hi Terry Shadyfeld for UAB School of Medicine Forest plots are commonly used in systematic reviews to graphically display the results of a meta analysis In this video we'll describe how to interpret the information that's given in a forest plot So this is a forest plot and forest plots have a lot of information in them So we'll walk through um each bit of this forest plot so that you'll understand what it's trying to tell you I'm going to focus on three zones this top sort of left zone that's the descriptive area about each study This right zone that's graphical nature of the information And then finally down here the bottom left the statistical components of the forest plot Now this is a fairly small forest plot There's some of them that will have 25 or 50 studies It just depends how many studies were included in this particular systematic review will make it into the forest plot But each study is represented by one line in this figure So study one here all the information about study one is located up here horizontally across this figure and extends over into this graphical area Also most of the time authors in forest plots will summarize the uh raw data from each of the individual studies In this case this these are this is a systematic review of randomized control trial So there's an intervention or control group Had it been a systematic review of a diagnostic test they would have given you numbers like false positives false negatives true positives true negatives Now each study has a result And typically in the forest plot the authors will present the final results of the information that they're interested in summarizing using metaanalytic techniques In this case the result was presented as a risk ratio So for study one the best estimate of the risk ratio of the intervention was 71 and over here in brackets is it's 95% conference interval This exact same information right here is presented over here graphically also Now usually studies are weighted differently in a meta analysis and most commonly the weight used in meta analysis is the inverse of the um variance of that particular study So on average bigger studies get greater weight in a meta analysis than smaller studies and that's on average not always the case Sometimes different weights are used but this gives you some sense of the strength or the weight of that particular study gives to the overall meta analysis Now as I mentioned earlier all this information here is also presented graphically So I personally like to look at the graphical representations uh of the information instead of these raw numbers over here But this box corresponds to the point estimate and the size of the box is related to the weight given to that particular study in the meta analysis So the bigger the box the greater the weight The smaller the box the smaller the weight And each of the lines emanating out of each box is a 95% confidence interval for that particular study So again this dot is the 33 These lines are 0.01 to 7.95 So that's how that's interpreted Now this vertical line that you see here is what's called the line of no effect where the intervention has no effect on the outcome In this particular case because this is a risk ratio this line of no effect is one ratio if the intervention is no better than control would have the same numerator or denominator So it would be a one This were a mean difference this number here would be zero And usually there's some labels put in at the bottom of a forest plot In this particular case the authors tell you the different risk ratios and whether it favors experimental or control group Just pay attention to how it's labeled um at the bottom of the forest plot you're trying to use Now because this is a meta analysis and what a meta analysis does is statistically combines all these studies into a more precise estimate of effect This information is given to you also in the forest plot in this red box here So we can see the weight adds up to 100% the event rates all add up And here is the metaanalytic summary of all these studies put together that the best estimate of the intervention is 64 and this is the confidence interval.36 to 1.15 That exact same information is represented here graphically And classically in a forest plot the metaanalytic answers presented as a diamond and where the peaks of the diamond um relate to this point estimate of 64 and the edges of the diamond are the confidence interval So over here this is.36 over here is 1.15 Usually they will give you some statistical test and tell you the p value of this finding and you can see it's not statistically significant but I already knew that because the edges of the diamond cross the line of no effect so I know it's not statistically significant but that information is given here for you And finally the last bit of information over here in this purple box is the test for heterogeneity The author here did two particular tests and they give you the results of their testing for heterogeneity Discussion of heterogeneity is beyond this video I have a separate video on what heterogeneity is and how do you assess for it There's lots of other um good information out on the internet also which can give you this uh same information Hope this video has helped you understand how to interpret a force plot Remember if you have any questions you can contact me through the course website or through the contact me section of my blog Have a great day