eGFR after Dialysis

[Samir]

Hi I have a query regarding the eGFR for patients after dialysis.
Basically we are running a trial in which the main outcome of interest is eGFR at 3 years for each patient.
The trial population is for patients with advanced CKD (i.e. stage 4-5 eGFR).
We have a maximum follow-up for each patient for 3 years from randomisation.
We want to estimate the patient GFR at 3 years post randomisation based on the 4 MDRD equation.
We know that the patients eGFR becomes somewhat invalid once the patient is on dialysis.
We are hence trying to understand what we should do in terms of data at 3 years if certain patients in the trial reach dialysis before the study endpoint which is at 3 years.
We have had various discussions around this and some of the clinicians we have spoken to have suggested on using either the last observation carried forward method or just imputing a flat value of 8.
So wanted to know if there have been a situation you may have faced similar to this and what you did?
And is the approach of last observation carried forward or imputing a flat value viable assumptions to make in this situation?
Many thanks in advance for any help with!

[SKM23435]

I suspect the answer to your question depends on what the trial is looking at.
The other point to note is that when you are on dialysis your native renal function continues to deteriate. Your eGFR is a measure of your own renal function plus the effect of dialysis.

Sorry I'm not a statistician so I can't add anything more sensible.

[George_UKRR]

Is this a UK based study?