Mixed Models for longitudinal analysis will be a good choice, depending on
the choice to focus on single outcome or multiple outcomes.
The other option to create a single variable for outcome that combine all
available scores for each patient.
Your units of analysis will be patients and and these patients will form
clusters depending on how many times
they have been observed and how many times they have been operated(source of
errors). You will be able to observe the trends
for each Doctor over time without comparing their performance.
Hope you will check the possibility for there models.
On Tue, Nov 3, 2009 at 3:57 PM, - <[log in to unmask]> wrote:
> hello all,
> methodology question here:
> Someone has data on 100 patients doctor A has been treating and 150
> patients doctor B has been treating. There are scores on the patients
> regarding symptoms, complications etc and we want to assess the
> performance of doctor A from his 1st patient to the last, i.e. for
> operative time we want to know if it has increased or not during the
> time he has been seeing patients. The same for doctor B. We do not
> want to compare the performance of doctor A with that of doctor B. It
> sounds like a series with patient 1 at time point 0 etc. What
> statistical method would be used in this situation?
> Thank you all in advance
Center for Statistics