Medical review articles are increasingly popular and necessary. The need for review articles stems from the huge volume of scientific publications that confront researchers and clinicians. “Keeping up with literature” has become an insurmountable task for many. As the magnitude of reviews has increased, the importance of quality and reliability of the reviews also has increased. An honest review article describes its sources of information and methods of selection; clarifies the types and strength of evidence for key statements; and declares the provenance, findings and competing interests.
Kinds of medical review articles
This treatise discusses only medical review articles; it does not discuss reviews in other areas of science, such as basic molecular and cellular biology. The various kinds of medical review articles include traditional clinical reviews, also known as updates; systematic reviews; and meta-analyses. Non-quantitative systematic reviews comprehensively examine the medical literature, seeking to identify and synthesize all relevant information to formulate the best approach to diagnosis or treatment; meta-analyses (quantitative systematic reviews) seek to answer a focused clinical question, using rigorous statistical analysis of pooled research studies. The traditional clinical reviews are the most common kind of review and are the focus of this document. Such reviews can provide readers with powerful summaries and sound clinical guidance.
Often, clinical review articles are requested by journals or publishers, but authors themselves may choose to write a review. In either case, the topic should be chosen with the interests and specialty of the audience foremost in mind, represent an important aspect of the field of inquiry, and focus on a well-defined issue. Some examples are: The diagnosis and management of Barrett’s esophagus; the management of hepatocellular carcinoma in hepatitis C patients; advances in the management of congestive heart failure in the elderly.
Searching the literature
The literature search should first seek evidence-based reviews (systematic reviews and meta-analyses), well-designed randomized control trials, and guidelines for the diagnosis, treatment, or prevention of the disorder being discussed. Some sources of guidelines are the Agency for Healthcare Research and Quality (AHRQ), http://www.ahrq.gov/clinic; the American College of Physicians Journal Club (ASPJC), http://acpjc.acponline.org; and the Centre for Evidence Based Medicine (CEBM), http://www.cebm.net. Ideally, research methods, investigated databases, and key words should be described in the review’s final report. Different databases are used according to the topic analyzed. In most of the clinical topics, Medline should be surveyed, but several other databases can be appropriate, such as Embase, Cumulative Index of Nursing and Allied Health Literature (CINAHL), Google Scholar, ISI Proceedings, JSTOR Research, Scopus, and Web of Science. 
In recent years, the concept of patient-oriented evidence that matters (POEM) vs. disease-oriented evidence (DOE) has emerged. POEM deals with outcomes of importance to patients, such as changes in morbidity, mortality, or quality of life. DOE deals with surrogate end points, such as changes in laboratory values or other measures of response.  In traditional clinical reviews, POEM-type evidence is generally preferred over DOE-type evidence because of its direct clinical relevance. In literature searching, anecdotal reports and hearsay of conventional wisdom should be avoided. 
Evaluating the literature
Readers will need to know the quality of evidence supporting any recommendations made through a clinical review. Various rating systems for the levels of evidence have been put forth, but none has been uniformly used. Siwek, et al. use an A, B, C rating, with level A designating randomized controlled trials and meta-analyses; level B designates well-designed nonrandomized clinical trials, such as clinical cohort studies and case-controlled studies with non-biased selection of study participants; and level C designates consensus/expert opinion. Gulpinar and Guclu use a descending hierarchical rating system consisting of systematic review, randomized controlled trial, non-randomized experimental study, and case series. Those authors also emphasized that an essential part of the review process is differentiating good research from bad and leaning on the results of the better studies. Authors of reviews should evaluate the quality of articles by a strategy of their choice or design.
When authors of a review cite their own work they run the risk of creating a conflict of nterest. They must be cautious and objective, neither overstating nor understating the importance of their work, and they must strictly avoid self-promotion.
The discussion may address topics such as etiology, pathophysiology, clinical presentation, diagnostic evaluation, differential diagnosis, treatment goals, prognosis, and future directions. The discussion should acknowledge controversies, unresolved questions, recent developments, other viewpoints, and possible conflicts of interest or bias that may have affected the strength of the evidence presented.
Conclusions or summary
Whenever possible, the clinical review should contain summary recommendations based on the findings of the review. If that is not possible, the authors should explain the reason(s), such as, “The result of thorough, critical review of articles published on the subject of inquiry do not permit definitive recommendations to be made. Additional well-designed studies, with sufficient study-population size, are needed.”
Format of the review
A formal format often is not appropriate for clinical review articles, but a format suggested by Siwek et al, may be appropriate and useful:
- Introduction. Define the topic and purpose of the study.
- Methods. Indicate how the literature search was conducted and what major sources of evidence were used.
- Discussion. Cite the major achievements in the reviewed field, methodological problems, areas of debate, and questions for future research.
- Recapitulation, take-home message(s).
The references should be selective and not exhaustive. They should emphasize evidence-based recommendations, meta-analyses, and landmark articles. They should include the most current sources of information, but older, well-substantiated, time-honored studies may be valuable inclusions. In the current age of accelerated publication of scientific papers, authors of reviews should closely follow electronic lists of papers in press, as it can take months before these appear in scientific databases.
References for this treatise
1. Groves T. Fundamentals of good medical writing. bmjopen.bmj.com/site/about/resources/Fundamentals_of_Good_Medical_Writing.ppt
2. Siwek J, Gourlay ML, Slawson DC, et al. How to write an evidence-based clinical review article. Am Fam Physician. 2002 Jan 15;65(2):251-8.
3. Gülpınar Ö, Güçlü AG. How to write a review article? Turk J Urol. 2013 Sep;39(Suppl 1):44-8.
4. Pautasso M. Ten simple rules for writing a literature review. PLoS Comput Biol. 2013;9(7):e1003149.