Table 1 Characteristics of all trial participants and prediction model-building cohort.
From: Predicting radiocephalic arteriovenous fistula success with machine learning
Baseline Characteristics | All Trial Participants | Predictive Modeling |
|---|---|---|
N = 914 | N = 591 | |
Age (Years) | 57 (13) | 57 (13) |
Sex (Female) | 203 (22%) | 130 (22%) |
Race | ||
White | 619 (68%) | 382 (65%) |
African American | 219 (24%) | 149 (25%) |
Other | 76 (8.3%) | 60 (10%) |
Hispanic | 156 (17%) | 99 (17%) |
BMI | 31 [26,37] | 31 [26,37] |
Smoking Status | ||
Current | 131 (14%) | 91 (15%) |
Former | 403 (44%) | 257 (43%) |
Never | 380 (42%) | 243 (41%) |
Medical History | ||
Diabetes | 580 (63%) | 382 (65%) |
Hypertension | 885 (97%) | 573 (97%) |
Heart Failure | 252 (28%) | 166 (28%) |
Coronary Artery Disease | 260 (28%) | 172 (29%) |
Peripheral Artery Disease | 95 (10%) | 47 (8.0%) |
Cerebrovascular Disease | 129 (14%) | 79 (13%) |
Any Antithrombotic | 499 (55%) | 336 (57%) |
Statin Use | 499 (55%) | 312 (53%) |
Prior Renal Transplant | 37 (4.0%) | 26 (4.4%) |
Prevalent HD | 409 (45%) | 326 (55%) |
Duration of Prior HD (Months) | 9 (5,19) | 9 (5,19) |
Current or Prior CVC | 444 (49%) | 343 (58%) |
Location of AVF | ||
Wrist | 669 (73%) | 445 (75%) |
Forearm | 220 (24%) | 132 (22%) |
Snuffbox | 25 (2.7%) | 14 (2.4%) |
Intraop. Vein Diameter | ||
≥ 4.0 mm | 282 (31%) | 186 (31%) |
3.0–3.9 mm | 451 (49%) | 296 (50%) |
< 3.0 mm | 181 (20%) | 109 (18%) |
Intraop. Artery Diameter | ||
≥ 3.0 mm | 433 (48%) | 274 (47%) |
2.0–2.9 mm | 450 (49%) | 293 (50%) |
< 2.0 mm | 28 (3.1%) | 22 (3.7%) |
Anesthesia | ||
General/Local | 205 (22%) | 124 (21%) |
Regional | 709 (78%) | 467 (79%) |
Site Enrollment Volume | ||
Lower (≤ 20) | 316 (35%) | 197 (33%) |
Mid (21–49) | 303 (33%) | 181 (31%) |
Upper (≥ 50) | 295 (32%) | 213 (36%) |
