Critical appraisal

Nearly all biomedical research is published in reputable, peer-reviewed journals in the form of scientific articles. The ability to read, understand and critically appraise scientific literature is an important skill to develop. These skills are fundamental to the effective practice of evidence-based medicine (EBM). This article will guide you through the basic structure of scientific articles and how to read them systematically in order to glean relevant information.

Structure of a Scientific Article

A scientific article is usually divided into several sections – the abstract, introduction, methods, results, discussion and conclusions. Reading these sections in a step-wise manner allows the reader to gain a comprehensive understanding of the subject matter discussed.

Abstract – The abstract is usually found on the first page of the article and functions as a summary of the article. Often, the abstract is the part of an article which is available following a literature search (in PubMed for example). In the abstract, authors list the aims of the study, provide a short description of the study design and methodology employed and the main findings of the study and their importance. Reading the abstract of a paper will help you decide if the article is of relevance to you and if you would benefit from reading the rest of the article.

Introduction – This section should provide you with background information relating to the study. Ideally, the amount of information provided should allow a person unfamiliar with the field of research to understand the remainder of the article without much difficulty.

This section should also provide justifications for the study conducted, highlighting the origins of the research question and how the authors think their research benefits the field as a whole.

The aims and objectives of the study should be clearly delineated in the Introduction. This can be divided into primary and secondary endpoints. Primary endpoints are the main parameter by which the authors of a study will deem their work to be successful (a ‘benchmark’ for the outcome of subsequent experiments). Secondary endpoints are any additional information obtained or parameters study in the study.

Methods – In the methods section, the authors of the paper outline details of the study or experiments conducted to obtain their results. Here the reader can identify the type of study conducted, be it a randomised control trial, cohort study, case control study, or case series.

Good scientific articles will clearly describe certain elements of the methods (and you should look out for these when reading an article), including demographics of the population studied (size, gender, age, co-morbidities, etc), intervention used and method of data collection.

Depending on the above factors, a reader can determine if the authors made sufficient effort to account for any bias or confounding factors in their study. In laboratory based studies, the Methods section outlines the experiments conducted and reagents used. Scientific articles in basic science and biomedical science journals often go into sufficient detail about their methods to allow for a reader to be able to re-create or build upon the methodology discussed should they want to.

Close scrutiny of the Methods section of an article help provide a reader with a sense of weather the findings of the study can be deemed valid.

Results – Here authors present the results of the study or their experiments. Readers should be able to correlate the results presented with the earlier aims and methods outlined. Good scientific articles are those which demonstrate this link in a clear and accessible manner. Another important consideration when reading the results section is to identify if the authors reach their primary outcome measures and any secondary outcomes.

Some knowledge of statistical methods may be necessary to interpret the results presented in an article. At the simplest level, it is important to assess if the authors present their results as statistical significant (where the ‘p’ value is at least <0.05) – indicating that the results obtained are not likely to be due to statistical error and are indeed ‘true’. In addition, knowledge of concepts such as Numbers Needed to Treat (NNT), Number Needed to Harm, Odds Ratios and Relative Risk can help a reader understand most clinical research articles.

 

Discussion and Conclusions – In this section, authors place their findings in the context of the background provided earlier in the article. Authors should also use this section to discuss the strengths and limitations of the study and any recommendations as to future work that needs to be conducted to build upon or improve their findings.

In addition, this section of the article will be used to address issues such as the applicability of the results from the study, and how results from this study correlate with existing and ongoing research in the area – again to provide readers with a clearer understanding of the context in which the study has taken place.

Critical Appraisal

Critical appraisal is the process of systematically assessing and interpreting research studies. When appraising an article, it is important to ask 3 questions:

  1. Is the study valid?
  2. Are the results reliable?
  3. Can I apply the findings of this study to the population that I work with?

These websites list approaches and tools that you can use to critically appraising scientific literature:

Critical Appraisal Skills Programme (CASP) International Network – http://www.caspinternational.org/?o=1012

Scottish Intercollegiate Guidelines Network (SIGN) Critical Appraisal Notes and Checklists – http://www.sign.ac.uk/methodology/checklists.html

An excellent guide to reading research articles and critical appraisal can be found in ‘How to Read a Paper’ by Trisha Greenalgh. This book discussed the above skills in an easy to read and accessible format. With practice, critical appraisal of the article can be done whilst reading the article the first time around.

Like any skill, practice will greatly improve your ability to read, understand and critically appraise scientific articles. In addition to your own reading, journal clubs are a great way to read and discuss papers in the company of others, which can help hone your abilities further. All the best!

by Prasad P. Velu

TOP