Table 1B.

Characteristics of patients in the ulcerative colitis training set (n=600), stratified by physician-confirmed diagnosis of CD, UC and non-IBD

	Physician assigned diagnosis by chart review
Parameter	Ulcerative colitis (n=378)	Crohn’s disease (n=72)	Non-IBD (n=150)
Age at first diagnosis code [(mean (SD)) (in years)	43 (18)	37 (19)	52 (21)
Female (%)	53	53	60
Total ICD-9 codes for UC [mean (SD)]	23.0 (30)	8.1 (23)	1.8 (2)
Total ICD-9 codes for CD [mean (SD)]	3.5 (14)	36.8 (41)	0.1	(0)
Inpatient ICD-9 codes for UC [mean (SD)]	3.1 (7)	1.3 (3)	0.40	(1)
Narrative mentions of symptoms [mean (SD)]	20.8 (55)	30.6 (45)	31.4 (64)
Narrative mentions of “Crohn’s disease” [mean (SD)]	5.6 (23)	62.4 (76)	1.5 (8)
Narrative mentions of “ulcerative colitis” [mean (SD)]	32.9 (51)	9.0 (23)	1.1 (5)
Number of facts [mean (SD)] †	1602 (2466)	1640 (2035)	2564 (3641)
Duration of follow up (in days) [mean (SD)] ‡	4362 (2544)	4000 (2197)	4791 (2543)

Note: Ulcerative colitis training set ≥1 ICD-9 code for UC (556.x)

^†Number of facts refers to number of distinct encounters with the medical system and is a marker of healthcare utilization. For example, an office visit, a laboratory test, and a colonoscopy each contribute 1 fact.

^‡Duration of follow-up is the difference between first and most recent clinical encounter (office visit, X-ray, laboratory test) within the Partners Healthcare system.