1
$\begingroup$

I have seen that in RCT’s with very long follow up times the power of the study can be impacted by various factors in Herbet, Kasza, and Bø, 2018, and I was wondering if anyone has heard of comparing an RCT that has a long follow up time with a retrospective cohort study which could be designed to decrease loss of information due to follow up issues (I’m not talking about death/non- adherence or anything that could be associated with the trial, but random events such as moving out of a study population etc).

Would it be interesting to compare to see if the results were different?

Herbert, R. D., Kosza, J., and Bø, K. (2018). Analysis of randomised trials with long-term follow-up. BMC Medical Research Methodology. 18(48).

Improve this question
$\endgroup$
1
  • 1
    $\begingroup$ You might want to look for articles by M. Zelen about "double randomized consent" trials where participants are randomized to something like an RCT, and to something like a cohort design, and where differential drop out accounts for some of the selection bias in who agrees to participate in an RCT. For a review see Altman, D. G., Whitehead, J., et al. (1995). Randomised consent designs in cancer clinical trials. European Journal of Cancer, 31(12), 1934-1944. $\endgroup$
    – Alexis
    May 14, 2021 at 13:21

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.