The cohort was selected as described in the case ascertainment section of the paper. Data was aggregated and described in the paper. No changes to the original data that we received (already deidentified) were made.