Systematic Reviews and Meta Analyses: What’s the difference?

Systematic reviews and Meta analyses are frequently mentioned terms in Evidence Based Healthcare and Evidence Based Medicine. However, many are unaware of the essential difference between the two. This post will attempt to shed some light on them.

Systematic Review:  A systematic review is a review of a clearly formulated question that uses systematic and explicit methods to identify, select, and critically appraise relevant research, and to collect and analyze data from the studies that are included in the review.

Meta Analysis: It refers to the statistical technique involved in extracting and combining data to produce a summary result.

A systematic review may or may not involve a meta analysis.

So what is a systematic review after all?

Let us look at an example to better understand the two terms:

Suppose you want to know whether administering steroids to a pregnant woman will improve the survival of her pre-term baby (due for delivery).

What will you do? Where will you look for the answer?

One way is to do a review of existing literature. However, there are some concerns with this approach.

We know that small studies often lack the power to detect meaningful differences between interventions. It is also known that all studies do not investigate the exact same question- they might be related, but different; or may have a different approach, etc. In short, it is difficult to look at a few studies and arrive at a decision.

Note: Studies published in top journals might be of better quality, but may be conducted in a different setting than yours…

This is where a systematic review comes in. It is like a large scale literature review, only it does the job in a systematic manner to minimize bias and other errors. It begins by asking a specific answerable research question. Then the reviewers establish rules for conducting the review.

This includes the exact procedure to be followed:

where all to search for articles (Google, PubMed, Embase, other databases, etc.);

which languages to search in (there are many fields in which much of the research has been done in non-English speaking areas- the results of such studies are published in languages other than English, so not including them would result in bias);

what search terms to use while searching;

what inclusion and exclusion criteria to have for studies (what study designs to include; all studies may not be relevant, as they may be addressing other concerns);

how the data will be extracted from studies (sometimes the authors may need to be contacted to obtain data/ clarifications)

how they will assess the quality of studies

who all will do the reviewing (the review team);

how long the review will take for completion;

how the results of studies will be consolidated/ pooled and analyzed (this is often complicated by the variations in studies);

how the findings of the review will be disseminated (to practitioners and the general public)

how often the review will be updated (with new evidence being generated daily, a review needs to be revised from time to time to stay current), etc.

[Please note that the steps outlined above are applicable only to systematic reviews. It is possible to conduct a review of literature without adhering to these procedures. However, such reviews are likely to suffer from the same problems mentioned earlier in this post.]

Thus, a systematic review looks at the findings of a large number of studies, and informs readers whether or not the original research question has been answered (insufficient or poor quality studies do not help, and are often cited as the reason for not finding an answer). If it has been answered, the reviewers provide guidelines for action based on that.

So what about meta analysis?

Remember the step where the data from various studies is to be extracted and analyzed? That is what a Meta Analysis is for- it is a statistical procedure that extracts and pools (combines) data to provide a summary result. A meta analysis consists of two essential steps: 1. Extract the data; 2. Calculating a pooled, average result across studies if it is appropriate to do so.

Can a systematic review be conducted without a meta analysis? Yes.

Is a meta analysis always undertaken as part of a systematic review? No.

The following article discusses systematic reviews and meta analyses:

Additional resources on systematic reviews and meta analysis are available here (this article may require a subscription/ fees)

The Cochrane Collaboration is probably the largest repository and publisher of systematic reviews related to health. It is a non-profit organization engaged in the creation and dissemination of evidence based healthcare.

Enhanced by Zemanta

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s