Reading an Article
Readers tend to focus on the ‘take home message’ and fail to consider the limitations of studies. This is understandable. We are short on time and wish to be positive. But this approach has severe limitations. Many studies are either too preliminary or have limited applicability to routine surgical practice. On the other hand, it is impossible to read every potentially useful article in depth. That is why it is important to develop an information plan.
Seven Golden Rules
1. Have a clear objective
Safe surgical practice is not heavily dependent on being up to date with the latest information – it depends on access to reliable information. That is why you need an information plan.
The starting point is to frame a specific question. The PICO format identifies the most important components of clinical questions i.e. Patient, Intervention, Comparison, and Outcome. This a useful starting point. It is better to study specific topics in depth rather than just flicking through the recent issues of journals. Every specialty has a limited number of key topics that need to be considered in detail – concentrate on them. Browsing is useful, but it is no substitute for the thoughtful accumulation of evidence.
2. Search for information
All of the peer-reviewed journals of quality are indexed in PubMed. Keep searches simple because many articles are incorrectly classified by subheading. It may be better to play around with a few simple terms and filters, and be prepared to scan the results, than to perform one complex search. The Cochrane Collaboration is a database of systematic reviews that includes many useful accounts of interventions. Databases other than PubMed and the Cochrane Collaboration tend to be more useful for investigators than readers.
3. Scan articles
The title and key words come first. Next glance at the Abstract, the last paragraph of the Introduction, and the final paragraph of the Discussion. Some readers then target the Methods to ensure that they are sound. It can also be useful to scan the illustrations.
Next, look carefully at the authors and their previous publications (PubMed presents the names of the authors in hypertext and this links to their recent publications) – this is important because good articles do not usually exist in isolation, they evolve out of a research program. Consider rejecting:
- Articles that do not offer the prospect of altering your clinical practice.
- Studies that lack an adequate control group i.e. retrospective and single-group studies.
- Studies that lack large numbers of patients.
- Studies where the observed difference between the groups lacks clinical significance (regardless of the P values).
These steps should result in the identification of a limited number of articles suitable for evaluation in depth.
4. Study selected articles in depth
There are key issues associated with each of the sections in an original article.
The introduction should answer the following questions:
- What is the problem?
- Why was the study performed?
- What is the aim of the study?
The methods must contain enough detail for others to replicate the study. The specific issues are: the design of the study, the patients, the intervention, assessment of outcome, and the use of statistics (including the calculation of the estimated sample size).
The Results should be concise and contain enough data for an interested reader to draw their own conclusion. That is why data must be presented as absolute numbers, not as derivatives such as percentages.
Discussions usually start with summation comments and then: explore possible explanations for the main findings, compare the results with other relevant studies, discuss problems, state the limitations of the study, and explain the implications of the findings. The final paragraph should contain useful conclusions that link to the aim of the study.
5. Store information
It is impossible to keep a copy of every article that may be of future interest. You need to develop your own system. One approach is to keep summary notes with web links and reference lists. It does mean that some key references will be copied several times: but it is otherwise efficient and you will never lose sight of a useful reference. It is a small price to pay for a lack of clutter. Less neurotic types will be content with keeping piles of paper copies or hard drives full of digital copies. As search engines improve it becomes less important to store prime source material. Regardless of the approach that you adopt, there is a need to systematically store useful information so that it can be retrieved when needed.
6. Pay attention to guidelines
Seek out documents that are freely available on-line from the following groups that are interested in the standards of articles about patients:
CONSORT: The CONsolidated Standards Of Reporting Trials
STROBE: STrengthening the Reporting of OBservational Studies in Epidemiology
STARD: STAndards for the Reporting of Diagnostic accuracy studies
SQUIRE: Standards for QUality Improvement Reporting Excellence
PRISMA: Preferred Reporting Items For Systematic Reviews And Meta-Analyses
In particular, download the discussion paper at the CONSORT site. It is more informative than many biostatistical texts.
7. Develop an interest in critical evaluation
It is impossible to reliably interpret the surgical literature unless you acquire the skills to critically evaluate evidence. Besides books and websites, most surgical journals contain comments about published research; and some, like this journal, publish lengthy critical evaluations of key articles from the literature. The weekly medical journals – especially the British Medical Journal, which is available for free on-line – are useful sources of information.
A. J. Spillane, P. J. Crowe. The Role Of The Journal Club In Surgical Training. ANZ J Surg 1998; 68: 288-291.
Ellis C, Hall JL, Khalil AA, Hall JC. The evolution of methodological standards in surgical trials. ANZ J Surg 2005; 75: 874-7.
Maier RV. What the surgeon of tomorrow needs to know about evidence-based surgery. Arch Surg 2006; 141: 317-23.
Mahid SS, Hornung CA, Minor KS, Turina M, Galandiuk S. Systematic reviews and meta-analysis for the surgeon scientist. Br J Surg 2006; 93: 1315-24.
Trisha Greenhalgh. How to read a paper: the basics of evidence-based medicine. Third edition. Blackwell Publishing: Oxford, 2006.
Guyatt GH, Oxman AD, Vist GE et al. for the GRADE Working Group. Rating quality of evidence and strength of recommendations. BMJ 2008; 336: 924-6.
Campbell MJ, Swinscow TDV. Statistics at Square One. 11th edition. Wiley-Blackwell: Oxford; United Kingdom. 2009.