Cox regression for the association of delirium and delirium recognition with inpatient mortality
| Beta | SE | Wald | Freedom | HR | CI | P value | ||
| Lower | Upper | |||||||
| Delirium unadjusted | 1.14 | 0.30 | 14.43 | 1 | 3.13 | 1.74 | 5.65 | <0.001* |
| Delirium adjusted* | 0.98 | 0.33 | 9.02 | 1 | 2.65 | 1.40 | 5.01 | 0.003* |
| Recognition unadjusted | −0.61 | 0.41 | 2.17 | 1 | 0.55 | 0.24 | 1.22 | 0.141 |
| Recognition adjusted† | −0.38 | 0.47 | 0.68 | 1 | 0.68 | 0.27 | 1.70 | 0.411 |
| Recognition adjusted‡ | −0.33 | 0.55 | 0.36 | 1 | 0.72 | 0.24 | 2.12 | 0.547 |
The presence of delirium was associated with increased risk of inpatient death in both univariable and multivariable analyses. Recognition of delirium did not statistically significantly impact on risk of inpatient mortality. The ORs represent the likelihood of death with recognised delirium compared with unrecognised delirium.
*Delirium adjusted for age, gender, dementia status, frailty and specialty.
†Recognition adjusted for age, gender, dementia status, frailty and specialty.
‡Recognition adjusted for variables above, duration, and subtype.