I have a large dataset (n=11804) that includes data from a cross sectional study conducted in 2013, 2014 and 2015 in patients in nursing homes...so its 'repeated cross sectional', right? However, some patients participated more than once, so in 2013 and 2014 or 2014 and 2015 or all 3 years...but others died/left the nursing home and new ones came in ass well... my problem is the repeated measures of the few participants that participated in more then one year... I want to analyse the dataset as one 'bulk' so not separated per year...what shall i do with data from people that participated more then once? Its categorical data so taking an 'average' is not really possible...

Thanks!