Skip to main content
. 2023 Jun 27;30(9):1532–1542. doi: 10.1093/jamia/ocad114

Table 1.

Demographic breakdown of retrospective and prospective cohorts

Demographic breakdown
Retrospective cohorts
Prospective cohorts
Diagnostic CBC Magnesium Metabolic CBC Magnesium Metabolic
N unique patients 13 362 11 771 13 410 18 982 5234 16 441
Age, mean (SD) 51.4 (23.7) 53.1 (24.2) 54.6 (21.4) 55.7 (21.4) 62.1 (17.8) 55.4 (21.1)
Sex, n (%) Female 7103 (53.2) 5432 (46.1) 6923 (51.6) 10 092 (53.2) 2425 (46.3) 8711 (53.0)
Male 6259 (46.8) 6338 (53.8) 6485 (48.4) 8884 (46.8) 2808 (53.6) 7724 (47.0)
Unknown 0 (0.0) 1 (0.0) 2 (0.0) 6 (0.0) 1 (0.0) 6 (0.0)
Race, n (%) White 6888 (51.5) 6038 (51.3) 6915 (51.6) 9009 (47.5) 2726 (52.1) 7636 (46.4)
Other 2870 (21.5) 2899 (24.6) 2680 (20.0) 4545 (23.9) 1067 (20.4) 4103 (25.0)
Asian 2444 (18.3) 1739 (14.8) 2600 (19.4) 3612 (19.0) 928 (17.7) 3125 (19.0)
Black 557 (4.2) 587 (5.0) 583 (4.3) 1060 (5.6) 327 (6.2) 884 (5.4)
Unknown 391 (2.9) 238 (2.0) 420 (3.1) 382 (2.0) 56 (1.1) 353 (2.1)
Pacific Islander 168 (1.3) 212 (1.8) 171 (1.3) 294 (1.5) 98 (1.9) 261 (1.6)
Native American 44 (0.3) 58 (0.5) 41 (0.3) 80 (0.4) 32 (0.6) 79 (0.5)

Note: Retrospective cohorts were sourced from Stanford’s clinical data warehouse (STARR). Prospective cohorts were collected in real-time through the EMR transactional database (Epic Chronicles) as diagnostic orders triggered model inferences between January 11 and February 15, 2023.

CBC: complete blood count.