Minimising bias

This is an expression you will see a lot in relation to systematic reviews. Because the aim is to provide reliable information, we need to do as much as possible to minimise the effects of anything that will cause the results to deviate from the truth. In other words, we need to minimise bias.

There are many possible sources of bias, which we will look at in various modules. For now, they can be grouped into two areas: bias arising from the studies included in the review, and bias arising from the way the review is done.

We'll come back to bias in the studies collected for a review shortly. The 'systematic' part of systematic reviews is all about minimising bias in the way the review is carried out. People have tried to identify the major sources of bias and error in reviews, and to design a system that will minimise them. The process involved in a systematic review goes like this:

• Define the question
• Look for all studies reliably addressing the question
• Sift the studies to select relevant ones
• Assess the quality of the studies
• Calculate results for each study (and combine them if appropriate)
• Interpret results

This general approach is followed in all systematic reviews, although the latter steps depend on finding some suitable studies.

The component studies of a review

There is no point taking a very systematic approach to preparing a review if the individual studies within it are not capable of answering the question your review poses. Just as we want to minimise bias in the review process, so we want to choose component studies that are likely to give us an answer close to the truth. There will always be the effects of chance and we can only overcome this by collecting as much evidence as possible, and taking account of it in the way we analyse and interpret results. But our only means of minimising the effects of bias, is to include only studies that are less likely to be biased.