Longitudinal Propensity Score Matching: A Demonstration of Counterfactual Conditions Adjusted for Longitudinal Clustering