Meta-analysis is a method which can be used to combine the results of two or more studies. But why should we want to do this?
In research, studies are conducted to try to answer questions about the clinical effectiveness of a treatment or intervention.
In some medical areas such as cancer and cardiovascular disease, several studies may have been conducted which have assessed the effectiveness of a particular treatment. These studies will have asked whether, for example, does treatment A provide significant benefits over treatment B in a particular group of people?
Although these studies have all assessed the same treatment comparison, the size of the benefit from these studies will vary either due to differences in study protocols or purely by chance. Some studies may find that treatment A is beneficial whilst others find it has no apparent effect or is even harmful. Therefore it can be useful to combine the results from these studies together to get a single answer about whether treatment A is more beneficial than treatment B.
Meta-analyses improve precision
Meta-analyses provide researchers with a single pooled result to answer whether treatment A is more beneficial than treatment B. This pooled result is usually more precise than the result from the individual studies. But why is this important?
Each of following studies has attempted to calculate the effectiveness of treatment A as compared to treatment B. However, it is likely that the results from these studies will vary either by differences in study protocols or purely by chance. Some studies may find that there is a large beneficial effect of treatment A as compared to treatment B. However, others may show that the treatment effect is a lot smaller or even that treatment A is harmful.
The precision with which each of these studies calculates the treatment effect depends on many factors, including the number of people in the study. Generally, as the number of people increases in a study the precision of the treatment effect will increase.
Therefore, by statistically combining all of the sample sizes together from the individual studies, the precision of our pooled result for the treatment effect can be improved.
Meta-analysis is an important and valuable tool for summarising data from multiple studies to provide researchers with a new pooled estimate of how much more beneficial a particular treatment is as compared to a different treatment.
Although meta-analysis does not involve human subjects or experimental animals directly, it is often considered an easy study that can be done with the minimum of effort and little attention is often paid to its design and implementation. A valid meta-analysis however, requires the same careful planning in the protocol stage as for any other research. The criterion for selecting which studies are included in the meta-analysis is important since the pooled result of the meta-analysis depends on which studies are included.
|Centre for Reviews and Dissemination.||The CRD undertakes reviews of research about the effects of interventions used In health and social care.|
|Cochrane library.||The Cochrane Library contains high-quality, independent evidence to inform healthcare decision-making. It includes reliable evidence from Cochrane and other systematic reviews.|
|Health Technology Assessment (HTA) Programme.||The HTA programme works to provide all those who make decisions in the NHS with high-quality information on the costs, effectiveness and broader impact of health care treatments and tests.|
|Steps in conducting a systematic review.||This RLO outlines the five fundamental steps to conducting a systematic review of health care research so as identify, select and critically appraise relevant research.|