Writing a review article: Not a piece of cake
As a facilitator of AuthorAID scientific writing courses and an academic as well as a researcher, I have witnessed a lot of curiosity from course participants and students about writing/conducting review articles. This has become amplified after impediments imposed by COVID-19 on primary data collection. Often, it is presumed that writing a review paper is an easy and fast process. However, it demands expertise in the field and time commitment by the researcher. I briefly highlight review articles, their types, guidelines and resources for conducting a review, especially for systematic and scoping reviews.
Review articles
Review articles summarize or synthesize the current understanding of a topic of interest. Reviewers use different standard methods to examine and pool findings and ideas from other studies of interest. Writing a review paper is a vigorous process. A reviewer must have expertise in the field and an understanding of the review process. In addition, it demands a significant time commitment. Often people ask, “I want to write a systematic review; how can I do it?” Here the point is, an author should be clear whether or not a systematic review of their topic is required. If one is needed, which type of review is necessary? There are different types of reviews. Each review type has different objectives, so based on need and gaps identified in the literature, a reviewer should take a decision as to the type of review. It should not only be conducted just for a sake of increasing the publication metrics of the researcher but also for scientific reasons.
Many journals consider narrative reviews for publication by experienced researchers (those with significant experience in their field), or such reviews may be published by invited authors. It is highly unlikely that a narrative review from a young or early-career researcher will be published without any collaboration with more senior researchers. However, some types of reviews (i.e., systematic or scoping reviews) could be published by young or early-career researchers, provided that the quality of the paper is technically and methodologically sound. It is recommended that early-career or novice researchers form a network or collaboration and get trained to conduct such reviews to ensure quality. I usually try to include some of my interested students in my review projects, which helps them with developing their expertise and this supports the review process too.
Types of reviews
Narrative/general reviews: A narrative review is an examination of the current literature. It can cover a wide range of subjects at various levels of completeness and comprehensiveness. These reviews aim to provide a critical interpretation by synthesizing the previous literature using the creative and expert judgement of the reviewer(s) (e.g., general review article).
Scoping reviews: A scoping review is conducted with more specific objectives than narrative reviews, following systematic synthesis of evidence. Typically, scoping reviews aim to identify knowledge gaps and provide recommendations for future research. They provide broad overviews of evidence on a given topic and help to clarify key concepts, particularly in emerging fields (e.g., Scoping review article).
Rapid reviews: Rapid reviews are performed when quick evidence summaries are required. They are classically conducted similarly to the traditional systematic review, but the review process is rationalized to save time. This involves narrowing the scope, searching fewer databases, or extracting limited data from each included study based on the immediate need, with proper justification (e.g., Rapid review article).
Systematic reviews (with or without meta-analysis): A systematic review (SR) is a type of review that follows a standard protocol as well as guidelines, inclusion and exclusion criteria, data extraction and synthesis of findings, etc. It provides a comprehensive summary of the scholarly literature related to a specific research topic or question. SRs have a defined focus that aims to answer relatively specific research questions. The process involves formulating strict criteria for inclusion and exclusion of studies and developing and documenting extensive search strategies. These may include assessing the quality or 'risk of bias' of individual studies and extracting pertinent information from study reports. PRISMA is a standard reporting guideline for systematic reviews and meta-analysis.
SRs are considered the gold standard in medical research and form the basis of decision-making in Evidence-Based Treatment (EBT) and evidence-based behavioral practice (EBBP). A SR is often written by a panel of experts after reviewing all information from both published and unpublished studies. SRs' protocols should be registered in PROSPERO, the Cochrane library or the JBI site to prevent duplication. Usually, each SR should have its own specific objectives, and not duplicate the same work done by others, but it can be updated as needed. Unregistered systematic review protocols are unlikely to get published. The key aspect is, conducting a SR demands skills and expertise, proper training and collaboration.
SRs can be conducted with or without inclusion of a meta-analysis. A meta-analysis is a statistical technique that summarizes/pools numerical data from similar individual research studies included in SRs to come to new statistical conclusions. Not all systematic reviews include meta-analyses, but all meta-analyses are conducted in SRs (example of a SR without meta-analysis). Simply put, it refers to a statistical approach for combining the data derived from a systematic review. Meta-analysis provides stronger statistical, conclusive evidence than the analysis resulting from any single study (e.g., SR with meta-analysis).
Umbrella reviews: Umbrella reviews are reviews of reviews. Other types of reviews synthesize original empirical studies. However, umbrella reviews are based on reviews and meta-analyses, and thus they provide a higher level of evidence. They are considered helpful in comparing two or more treatments or interventions for the same condition (example of an Umbrella review).
Usually, a general/narrative review could lead to a scoping review, and a scoping review could inform a SR. However, there is no hard and fast rule of requiring a general or scoping review to conduct a SR In fact, there should be a clear need/justification for doing a SR. There are various guidelines for determining the quality of reviews as well.
Guidelines/resources for conducting a systematic or scoping review
To learn more details about the process of conducting SRs, please consult the Cochrane website. This is an organization dedicated primarily to providing health-related SR guidelines and training. Cochrane handbooks include clear guidelines for doing SRs and meta-analyses. In addition, the JBI manual for evidence synthesis is also considered a standard for conducting scoping and systematic reviews. These sites have learning materials and STEP by STEP guidelines/Training programs for conducting systematic and scoping reviews. In addition, there is a course on SRs and meta-analyses by Coursera. There is some software that can be used to facilitate the process of conducting systematic/scoping review, such as Covidence, Review Manager (Rev Man) and others. I find Covidence very helpful in managing SRs. However, Revman is freely available and I have used it for meta-analyses, too; Covidence carries a subscription charge. Haseeb Irfanullah, one of the AuthorAID Stewards, has shared his experience and tips for making review articles interesting in his blog.
Please go through all available resources and gain expertise. Only then start conducting your review based on identified gaps. Happy writing review papers!!
Buna Bhandari (PhD: UNSW, Australia) is an assistant professor at the Central Department of Public Health, Tribhuvan University Institute of Medicine, Nepal. She has been a guest facilitator of AuthorAID online courses since 2016 and an AuthorAID Steward since 2020.