Baseline characteristics of the identified study cohort of type 2 diabetes, by full dataset and by training and validation cohortsdatasets
| Characteristics | Full dataset | Training dataset | Validation dataset |
| Patient count, n (%) | 139 626 | 111 748 (80) | 27 878 (20) |
| Age, years±SD | 63.0±12.5 | 63.0±12.5 | 63.1±12.6 |
| Gender (female), n (%) | 62 168 (44.5) | 49 686 (44.5) | 12 482 (44.8) |
| Number of general practices | 400 | 400 | 396 |
| Mean baseline HbA1c,% (mmol/mol±SD)* | 7.8 (62±22) | 7.8 (62±22) | 7.8 (62±22) |
| Cases with HbA1c data, n (%) | 122 294 (88) | 97 858 (88) | 24 436 (88) |
| Ethnicity, n (%) | |||
| White | 116 393 (83.4) | 93 157 (83.4) | 23 236 (83.3) |
| Non-White | 11 024 (7.9) | 8842 (7.9) | 2182 (7.8) |
| Unknown | 12 209 (8.7) | 9749 (8.7) | 2460 (8.8) |
| IMD quintiles, n (%) | |||
| Quintile 1 (affluent) | 26 930 (19.3) | 21 437 (19.2) | 5493 (19.7) |
| Quintile 2 | 29 534 (21.1) | 23 580 (21.1) | 5954 (21.4) |
| Quintile 3 | 29 539 (21.1) | 23 742 (21.2) | 5797 (20.8) |
| Quintile 4 | 27 883 (20.0) | 22 341 (20.0) | 5542 (19.9) |
| Quintile 5 (deprived) | 25 641 (18.4) | 20 571 (18.4) | 5070 (18.2) |
| Unknown | 99 (0.1) | 77 (0.1) | 22 (0.1) |
| Mean follow-up, years±SD | 7.6±4.8 | 7.6±4.7 | 7.6±4.8 |
| Region in England, n (%) | |||
| North East | 3202 (2.3) | 2560 (2.3) | 642 (2.3) |
| North West | 24 115 (17.3) | 19 412 (17.4) | 4703 (16.9) |
| Yorkshire and the Humber | 6036 (4.3) | 4799 (4.3) | 1237 (4.4) |
| East Midlands | 4553 (3.3) | 3660 (3.3) | 893 (3.2) |
| West Midlands | 17 838 (12.8) | 14 252 (12.8) | 3586 (12.9) |
| East of England | 14 540 (10.4) | 11 573 (10.4) | 2967 (10.6) |
| South West | 18 658 (13.4) | 14 888 (13.3) | 3770 (13.5) |
| South Central | 15 617 (11.2) | 12 441 (11.1) | 3176 (11.4) |
| London | 17 250 (12.4) | 13 829 (12.4) | 3421 (12.3) |
| South East Coast | 17 817 (12.8) | 14 334 (12.8) | 3483 (12.5) |
*Most recent measure within 1 year before or 6 months after index date.
HbA1c, glycated hemoglobin; IMD, Index of Multiple Deprivation.