I have collected data on scores for a sample population before, 1 year after and 6 years after an intervention.

All participants in the sample received the intervention. However, participants required the intervention for different reasons.

I want to find out whether there is a significant difference in the improvement between the pre-, 1-year, 6-year scores on the basis of reason for intervention.

I was wondering what test I should do, as some of the data is parametric and some is not. I am also finding conflicting answers in books and online about what to do in reference to missing data.

Some participants only have 1 year, some only 6 year scores etc. Am I able to compare means if different people are missing different data?

Thank you.