Rebecca McCracken, MSPH, ISMPP CMPP™, the Lockwood Group, Stamford, CT, USA; Kelly Mayer, BSc, the Lockwood Group, Stamford, CT, USA

This article is part of the Medical Publications 101 article series, which serves as an introduction to various topics that are relevant to medical publication professionals.

In the current publication landscape, medical writers and other publication professionals need to distinguish and identify the key elements that go into the various types of reviews and analyses. This article will provide an overview of review articles, meta-analyses, and computer modeling, such as in-silico trials, and how they apply to the publication landscape.

Review Articles and Meta-Analyses

With the number of scientific publications increasing every year, it can be a challenge for those in the healthcare community to obtain every bit of information available about a specific topic of interest. Review articles are standard elements of many publication plans and can provide a practical summary of recent scientific or clinical developments. Review articles allow a lot of information to be collated into one paper, but the type of review article is an important consideration when interpreting or applying information from the review. They can take a variety of forms, including narrative reviews, systematic literature reviews (SLRs), or systematic reviews with meta-analyses.

Narrative reviews summarize and comment, and provide a broad overview of a specific topic. Articles selected for inclusion are not selected systematically; inclusion is up to the authors. Narrative review articles, therefore, often focus on a small subset of studies with no quantitative analysis performed. These articles are mainly descriptive. While narrative reviews can be informative, they often include an element of selection bias and are usually considered, in the peer-reviewed journal community, a lower standard of evidence.

SLRs also summarize large amounts of information, but in an objective, systematic, or methodical way. These reviews follow a process that aims at removing bias and assessing the quality of all literature included. The process of creating SLRs typically starts with the development of a protocol that defines a specific research question, followed by a definition of search terms and databases to be used. [1] It specifies inclusion and exclusion criteria for studies to be considered, outlines how studies will be assessed, and defines the outcomes/data variables that will be pulled into a data extraction grid. [1] SLRs are descriptive by nature. They describe and summarize a large amount of information that has been identified systematically.

Often, SLRs include a meta-analysis component, when they are reviewing appropriate data. Meta-analyses use statistical techniques to synthesize data from several studies into a single quantitative estimate. [2,3] SLRs with meta-analysis follow a quantitative, formal, epidemiological study design that is used to systematically assess previous research studies to derive conclusions about that body of research. [3,4]

The Cochrane Collaboration ( is an international organization that promotes, supports, and disseminates SLRs and meta-analyses in the healthcare field. [1] It also provides guidance for conducting SLRs. Two additional resources essential to developing and implementing SLRs are the PRISMA-P and PRISMA guidelines and checklist. [5-8]

Table 1 below summarizes key differences in the types of reviews.

Table 1. Types of Reviews

Review articles play an important role in disseminating large amounts of data to healthcare providers, researchers, and industry. Knowing when and where narrative and SLRs are beneficial is important. For instance, while a narrative review may be suitable for an article regarding a new MOA, an SLR is more appropriate for an article covering compounds with established competitor products.

Be aware that many journals will not accept industry-sponsored narrative reviews, and some will not accept any narrative reviews. Given this, identifying the type of review to conduct is important, as is making sure it is appropriate for the target audience in order to avoid negative feedback from journals. Teams should look into target journals of interest before writing a review manuscript to assess what type of review articles the journals of interest accept and if they accept industry-sponsored reviews, as well as to determine any other guidelines regarding review articles that need to be followed.

Model-Based Analyses and In-Silico Trials

Model-based analyses are an element that is often included in publication planning. There are numerous types of model-based analyses, ranging from financial modeling (budget impact analysis) to regression analysis (which measures effect or association), and there are also computer models, such as in-silico trials.

