Skip to main content
. 2015 Feb 27;22(3):649–658. doi: 10.1093/jamia/ocu022

Table 3:

Coverage of the CORE Subset for a New Dataset Estimated by “Leave-One-Out” Analysis

Dataset left out CORE Subset size (based on the other seven datasets) Coverage of frequently used concepts in dataset left out (%) Usage coverage % Frequently used concepts in dataset left out that are missing from CORE Subset
with CORE parent (percentage of missing concepts) with CORE child (percentage of missing concepts) with either CORE parent or child (percentage of missing concepts)
KP 5733 1745 (79.8) 79.6 246 (55.5) 106 (23.9) 287 (64.8)
VA 5856 1222 (79.2) 87.6 170 (53.1) 67 (20.9) 195 (60.9)
MA 5286 1977 (69.0) 77.8 449 (50.4) 82 (9.2) 487 (54.7)
NU 5508 2050 (75.4) 85.4 290 (43.4) 127 (19.0) 351 (52.5)
HA 5313 1232 (58.8) 66.8 419 (48.6) 87 (10.1) 456 (52.8)
IH 6091 899 (91.4) 92.6 37 (43.5) 23 (27.1) 52 (61.2)
RI 6112 660 (91.2) 86.8 29 (45.3) 21 (32.8) 38 (59.4)
BI 6167 374 (97.7) 93.1 2 (22.2) 2 (22.2) 3 (33.3)
Mean 5758 1270 (80.3) 83.7 205 (45.3) 64 (20.7) 234 (55.0)

KP, Kaiser Permanente; VA, Veterans Administration; MA, Mayo Clinic; NU, University of Nebraska Medical Center; HA, Hong Kong Hospital Authority; IH, Intermountain Healthcare; RI, Regenstrief Institute; BI, Beth Israel Deaconess Medical Center.