis in the Abstract and may present tables with long lists of selected articles. Extracted data from trials should, when available, include report of randomization method, study population, intervention methods and delivery, reasons to losses at follow-up, information related to treatment monitoring, post-intervention assessments and follow-up. Report the major outcomes, which were pooled, and include odds ratios or effects sizes. Use when applicable meta-analysis. Numerical values should, when possible, be accompanied with confidence intervals. State the major identified sources of variation between reported studies, as differences in treatment protocols, co-interventions, confounders, outcome measures, length of follow-up, and dropout rates. Tables and figures must be self-explanatory and have appropriate title or caption. The methods for synthesis of evidence should be pre-determined. Sometimes it may not be possible to pool the data, but a synthesis of best evidence ought to be given.
The Discussion section should be structured similar to an original report. The findings should be discussed with respect to the degree of consistency, variation, and generalisability. New contribution to the literature based on the review conducted and where information is insufficient must be stated. Providing the limitations of the review would be helpful. Suggest the need for new studies and future research agenda.
Length of paper: The total length of the text should usually not be more than 5000 words (corresponding to 8-9 printed pages) and in addition tables and the reference list. The reference list should be comprehensive and will therefore often be rather long. However, in the printed version of a review paper normally not more than more than 100 references will be accepted. If needed and without an upper limit, additional references may be published only electronically with a link to such an Appendix given in the original version of the paper.
If the review is not invited, consult always the Editor-in-Chief before submitting such a manuscript to make sure that the topic is relevant for the journal, but also that it fits in a more general plan for publishing review papers and selection of specific topics. It is important for the journal to keep the time from submission to publishing as short as possible, as a review should be considered as “fresh goods”, keeping in mind the very fast development in certain areas and that one would like to have very recently published papers included among the references.
In a narrative review the selection of the papers reviewed depend more or less on the experience and attitude of the authors. If you intend to write a narrative review you must be critical to your own selection procedure. It is important that the selection criteria are defined and strictly followed. However, JRM, as many other journals, is reluctant to publish narrative reviews unless there is a compelling reason for doing so.
You should therefore always consult the Editor-in-Chief before submitting such a manuscript. A particular type of narrative reviews, which is published, is educational reviews on specific topics. They are usually invited and at present JRM has a collaboration with the European Board of Physical and Rehabilitation Medicine (EBPRM) and the International Society of Physical and Rehabilitation Medicine (ISPRM) on such programs. Non-invited educational reviews may occasionally be considered, but must fall into the scheme and topics chosen for such reviews during a certain period.
A narrative (educational) review may or may not have a structured Abstract; often an unstructured Abstract may be most relevant and should not exceed 200 words, followed by at least 3 key words for indexing.
In the Introduction section a survey of relevant literature and the aim and goal for the review should be presented. The headlines in the review have to be chosen according to the need of that particular review and may be subject for discussion with the editorial office.
There is usually no Method section.
The Discussion section could be structured along the lines for an original report being aware to discuss the limitation and its clinical message. Also for a narrative (educational) review paper a long reference list is accepted (see information about structured reviews). It is important by definition that an educational review has an educational approach, but at the same time to be published in a scientific journal, it should be sufficiently scientific and meant to be cited.
All reviews will go through the peer review process, even in cases of invited educational reviews. In all review papers, any conflict of interest must be presented. See general information for authors.
1. Liberati A, Taricco M. How to do and report systematic reviews and meta-analysis. In Franchignoni F, editor. Research in Physical & Rehabilitation Medicine. Pavia: Maugeri Foundation Books; 2010, p. 137-164.
2. Moher D, Liberati A, Tetzlaff J, Altman DA. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009; 151: 264-269.
3. Liberati A, Altman DG, Tetzlaff J et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009; 151: W-65-W-94.
4. Sign 50: A guideline developer's handbook. Edinburgh: Scottish Intercollegiate Guidelines Network; 2008.
5. van Tulder MW, Assendelft WJ, Koes BW, Bouter LM. Method guidelines for systematic reviews in the Cochrane Back Review group for Spinal Disorders. Spine 1997; 22: 2323–2330.
6. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso- Coello P, et al. Rating quality of evidence and strength of recommen- dations: GRADE: an emerging consensus on rating quality of evi-dence and strength of recommendations. BMJ 2008; 336: 924–926.
7. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Schunemann HJ et al. GRADE: what is “quality of evidence” and why is it important to clinicians? BMJ 2008; 336: 995-998.
8. Atkins D, Briss PA, Eccles M. et al. Systems for grading the quality of evidence and the strength of recommendations II: a pilot study of a new system. BMC Health Serv Res 2005; 5: 25.
9. Department of Health Long-term Conditions NSF Team. The National Service Framework for Long–term Conditions. Annex 2 Research and Evidence. UK: Department of Health; 2005, p. 88.
10. Turner-Stokes L, Harding R, Sergeant J, Lupton C, McPherson K. Generating the evidence base for the National Framework for long term conditions: a new research typology. Clin Med 2006; 6: 91-97.