Considerations for pooling real-world data as a comparator cohort to a single arm trial: a simulation study on assessment of heterogeneity