Table 3.
Comparison of validation and external cohort variables with those of the development cohort.
| Variable | Development | Validation | External |
|---|---|---|---|
| Sample number | 128 | 330 | 97 |
| Outcome prevalence (n (%)) | 25 (20) | 76 (23) | 28 (29) |
| General clinical characteristics and laboratory variables | |||
| Age (years) | 65 ± 19 | 67 ± 16 | 71 ± 18 * |
| Female sex (n (%)) | 55 (43) | 134 (41) | 30 (31) |
| Obesity (n (%)) | 19 (15) | 80 (24) | 25 (26) |
| Diabetes (n (%)) | 15 (12) | 46 (14) | 21 (22) |
| Hypertension (n (%)) | 56 (44) | 183 (55) | 65 (67) ** |
| Past or active smoking (n (%)) | 12 (9.4) | 52 (16) | 15 (15) |
| Congestive heart failure (n (%)) | 12 (9.4) | 20 (6.1) | 9 (9.3) |
| Coronary artery disease (n (%)) | 15 (12) | 35 (11) | 9 (9.3) |
| Stroke or TIA (n (%)) | 10 (7.8) | 29 (8.8) | 13 (13) |
| 30-day death (n (%)) | 23 (18) | 64 (19) | 27 (28) |
| Oral intubation (n (%)]) | 5 (4.0) | 23 (7.0) | 3 (3.1) |
| Plasma creatinine (mg/dl) | 0.9 (0.7–1.1) | 0.9 (0.8–1.2) | 1.0 (0.8–1.3) |
| Blood urea nitrogen (mg/dl) | 16 (11–23) | 18 (13–26) | 21 (16–28) *** |
| eGFR (ml/min/1.73 m2) | 78 ± 28 | 73 ± 29 | 69 ± 27 |
| C-reactive protein (mg/dl) | 3.1 (0.8–7.0) | 5.1 (1.7–12.1) *** | 5.1 (1.6–11.1)* |
| Procalcitonin (ng/mL) | 0.02 (0.01–0.08) | 0.04 (0.01–0.13) | 0.05 (0.01–0.14) |
| Vital signs | |||
| Glasgow Coma Scale | 14.9 ± 0.6 | 14.8 ± 1.1 | 14.7 ± 1.1 |
| Body temperature (°C) | 36.8 ± 0.8 | 37.1 ± 1.0 ** | 37.3 ± 0.9 *** |
| Systolic blood pressure (mm Hg) | 132 ± 19 | 137 ± 22 | 135 ± 19 |
| Mean arterial pressure (mm Hg) | 96 ± 13 | 98 ± 14 | 96 ± 11 |
| Heart rate (beats per minute) | 82 ± 15 | 85 ± 17 | 83 ± 14 |
| Respiratory rate (acts per minute) | 20 ± 6 | 20 ± 6 | 20 ± 6 |
| Arterial oxygen saturation (%) | 96 (94–98) | 95 (91–97) *** | 95 (93–97) * |
| Variables used for calculating the aPNea score | |||
| Hospitalization COVID-independent (n (%)) | 23 (18) | 39 (12) | 21 (22) |
| Arterial O2 partial pressure (mm Hg) | 75 (64–84) | 67 (59–78) *** | 68 (61–80) * |
| Horowitz index | 332 ± 86 | 290 ± 96 *** | 279 ± 97 *** |
| Positive walk test (n (%)) | 16 (13) | 31 (9.4) | 1 (1.0) ** |
| Calculated A-a O2 gradient (mm Hg) | 40 (28–55) | 47 (33–93) ** | 53 (39–150) *** |
| Expected A-a O2 gradient (mm Hg) | 21 (17–24) | 22 (18–24) | 24 (19–25) * |
| Lung involvement ≥ 50% on X-ray (%) | 32 (25) | 113 (35) | 39 (42) |
| Lung involvement ≥ 50% on US (%) | 32 (25) | 113 (35) | 39 (42) |
| Lung involvement ≥ 50% on CT (%) | 32 (25) | 113 (35) | 39 (42) |
| D-dimer (ng/mL FEU) | 656 (423–1155) | 709 (453–1356) | 814 (559–1569) |
| Prognostic scores | |||
| aPNea score | 4 (1–8) | 7 (2–12)* | 6 (2–11) |
| NEWS2 score | 1 (1–3) | 3 (1–4)* | 2 (1–4) |
Bonferroni-corrected * p < 0.050; ** p < 0.010; *** p < 0.001 vs. the development cohort; TIA, transitory ischemic attack; SBP, systolic blood pressure; eGFR, estimated glomerular filtration rate; US, ultrasound; CT, computerized tomography; FEU, fibrinogen equivalent units.