I am still confused how if we don't know the outcome of the dropped out subjects? For example, 100 subjects in control group vs 100 subjects in experimental group are followed for survival after 5 years. If there are 10 dropped out subjects in experimental group and (for the sake of simplicity) the rest of them is alive, how should we calculate the survival rate of experimental group? PP analysis: 90/90 = 100% ITT analysis: 90/100? And how about if I study their mortality rate? PP analysis: 0/90 = 0% ITT analysis: 10/100 = 10% (we count the dropped out subjects as "failure") or 0/100 (because we only include the dropped out as they were randomized, without any outcome)? Thank you
This is probably way too late of a response to matter, but looking at the worst case scenario can help you gauge whether the amount lost to follow up will skew the results significantly. For example, in ITT, you assume those lost to follow-up in the treatment group died, and everyone lost in the control group survived. This will give a more conservative estimate of effect since it will make your treatment look worse. However, with PPA, as you showed above, the mortality rate doesn't look as bad (0% vs 10%). Without 100% followup, we'll never know with certainty what happened to those people. This is why researchers tend to prefer ITT, since even when considering the worst case scenario, if the results still look good, you can be quite certain that the treatment is effective. Reality likely falls somewhere in between the ITT and PPA results, but using conservative statistics (ITT) makes the final conclusion, if it's a good one, more trustworthy.
Thanks buddy
I am still confused how if we don't know the outcome of the dropped out subjects? For example, 100 subjects in control group vs 100 subjects in experimental group are followed for survival after 5 years. If there are 10 dropped out subjects in experimental group and (for the sake of simplicity) the rest of them is alive, how should we calculate the survival rate of experimental group?
PP analysis: 90/90 = 100%
ITT analysis: 90/100?
And how about if I study their mortality rate?
PP analysis: 0/90 = 0%
ITT analysis: 10/100 = 10% (we count the dropped out subjects as "failure") or 0/100 (because we only include the dropped out as they were randomized, without any outcome)?
Thank you
This is probably way too late of a response to matter, but looking at the worst case scenario can help you gauge whether the amount lost to follow up will skew the results significantly. For example, in ITT, you assume those lost to follow-up in the treatment group died, and everyone lost in the control group survived. This will give a more conservative estimate of effect since it will make your treatment look worse. However, with PPA, as you showed above, the mortality rate doesn't look as bad (0% vs 10%). Without 100% followup, we'll never know with certainty what happened to those people. This is why researchers tend to prefer ITT, since even when considering the worst case scenario, if the results still look good, you can be quite certain that the treatment is effective. Reality likely falls somewhere in between the ITT and PPA results, but using conservative statistics (ITT) makes the final conclusion, if it's a good one, more trustworthy.