Table 2.
Population characteristics for the study cohorts and the prospective validation set
|
Study population |
Prospective validation cohorts |
||||||
|---|---|---|---|---|---|---|---|
| Presenting to hospital |
Admitted to hospital |
Presenting to hospital (n=3326) | Admitted to hospital (n=1715) | ||||
| COVID-19 negative (n=114 957) | COVID-19 positive (n=437) | COVID-19 negative (n=71 927) | COVID-19 positive (n=383) | ||||
| Patients positive for COVID-19 | 0 | 437 | 0 | 383 | 107 | 91 | |
| Age, years | 60 (38) | 69 (26) | 65 (33) | 71 (26) | 56 (37) | 64 (34) | |
| Sex | |||||||
| Men | 53 570 (46·6%) | 246 (56·3%) | 34 381 (47·8%) | 211 (55·1%) | 1513 (45·5%) | 832 (48·5%) | |
| Women | 61 387 (53·4%) | 191 (43·7%) | 37 546 (52·2%) | 172 (44·9%) | 1813 (54·5%) | 883 (51·5%) | |
| Previous EHR encounter | 85 183 (74·1%) | 367 (84·0%) | 53 370 (74·2%) | 33 091 (86·4%) | 2671 (80·3%) | 1367 (79·7%) | |
| Ethnicity | |||||||
| White British | 76·0% | 65·4% | 78·5% | 68·4% | 66·3% | 68·2% | |
| Not stated | 11·8% | 17·4% | 11·0% | 16·2% | 19·5% | 20·5% | |
| Any other White background | 5·0% | 3·7% | 4·0% | 3·4% | 6·5% | 4·7% | |
| Pakistani | 1·3% | 1·1% | 1·1% | 1·0% | 1·2% | 1·0% | |
| Any other Asian background | 0·9% | 2·5% | 0·8% | 1·8% | 1·4% | 1·2% | |
| Indian or British Indian | 0·8% | 1·1% | 0·7% | 0·8% | 0·9% | 0·8% | |
| White Irish | 0·7% | 0·7% | 0·7% | 0·8% | 0·7% | 0·8% | |
| African | 0·6% | 3·0% | 0·6% | 2·9% | 0·6% | 0·8% | |
| Any other Black background | 0·3% | 0·9% | 0·3% | 0·5% | 0·5% | 0·3% | |
| Bangladeshi | 0·2% | 0·7% | 0·2% | 0·8% | 0·3% | 0·3% | |
| Chinese | 0·2% | 0·2% | 0·2% | 0·3% | 0·4% | 0·3% | |
| Any other ethnic group | 2·0% | 3·2% | 1·8% | 3·2% | 1·6% | 1·3% | |
| Patients positive for influenza | 484 (<0·1%) | 0 | 466 (<0·1%) | 0 | 0 | 0 | |
Data are n (%) or median (IQR). EHR=electronic health-care record.