How a systematic review makes sense of several studies

Imagine one study reports that extending weekend hours increased library visits and another reports little change. Their findings may differ for several reasons. The places could have different starting hours, or the studies might count visits in different ways. A systematic review follows a stated method to find relevant studies and consider what they show together.

Start with the question and the search

Cochrane’s methods guide starts by defining a question and agreeing how to answer it. For the fictional library example, reviewers might compare extended opening with ordinary hours and define the outcome as visits per branch per week. That tells readers why some studies belong in the review and why others may fall outside its scope.

The review should describe where researchers looked, how they selected studies, and what they extracted from each one. If the search is limited to particular languages or dates, those limits help explain what evidence is represented. A study that falls outside the eligibility criteria may still be useful for a different question.

Compare how the studies were done

One library study might compare otherwise similar branches with and without extended hours. Another might compare visits in the same branch before and after the change. That second approach could pick up changes in a new school term or a nearby branch closing. These are invented examples of why researchers consider the design and possible sources of bias alongside the results.

The studies might also measure related but different outcomes, such as visits to the building and borrowing transactions. Read the review’s study table to see who or what was studied, how the outcomes were recorded and for how long.

Look at whether results were combined

When the questions, measurements and designs are sufficiently compatible, reviewers may use a meta-analysis to calculate a combined estimate. When they differ in important ways, a structured account of the separate findings may be more informative. The Cochrane Handbook discusses arranging studies for a synthesis that respects their characteristics.

If there is a combined estimate, look at which studies contributed and how uncertain the result is. Then check whether it applies to the kind of place or population you care about. A review can provide a careful answer about the studies it found and leave room for further work on a different setting. Its method gives you a way to see where that boundary lies.