This document discusses various topics related to outcomes research, including comparative effectiveness research, multilevel data analysis, investigating change over time, and estimating treatment effects from observational data. Comparative effectiveness research directly compares existing healthcare interventions to determine their benefits and harms. Multilevel analysis is helpful for comparing patient outcomes across hospitals while accounting for risk factors. Propensity score adjustment and other statistical techniques can be used to estimate causal treatment effects from observational data and reduce selection bias. Bayesian statistics are increasingly being used in areas like early-phase cancer trials.