I want to check whether there is a significant difference between the CD4 cells at three different time points after virus infection. I can see the increase in CD4 population, but how do I know whether the increase is statistically significant? I have done this study in mice and at each time point I had 3-6 mice.
With multiple time points for each mouse, you want some sort of multilevel model. Which type depends on the nature of CD4. If, as I suspect, CD4 is always a fairly large number, then a linear multilevel model should be fine. If CD4 is a relatively small number, you may want to try a Poisson or negative binomial multilevel model.
MLMs deal well with missing data, provided that missingness is not informative.