Comparison of Methods that Combine Multiple Randomized Trials to Estimate Heterogeneous Treatment Effects