The effect estimates obtained using mixed effects models tend to be considerably larger than those obtained using marginal models but, since the standard errors in mixed effect models are also larger, the significance levels obtained are often similar in both types of models. Interestingly, in our dataset we found that the results of GEE and mixed effects models had these expected differences, but the WLW and frailty models provided very similar effect estimates. Close agreement between WLW and frailty models in our data may reflect small correlation between the incident detections of different types of HPV, whereas the large differences observed between the GEE and mixed effects model estimates likely reflects strong correlations between the cross-sectional detection of the same HPV types over serial semi-annual visits (i.e., the high frequency of persistent and recurrent infections).
