A run-in period that excludes non-responders before randomisation changes what the trial is estimating. It is legitimate design and it must be stated in any summary.
Posting this under the heading it deserves: Immortal time bias in a claims-database study — one year on Everything below is what sits behind that.
I would like to disagree carefully with the settled view on Immortal time bias, and I would like to be argued out of it if the disagreement is bad.
The disagreement is about one step, not about the conclusion. If the step holds I withdraw it entirely.
Reporting bias within a paper — a result in a figure that is absent from the abstract — is checkable and is worth checking when the claim matters.
Second-hand on Immortal time bias, so weight it accordingly — someone whose method I trust told me this and I have not verified it myself.
The bit of Immortal time bias that nobody enjoys is that the answer changes depending on what you are trying to decide with it. Say what the decision is and the thread will converge.
The arithmetic on Immortal time bias is the easy part and it is where the errors are, which is an uncomfortable combination. Show your working and someone will catch it.
I had written a reply contradicting post #6 and deleted it. Here is what survived.
Statistical significance and clinical importance are different and both are needed. A significant difference below the minimal important difference is a real finding of no practical consequence.
Written in the hope of being told what I have missed.
My position on Immortal time bias is current rather than settled. I have revised it once already and I expect to again, so treat it accordingly.
Defending a design that is being criticised unfairly is as useful as criticising one that deserves it, and it happens far less often.
Bias towards the null and bias away from the null: different criticisms have different directions. Differential dropout might bias away from null; conservative statistical analysis might bias toward null.
That is my reading. Someone else read the same page differently and was reasonable.
I read post #9 twice before replying, because I had assumed the opposite.
Trying to state the Immortal time bias position in a way that someone who disagrees would recognise as fair, because I do not think the version in this thread passes that test.
Immortal time bias is worth one more sentence than it usually gets, and the sentence is the one about how the number was arrived at.
Picking up post #13: that is the part I would want checked first.
Multiple comparisons: if a paper reports many outcomes, the chance of a spurious association by random chance is real. Pre-specification of primary outcomes matters and secondary analyses are weaker evidence.
Second-hand, so weight it accordingly.
Where I would push back on the Immortal time bias consensus is the confidence, not the direction. The direction looks right. The confidence is borrowed.
Hold a trial to the standard something could actually have met. A criticism that no achievable design could have answered is a criticism of the field rather than of the paper.
Careful with the language on Immortal time bias. "Not detected" and "not present" are different findings and the first is a statement about the method.
Blinding, duration and comparator choice trade against each other. Improving one usually costs another, and a critique that ignores the trade is incomplete.
Adding a small correction to the Immortal time bias summary above rather than a disagreement with it. The substance holds; one of the figures is out by a factor that matters.
Everything in post #20 holds. The case it does not cover is the one I have.
What I want from this Immortal time bias thread is the list of things that would need to be true for the claim to hold. If we can write that list, we can check it.
Generalisability and validity are separate axes. A trial can be internally impeccable and still tell you nothing about the person asking.
Post #26 answers the question as asked. The question underneath it is different.
The version of Immortal time bias that I was taught turned out to be a teaching simplification. Useful, and not true in the way I had assumed it was.
Helpful, and short, which on this subject is harder than long.
Attrition is the failure mode most likely to invalidate a result and the least likely to be discussed. Differential attrition between arms is the specific thing to look for.
Reporting the observation and leaving the explanation open deliberately.
Where I part company with post #27, and it is a narrow parting.
On Immortal time bias the community has more anecdote than the confidence in this thread implies, and I include my own contribution in that.