As discussed in Part 1 of this series, identifying subgroups of patients with regard to their self-reported symptoms could be useful in clinical decision making. Further, we argued that latent variable modeling of patient reported outcomes (PRO’s) could provide advantages over “traditional” cluster analysis approaches. The objective of this Part 2 is to introduce this latent variable approach.
Take a moment to listen to the most recent video blog from our President and CEO, Dr. William Dudley.