Table 1 Data description for the Welldoc dataset, showing the number of instances and patient counts across subsets by diabetes type (Type 1 Diabetes [T1D] and Type 2 Diabetes [T2D]), age groups, and gender
| Â | Records (Patients) | Type 1 Diabetes (T1D) | Type 2 Diabetes (T2D) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Complete dataset | Dataset | Training | Internal | Temporal | Held-Out | Complete dataset | Dataset | Training | Internal | Temporal | Held-out | ||
Patients | 15,961,183 (592) | 7,788,836 (291) | 779,111 (290) | 560,184 (257) | 61,979 (256) | 68,762 (257) | 88,186 (33) | 8,172,347 (301) | 817,125 (301) | 597,975 274) | 66,257 (273) | 73,293 (273) | 79,600 (27) |
Age group | |||||||||||||
18–39 years old | 3,425,955 (129) | 2,901,601 (109) | 290,615 (108) | 209,054 (94) | 23,321 (94) | 25,837 (94) | 32,403 (14) | 524,354 (20) | 52,077 (20) | 39,252 (19) | 4,440 (19) | 4908 (19) | 3,477 (1) |
40–64 years old | 8,040,123 (299) | 3,249,786 (123) | 324,649 (123) | 231,576 (108) | 25,408 (107) | 28,412 (108) | 39,253 (15) | 4,790,337 (176) | 478,441 (176) | 347,663 (158) | 38,481 (158) | 42,483 (158) | 49,814 (18) |
65+ years old | 4,495,105 (164) | 1,637,449 (59) | 163,847 (59) | 119,554 (55) | 13,250 (55) | 14,513 (55) | 16,530 (4) | 2,857,656 (105) | 286,607 (105) | 211060 (97) | 23,336 (96) | 25902 (96) | 26,309 (8) |
Gender group | |||||||||||||
Female | 7,133,594 (283) | 4,150,047 (165) | 415,869 (164) | 295,753 (144) | 32,910 (144) | 36,478 (144) | 50,728 (20) | 2,983,547 (118) | 297,557 (118) | 224,985 (109) | 24,768 (108) | 27,582 (108) | 20,222 (9) |
Male | 8,827,589 (309) | 3,638,789 (126) | 363,242 (126) | 264,431 (113) | 29,069 (112) | 32,284 (113) | 37,458 (13) | 5,188,800 (183) | 519,568 (183) | 372,990 (165) | 41,489 (165) | 45,711 (165) | 59,378 (18) |
