Clinical Evidence aims to promote better treatment decisions by providing systematic reviews of the latest evidence about healthcare interventions. Crucially, we present that evidence in the context of clinical questions or decision points, and report only on those outcomes that matter to patients.
Interventions page
The interventions page for each systematic review presents the questions addressed and a list of the interventions covered, categorised according to whether they have been found to be effective or not for the clinical outcomes we have specified. We have developed these categories of effectiveness from one of the Cochrane Collaboration's first and most popular products, “A guide to effective care in pregnancy and childbirth”.[1]
The categories we now use are explained in the table below:
| Intervention | Icon | Description |
|---|---|---|
| Beneficial |  |
for which effectiveness has been demonstrated by clear evidence from systematic reviews, RCTs, or the best alternative source of information, and for which expectation of harms is small compared with the benefits. |
| Likely to be beneficial |  |
for which effectiveness is less well established than for those listed under “beneficial”. |
| Trade off between benefits and harms |  |
for which clinicians and patients should weigh up the beneficial and harmful effects according to individual circumstances and priorities. |
| Unknown effectiveness |  |
for which there are currently insufficient data or data of inadequate quality. |
| Unlikely to be beneficial |  |
for which lack of effectiveness is less well established than for those listed under “likely to be ineffective or harmful”. |
| Likely to be ineffective or harmful |  |
for which ineffectiveness or associated harm has been demonstrated by clear evidence. |
Fitting interventions into these categories is not always straightforward. For one thing, the categories represent a mix of several hierarchies: the size of benefit (or harm), the strength of evidence (randomised or observational data), the individual quality of studies, and the degree of uncertainty around the finding. Another issue is that much of the evidence that is most relevant to clinical decisions relates to comparisons between different therapeutic interventions in the absence of any comparisons with placebo or no intervention. Where necessary, we also indicate the comparisons (rather than the interventions alone) which we have categorised. A further challenge is that interventions may have been tested, or found to be effective, in only one group of people, such as those at high risk of an outcome, and this may limit the generalisability of our findings. Again, we have indicated this where possible. Most difficult of all has been trying to maintain consistency across different reviews as the volume and quality of evidence available, or likely to be available, can vary widely among different subject areas. We continue to work on refining the criteria for putting interventions under each category and always welcome feedback about this aspect of Clinical Evidence
Interventions that cannot be tested in an RCT for ethical or practical reasons are sometimes included in the categorisation table and are identified with an asterisk.
Benefits
There are many different types of study design, each of which has its own strengths and weaknesses. Whilst Clinical Evidence recognises that a variety of study designs may be informative in certain circumstances, we primarily report RCTs (and systematic reviews of RCTs) in the benefits section of our reviews. This is because this type of study design gives the best evidence of causality (that is, that a particular intervention is responsible for a specific effect) and is less subject to the effects of bias and confounding than other study designs. We report appropriate meta-analysis of RCTs when they are available. Only when RCTs are unlikely to be undertaken (for example, due to ethical reasons such as clear longstanding evidence of benefit of an intervention) do we include other levels of evidence in the benefits section.
Key point summaries
We recognise that readers may require different levels of information regarding an intervention or condition. Whilst we provide very detailed information about reviews, trials, and interventions, each of our systematic reviews also includes a key point summary which gives an “at a glance” overview of the key issues and findings of that review.
Negative findings
A surprisingly hard aspect to get right is the reporting of negative findings. Saying that there is no good evidence that a treatment works is not, of course, the same as saying that the treatment doesn't work. We recognise that to address this issue fully we need a better handle on the power of individual systematic reviews and trials to demonstrate statistically significant differences between groups, and better information on what constitutes clinically important differences in the major outcomes for each intervention. In the meantime, we hope that the text makes a clear distinction between where there is insufficient evidence to say whether a treatment is effective or not, and where there is clear evidence of ineffectiveness
Outcomes
Clinical Evidence focuses on clinical outcomes that matter to people with that particular condition, meaning those that patients themselves are aware of, such as symptom severity, quality of life, survival, disability, walking distance, and live birth rate. We are less interested in often proxy outcomes such as, for example, blood lipid concentrations, blood pressure, or ovulation rates. Each review includes a list of the main outcomes, and where possible describes how these are measured. We have for the moment decided not to address the vexed question of what constitutes a clinically important change in an outcome, but we would welcome suggestions on how to do this.
