Study conclusion: Angiography guided PCI was associated with significantly higher risk for MACE, However, IVI-guided PCI had comparable risk with CABG in DM patients.
But there's a methodological concern hiding in the study design that the authors didn't discuss:
Immortal time bias.
1/
Thank you for the wonderful explanation. Two (more!) naive questions about immortal time bias.
With re: to this particular analysis, where CABG pts were part of a registry but the PCI cohorts were part of an RCT, is it fair to say that:
1) Immortal time bias should favor PCI (if at all) here since there was no delay to treatment in the CABG arm? Presumably the pts had surgery exactly at the start of the follow up period.
2) Immortal time bias is unlikely to explain the difference bet. angio and IVI guided PCI because the assigned treatment was delivered immediately after randomization?
Perhaps there are other confounding factors to explain the early high event rate in the angio PCI arm? š¤