Table 1 Demographic distribution of UCSF and Vanderbilt cohorts. We identified women with preterm and not preterm deliveries at UCSF and Vanderbilt using similar ascertainment (the “Methods” section). For each woman, we predicted the earliest delivery in their EHR. We report age at delivery (patient age) as mean with standard deviation (SD) in parenthesis and self- or third-party-reported race for both cohorts as the count and the column-wise proportion in parenthesis. The T-tests and chi-squared tests of independence were used to compare distributions stratified by delivery label

From: Dense phenotyping from electronic health records enables machine learning-based prediction of preterm birth