The Cochrane Schizophrenia Group

Day 3: (Data) Extraction…oo sounds painful

on July 5, 2013

Today is about extraction.  Extracting information from the studies identified through your search.  It sounds like it could be a painful process but actually once you get the hang of it it’s really quick and easy to sift, decide which studies aren’t relevant to your review.  the hardest part is drilling down in the one’s that you do deem relevant.  During the practical, Clive Adams asks a delegate how long it took him to work out that the study in question wasn’t relevant.  It took him 4mins!  This just shows that it doesn’t have to take ages, even when you’re only just learning about systematic review writing.  If this delegate wasn’t just learning and the study was on his profession he probably could have decided its relevance in 2mins.  It’s a bit like the new hazard perception test you get as part of your driving theory test…read, read, read, you spot something that doesn’t match your criteria, click the button like a ninja…next study.

Now you have to give your reasons for exclusion.

No pointer?  Just use an umbrella.  Statisticians are very resourceful people.

No pointer? Just use an umbrella. Statisticians are very resourceful people.

Systematic reviews are the best form of Evidence Based Research

Systematic reviews kick ass in the Evidence Based Research arena

Next up was  Sarah Lewington, Senior Research Fellow in the Clinical Trial Service Unit at the University of Oxford talking about meta-analysis.  To produce a worthwhile conclusion from a review the stats extracted from studies needs to be vast, which is hard, especially if most of the studies are small ones.  Small studies generally highlight something massively significant, but is it really significant if the number of participants was 5?  This is why Risk of Bias and lots of data needs to be incorporated into a review, otherwise you are in just as much danger of producing a biased conclusion based on zilch.  It’s also important to recognise that a systematic review by definition does not necessary incorporate statistics.  Systematic reviews do not need to include statistics to be valid.  If you don’t uncover any studies about your topic during the search means no statistics to review for instance.  This means the purpose of the review becomes highlighting the gaps in research.  Why is no one researching the value of supportive therapies (not CBT but more community support), for instance, especially when government policy favours it (issue raised in CSzG review)?  Sometimes the gaps are more significant than the results of statistics.

Related articles

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 )

Google photo

You are commenting using your Google 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 )

Connecting to %s

%d bloggers like this: