Would you please briefly explain multiple imputation and why it can mitigate the missing preoperative MOX-FQ data, but not the missing follow-up MOX-FQ data?

If one of the covariates in the multiple regression analysis is missing, all data from that patient is excluded from the analysis.

And these patients, 75 patients, they had a missing baseline Manchester-Oxford score for a purely administrative reason, because this PROM was introduced after the registry was started.

So in the beginning of the registry, we didn't have that PROM for all patients because we used AOFAS score as the main outcome.

And this means that we can judge it as missing at random.

It's not missing in selected patients.

And then you can use multiple imputation to impute a reasonable value for that missing data using other information about the same patient.

And if you just give a little more clarity on this, I realize that you considered age as a covariate but would you give greater clarity on how your outcome analysis, in your outcome analysis, you isolated the influence of psychological distress from younger age?