Effects, not effectiveness
A key aim of Clinical Evidence is to emphasise the important trade offs between advantages and disadvantages of different treatment options. We therefore talk about the effects of interventions, both positive and negative, rather than the effectiveness, and for each intervention we present, when available, data on benefits and harms under separate headings.
Harms
Information about harms is often more difficult to synthesize than information about benefits.[2] Most RCTs are designed to investigate benefits and are less reliable to report harms, especially rare harms. Many RCTs either fail to document harms or present the information in a form that is difficult to analyse or interpret. When drugs are licensed they may have been used clinically in only a few thousand people; the absence of documented harms is not strong evidence that harms will not be discovered in the years after licensing.
Clinical Evidence recognises that the evidence about harms is often weaker than that about benefits. In an attempt to correct for this potential bias, Clinical Evidence has altered the threshold for evidence to be included in the harms section. Much of the evidence for harms comes from observational studies ranging from prospective cohort studies to case reports, and these are included when the harm is serious or when there is good corroborating evidence that the harm can be attributed to the treatment.
Clinical Evidence is not, and cannot be, a source of all potential adverse effects for a particular intervention. Many textbooks and prescribing guides exist which are far more suitable sources for this type of information. However, recognising the importance of emerging information on potential harms, Clinical Evidence now includes important new drug safety alerts relating to existing interventions.
Drug names
Clinical Evidence has an international audience. Difficulties can arise when different names for the same drug are used in different parts of the world. We state the Recommended International Non-proprietary Name (rINN) and give only the generic or non-proprietary names of drugs rather than the brand names.
Information on cost
We have decided not to include information on the cost or cost effectiveness of interventions. This is not because we believe cost to be unimportant, but because the question of what constitutes good evidence on cost is much disputed and because costs vary greatly within and between countries, and over time. However, we believe that it will become increasingly untenable for clinicians to act without paying attention to the absolute cost or cost effectiveness of treatments. Future companion publications of Clinical Evidence may provide relevant information on such costs
Numerical data
Whenever possible, data (including statistical results) are presented in the same form as in the original studies. However, sometimes we have changed the units or type of information in an attempt to present the results in a systematic and easily interpretable form.
An international approach
Clinical Evidence takes an international approach to the evidence. This may mean including drugs that are not licensed in some countries. It may also mean keeping in mind the practicalities of treating people in countries with limited resources, by covering some interventions even if they have been superseded elsewhere
Competing interests
In line with the BMJ's policy,[3] our aim is not to try to eliminate conflicts of interest but to make them explicit so that readers can judge for themselves what influence, if any, these may have had on the contributors' interpretation of the evidence. We therefore ask all contributors (and peer reviewers) to let us know about any potential competing interests, and we append any that are declared to the end of the contribution. Where the contributor declares that they have no competing interests, we also record this. Our editorial team are also asked to declare any possible competing interests and these are listed on our website.
Update cycle
We update this website monthly, and produce twice yearly Handbook editions. We aim to update each systematic review in Clinical Evidence every 12 months, and we also provide updates on the site to let users know about important studies that are published after the search date of the review. With each update we aim to increase the coverage and include stronger information about the adverse effects of treatments.
Changes since the last update
Substantive changes to each review published are listed on our home page. These may include:
- Presentation of additional evidence that either confirms or alters the previous conclusions;
- Re-evaluation of the evidence which may or may not affect the previous categorisations;
- Correction of an important error.
In addition, any new interventions or questions added to the review since the last update are clearly marked.
Reference links to full text
Clinical Evidence references link to the full text on PubMed when available.
Email alerting service
If you wish to be notified by email about new reviews, updates, or corrections, you can register for our alerting service.
How to use the information on Clinical Evidence
The type of information contained in Clinical Evidence is necessary but not sufficient for the provision of effective, high quality health care. It is intended as an aid to clinical decision making, to be used in conjunction with other important sources of information. These other sources include estimates of people's baseline risk of a condition or outcome based on history, physical examination and clinical investigations; individual preferences; economic arguments; availability of treatments; and local expertise.
References
- Enkin M, Keirse M, Renfrew M, et al. A guide to effective care in pregnancy and childbirth. Oxford: Oxford University Press, 1998. Zadig, 2003.
- Derry S, Loke YK, Aronson JK. Incomplete evidence: the inadequacy of databases in tracing published adverse drug reactions in clinical trials. BMC Medical Research Methodology 2001;1:7. http://www.biomedcentral.com/1471-2288/1/7 (last accessed 13 October 2003).
- A Smith R. Beyond conflict of interest. BMJ 1998;317:219–292.