Table 3.
Multivariable regression analysis results of various outcomes for patients admitted to hospital as an emergency for COVID-19 or influenza
| Model outcome |
Model 1: length of stay in upper quartile (>10 days)*† N=3680 |
Model 2: died in hospital N=4953 |
Model 3: died in hospital/in 30 days of discharge N=4953 |
Model 4: died in hospital/in 90 days of discharge N=4953 |
Model 5: readmitted within 90 days of discharge* N=3680 |
Model 6: primary care activity* N=3680 |
| Logistic regression models OR (95% CI), p value | Negative binomial model IRR (95% CI), p value | |||||
| Cohort (influenza rc.) | ||||||
| COVID-19 | 3.81 (3.14 to 4.65), p<0.001 | 11.85 (8.58 to 16.86), p<0.001 | 11.01 (8.28 to 14.96), p<0.001 | 7.92 (6.20 to 10.25), p<0.001 | 1.07 (0.89 to 1.29), p=0.48 |
1.30 (1.23 to 1.37), p<0.001 |
| Age (18–54 rc.) | ||||||
| 55–64 | 1.63 (1.29 to 2.06), p<0.001 | 2.85 (2.07 to 3.97), p<0.001 | 3.07 (2.29 to 4.15), p<0.001 |
3.10 (2.34 to 4.15), p<0.001 |
1.48 (1.13 to 1.93), p=0.0045 |
1.29 (1.20 to 1.38), p<0.001 |
| 65–79 | 2.27 (1.84 to 2.81), p<0.001 | 6.90 (5.21 to 9.26), p<0.001 | 6.82 (5.27 to 8.94), p<0.001 |
6.67 (5.19 to 8.66), p<0.001 |
2.12 (1.68 to 2.69), p<0.001 |
1.47 (1.38 to 1.56), p<0.001 |
| 80+ | 4.40 (3.52 to 5.52), p<0.001 | 11.75 (8.86 to 15.82), p<0.001 | 12.26 (9.43 to 16.12), p<0.001 | 13.05 (10.11 to 17.02), p<0.001 | 2.75 (2.15 to 3.54), p<0.001 |
1.50 (1.41 to 1.61), p<0.001 |
| Sex (male rc.) | ||||||
| Female | 0.94 (0.80 to 1.10), p=0.43 | 0.83 (0.71 to 0.97), p=0.022 | 0.79 (0.68 to 0.92), p=0.002 |
0.79 (0.68 to 0.91), p=0.0012 |
0.86 (0.72 to 1.02), p=0.090 |
1.07 (1.02 to 1.12), p=0.010 |
| Ethnicity (white rc.) | ||||||
| Asian | 0.70 (0.59 to 0.85), p=0.00020 | 1.07 (0.89 to 1.28), p=0.47 | 1.08 (0.91 to 1.28), p=0.40 |
1.08 (0.91 to 1.28), p=0.38 |
0.96 (0.79 to 1.18), p=0.71 |
1.02 (0.96 to 1.08), p=0.49 |
| Black | 0.88 (0.68 to 1.12), p=0.30 | 1.11 (0.87 to 1.41), p=0.40 | 1.11 (0.88 to 1.40), p=0.38 |
1.08 (0.86 to 1.36), p=0.50 |
0.95 (0.72 to 1.25), p=0.72 |
1.01 (0.93 to 1.09), p=0.82 |
| Mixed | 0.76 (0.47 to 1.19), p=0.24 | 0.95 (0.57 to 1.54), p=0.85 | 0.98 (0.61 to 1.55), p=0.94 |
1.04 (0.65 to 1.61), p=0.87 |
0.57 (0.30 to 0.99), p=0.062 |
0.93 (0.81 to 1.07), p=0.28 |
| Other | 0.81 (0.61 to 1.07), p=0.14 | 1.10 (0.82 to 1.45), p=0.52 | 0.984 (0.748 to 1.29), p=0.91 | 0.98 (0.75 to 1.27), p=0.87 |
0.76 (0.55 to 1.05), p=0.10 |
0.92 (0.85 to 1.00), p=0.06 |
| IMD quintile (1—most deprived rc.) | ||||||
| 2 | 0.92 (0.74 to 1.15), p=0.47 | 1.04 (0.84 to 1.29), p=0.72 | 1.07 (0.87 to 1.32), p=0.54 |
1.06 (0.87 to 1.31), p=0.56 |
0.97 (0.76 to 1.23), p=0.77 |
0.96 (0.90 to 1.03), p=0.30 |
| 3 | 0.96 (0.76 to 1.20), p=0.71 | 0.76 (0.61 to 0.96), p=0.022 | 0.87 (0.70 to 1.08), p=0.21 |
0.88 (0.71 to 1.09), p=0.25 |
0.84 (0.65 to 1.08), p=0.17 |
0.96 (0.90 to 1.03), p=0.28 |
| 4 | 0.83 (0.63 to 1.10), p=0.19 | 0.85 (0.64 to 1.11), p=0.23 | 0.90 (0.69 to 1.17), p=0.43 |
0.84 (0.65 to 1.09) p=0.19 |
0.79 (0.58 to 1.08), p=0.14 |
0.97 (0.89 to 1.06), p=0.48 |
| 5—least deprived | 1.07 (0.74 to 1.54), p=0.72 | 0.92 (0.64 to 1.30), p=0.63 | 1.01 (0.72 to 1.42), p=0.94 |
1.04 (0.75 to 1.45), p=0.81 |
0.58 (0.36 to 0.90), p=0.018 |
0.96 (0.86 to 1.08), p=0.51 |
Admissions were to hospitals in Northwest London, between 1 February 2020 and 2 November 2020 for the COVID-19 cohort and between 1 January 2018 and 31 December 2019 for the influenza cohort. Regression model output is shown as ORs for binary outcomes, and IRRs for primary care activity, with 95% CI and p value. The reference category for each covariate is denoted by rc. Ratios for all terms in the regression models are reported in this table. No interaction terms were included.
*Readmissions, length of stay and primary care activity are evaluated only for patients who did not die in hospital or within 90 days of discharge.
†Comparison for length of stay is Q4 vs Q1–3.
IMD, Index of Multiple Deprivation; IRR, incidence rate ratio.