Causal inference with missing data for continuous and survival outcomes in the practice of clinical trials and observational studies