Conclusions

Systematic reviews and meta-analyses formally appraise and, where appropriate, combine the results of trials that have considered similar questions. The methods aim to reduce the potential influence of both random error and bias, and to establish more reliably whether there is a real difference between treatments. This can be done in a relatively short period of time compared to initiating new prospective studies. Given that in many cases there may already be sufficient evidence to resolve therapeutic issues, it could be considered both scientifically inappropriate and unethical not to carry out a formal systematic review of existing information, before embarking upon a new prospective study. However, a systematic review is not a quick fix and if done properly, is likely to take many months to complete.

The overall results of a well conducted systematic review should provide the most comprehensive and least biased appraisal of a therapeutic question, the meta-analysis element providing an estimate of the average effect of a treatment. This is probably the most reliable and best available evidence to guide treatment policy for future patients. Further information from subgroup analyses and knowledge of underlying risks for different types of patient may also guide decision making. However, when interpreting the results of any systematic review, it is important to remember that neither individual trials nor meta-analyses can provide prescriptions on how individual patients should be treated. The estimated average treatment effect is, however, an essential piece of information to be considered by both patients and doctors alongside other factors, such as toxicity, cost, patient preference and quality of life, when making individual treatment choices.

In summary, there are very many advantages of systematic review and meta-analysis. The process enables us to take a more global perspective as the value of any individual study depends on how well it fits with or expands other work as well as its own merits [66]. Not only are the power and precision of estimated benefit and risk improved and the influence of bias limited by adopting appropriate methodology, but we can also assess whether findings are consistent and can be generalized across different populations, settings and treatment variations [2]. The process can be thought of as being like assembling a well-worn jigsaw puzzle, whereby all the individual pieces have to be found, checked that they are for the correct puzzle and where necessary repaired before all the individual pieces can be put together. The resulting big picture should provide us with far greater insight than any single piece (no matter how new and shiny) in its own right. Despite the many advantages, it is important that systematic reviews and meta-analyses are neither used nor interpreted blindly. We must not be fooled by the apparent authority of large numbers but critically appraise the quality and value of the evidence on which any systematic review or meta-analysis is based and ensure that the methodology used is appropriate. In this way we can use the process both to obtain valuable answers to unresolved questions that can be put into clinical practice and to establish baselines and the basis on which future research can be built.

0 0

Post a comment