Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Demystifying systematic reviews

  • Comment

VOL: 97, ISSUE: 48, PAGE NO: 61

June Jones, MSc, RGN, BA, NDNCertEd, PGDip, is clinical nurse specialist, leg ulcers, Southport and Formby Primary Care NHS Trust

Andrea Nelson, PhD, BSc, RGN, is research fellow, Centre for Evidence-Based Nursing, Department of Health Studies, University of York

Nursing aims to deliver the best available care. This can be achieved most effectively through decisions that combine research findings, clinical expertise and patient preferences and values, as well as information on the cost and availability of resources. The challenge for clinicians is to integrate all these sources of information.

Nursing aims to deliver the best available care. This can be achieved most effectively through decisions that combine research findings, clinical expertise and patient preferences and values, as well as information on the cost and availability of resources. The challenge for clinicians is to integrate all these sources of information.

Evidence-based care is the new buzz phrase in practice, but finding relevant information and integrating it into the clinical context is a problem for many nurses. There is a vast amount of written evidence available via professional journals, guidelines, textbooks, the internet and commercial representatives, but much of this material is of dubious quality and may represent one opinion about a treatment, rather than an objective summary. It is also important to be able to weed out biased information so that decisions are made on the best available evidence.

Nurses make many types of decisions - for example, what information needs do people have before being transferred to a nursing home? What are the chances that this leg ulcer will heal? These are examples of questions of communication needs and of prognosis. Many questions asked by nurses are questions of effectiveness - that is, how do I prevent maceration around a heavily exuding burn? What is the best way to teach older people about their medications?

Questions of effectiveness are best answered by a research design called a randomised controlled trial (RCT). In this sort of research, people are randomly assigned to different treatments groups - for example, a study comparing preparation of surgical sites might allocate some people to the pre-operative shaving group and others to pre-operative clipping. Allocating people to the groups randomly helps ensure guarantees that any differences in the types of patients in the two groups of the study, whether known (for example age, sex, type of surgery) or unknown (for example type of microbiological flora on the skin, unknown risk factors for infection) are distributed evenly between the groups. This ensures that any difference in eventual outcome is likely to be due to the difference in interventions and not to the two groups being different at the start of the study.

Regardless of the type of question being answered - whether it covers aetiology, diagnosis, prognosis, treatment, prevention or information needs - an efficient way of answering the clinical question is to find a good quality summary of all the relevant evidence in an area. This is the purpose of a systematic review.

Definition
A systematic review uses a systematic approach to prepare a literature review. It aims to minimise bias by accessing not only the literature that is available easily or agrees with current practice but attempts to include all the good-quality literature on a given topic.

The design of a systematic review is similar to that of a primary piece of research but concerns itself with the re-analysis of research data already collected. Eligible research studies are viewed as a population to be systematically sampled and surveyed. Individual study characteristics and results are studied and summarised.

Occasionally, if appropriate, the data can be statistically analysed much like other quantitative data. By considering a number of studies one can effectively increase the amount of information collected. For example, when seven small studies examined the use of steroids in pre-term babies there was not enough information in any one of the trials to say whether the steroids helped babies survive. As the trials were all similar, however, it was possible to combine them, thus increasing the overall number of babies in the study. This process of combining the results of more than one study using statistical techniques is called meta-analysis - and it effectively increases the number of people in your studies. This results in a more precise estimate of treatment effect than can be obtained from any of the individual studies (DiCenso et al, 1998).

However, meta-analyses are not always appropriate in systematic reviews if the trials are not similar. A narrative review would be a more appropriate method for describing the effects of treatment if the treatments are dissimilar or there are different sorts of patients (Nelson, 1998). For example, in the systematic review of compression bandages for healing leg ulcers, Cullum et al (1999) found six studies that compared the healing rates with and without compression bandages. As all the bandage systems and patients in the trials were dissimilar these studies were not combined in a meta-analysis.

A systematic review can thus be defined as a review of a clearly formulated question that uses systematic and explicit methods to identify, select and critically appraise relevant research and to collect and analyse data from the studies that are included in the review (Mulrow and Oxman, 1997). What distinguishes systematic reviews from more traditional reviews or overviews of research are the following criteria:

- They attempt to identify all the relevant, robust research studies on a topic, including those that have not been published, as well as articles in any language;

- They systematically appraise the validity of each study;

- The use of statistical techniques, where appropriate (Cullum, 1999).

Appraising a systematic review
Before using the information contained in a systematic review the clinician needs to check whether it is relevant and reliable. A reliable systematic review will meet a number of quality criteria and there are a number of checklists available to help the reader decide whether a review is worth reading or not. A particularly useful checklist has been developed by the Critical Appraisal Programme (CASP) team at http//www.phru.org/casp/review/html (Box 1). This checklist can be illustrated with reference to the systematic review of skin grafting for venous leg ulcers in the Cochrane Database of Systematic Reviews (Jones and Nelson, 2001) (Box 2).

Where can you find systematic reviews?
Many reviews are published in summary form in paper journals and in full in electronic journals. Paper journals not only have the problem of limited space for articles but also become quickly out of date as new trials are published. Evidence-based journals are particularly useful, as they regularly abstract systematic reviews and present them as a half-page abstract alongside a comment from a clinician to put them into context - for example, evidence-based nursing at http://www.evidencebasednursing.com/

The Cochrane Library contains four databases. One of these, the Cochrane Database of Systematic Reviews (CDSR) contains the full text of systematic reviews prepared according to strict quality criteria. Cochrane reviews are published electronically and updated in the light of new evidence or comments from users. The Cochrane Library is available via the National Electronic Library for Health (NELH) free to NHS staff at http://www.cochrane.org/cochrane/revabstr/

mainindex.htm and in many hospital libraries. Abstracts of all the Cochrane reviews are available free of charge on the internet.

More information on systematic reviews may also be found from the NHS Centre for Reviews and Dissemination at http://www.york.ac.uk/inst/crd

Their site also points you to a database of Abstracts of Reviews of Effectiveness (DARE) at http://agatha.york.ac.uk/welcome.htm

Systematic reviews published in paper journals are referenced in electronic databases such as Medline and CINAHL, among others. You can find them by combining the term 'systematic review' with your topic. If you were looking for a paper summarising the best way to help patients to follow their medication schedules, you might use the search terms 'compliance' and 'medication' and 'systematic review' in the title. On Cinahl 1999-2001, this search yields 'Dodds-F; Rebair-Brown-A; Parsons-S. A systematic review of randomised controlled trials that attempt to identify interventions that improve patient compliance with prescribed antipsychotic medication. Clinical-Effectiveness-in-Nursing 2000 Jun; 4(2): 47-53.'

Conclusion
It is vital that nurses become familiar with the process of accessing and appraising systematic reviews. Likewise, it is important that those undertaking reviews use a systematic approach, describing the methods in order to determine if the findings have any use in everyday clinical practice.

  • Comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.