Budget impact analyses are often part of a publication plan, as they are increasingly required by reimbursement authorities; they are often an essential part of a comprehensive economic assessment of new therapeutic interventions. [9] They address the expected changes in the spending for a healthcare system after the adoption of a new therapeutic intervention. They are also often used when assessing budget or resource planning. [9] The target audience for these publications consists of private and government health insurer plans, administrators of healthcare delivery systems, pharmacy benefit managers, etc. Several different variables go into these analyses, such as the time frame, the population characteristics and size, and disease-specific related costs. These analyses assess the financial feasibility of an intervention and, therefore, aid payers in making decisions. [10] Model-based analyses can help shape decisions for healthcare providers, administrators, and payers by estimating the financial feasibility of a strategy in a health service or system.

An in-silico clinical trial is an individual computer simulation used in the development or regulatory evaluation of a medicinal product, device, or intervention. [11] In-silico trials are computer models in which virtual patients are given virtual interventions, enabling observation of how the intervention performs and whether it produces the intended effect, without inducing adverse effects in real patients. [11] For these types of studies, a model is often built off of existing clinical trial data, and then extrapolates conclusions from the virtual patient population and size under the in-silico study.

Pharmaceutical products must undergo a development process that demonstrates efficacy and safety of the therapeutic intervention in healing or alleviating disease effects or disability to be commercially viable. Publications report on studies covering these preclinical and clinical testing phases. When clinical testing does not succeed, the economic losses can be disastrous for sponsors, and can delay or prevent potentially helpful therapeutic interventions from being studied further. In addition, when these studies do not succeed, sponsors usually only know that an intervention was not effective or not safe, without any further insight into the reason for the failure or guidance on how to improve the intervention’s performance. These potential interventions are often abandoned. Analysis through in-silico clinical trials may provide a better understanding of why an intervention failed and may be able to provide information that could be used to refine the product so it can successfully complete clinical trials.

In-silico clinical trials may also provide significant benefits over current preclinical practices. Unlike animals used as test subjects, virtual human models can be reused indefinitely, providing significant cost savings. Compared with trials in animals or a small sample of humans, in-silico trials might more effectively predict the behavior of the drug or device in large-scale trials, identifying side effects that were previously difficult or impossible to detect, and helping to prevent unsuitable candidates from progressing to the costly phase 3 trial stage.


Review articles and model-based analyses are important tools for healthcare providers and payers, as they provide information used in decision making. Publication teams should be familiar with these types of articles and analyses, as they are important components of a comprehensive publication plan. Teams should also keep up on best practices to ensure they are communicating these data correctly to their target audiences.

1. 2017 The Cochrane Collaboration. Accessed September 12, 2017.
2. Petticrew M, Roberts H. Systematic Reviews in the Social Sciences: A Practice Guide. Hoboken, NJ: Wiley-Blackwell; 2005.
3. Ressing M, Blettner M, Klug SJ. Systematic Literature Reviews and Meta-Analyses: Part 6 of a Series on Evaluation of Scientific Publications. Deutsches Ärzteblatt International. 2009;106(27):456-463.
4. Haidich AB. Meta-analysis in Medical Research. Hippokratia. 2010;14(Suppl 1):29-37.
5. Equator Network. Accessed September 12, 2017.
6. Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 2009;6(7):e1000097.
7. 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. PLoS Med. 2009;6(7):e1000100.
8. Moher D, Shamseer L, Clarke M, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 Statement. Syst Rev. 2015;4(1):1.
9. Sullivan SD, Mauskopf JA, Augustovski F, et al. Principles of Good Practice for Budget Impact Analysis II: Report of the ISPOR Task Force on Good Research Practices – Budget Impact Analysis. Value Health. 2014:17:5-14.
10. Silva MT, Nunes da Silva E, Pereira MG. Budget Impact Analysis. Epidemiol Serv Saude. 2017;26(2):421-424.
11. Viceconti M, Henney A, Morley-Fletcher E. In Silico Clinical Trials: How Computer Simulation Will Transform the Biomedical Industry. Int J Clin Trials. 2016;3(2):37-46.

%d bloggers like this: