Table 4.
Adverse outcomes for the total, control, and intervention cohorts
| Adverse outcomes | Total (N = 218) | Control cohort (N = 125) | Review cohort (N = 93) | B, MD or OR | 95% CI | P value | |
|---|---|---|---|---|---|---|---|
| Lower | Upper | ||||||
| Duration of delirium in days, mean (range)a | 12.52 (1–99) | 15.47 (1–99) | 8.56 (1–45) | ||||
| Interaction group × number of medications | B = 0.08 | 0.02 | 0.15 | 0.01 | |||
| For patients taking 0–6 medications, mean | 20.34 | 4.88 | MD = 15.46 | 9.25 | 21.67 | < 0.001 | |
| For patients taking 7–11 medications, mean | 13.6 | 9.86 | MD = 3.74 | − 0.35 | 7.83 | 0.07 | |
| For patients taking 12 + medications, mean | 10.57 | 8.76 | MD = 1.81 | − 4.15 | 7.78 | 0.55 | |
| Length of stay in days, median (range)a,b | 16 (1–105) | 17 (1–105) | 15 (3–80) | B = 0.88 | 0.71 | 1.09 | 0.22 |
| Died in hospital, n (%) | 24 (11%) | 15 (12%) | 9 (10%) | OR = 0.74 | 0.305 | 1.814 | 0.51 |
| Discharged back homec, n (%) | 66 (37%) | 39 (38%) | 27 (36%) | OR = 1.01 | 0.536 | 1.916 | 0.97 |
All models were corrected for age, sex, Charlson Comorbidity Index, and number of medications used by the patient
B unstandardised coefficient, CI confidence interval, MD mean difference from the generalised linear regression model, OR odds ratio
aFor the duration of delirium, an interaction effect between cohort and number of medications was found. To illustrate this interaction effect, the mean duration of delirium for patients taking 5, 10, 15 or 20 medications for both cohorts has been included in the table
bFor the length of stay, none of the covariates were significant; the table therefore displays the unadjusted model
cDischarged back home was only calculated for those patients living at home before hospital admission, and therefore has an N of 179 (N = 104 in the control condition, and N = 75 in the medication review condition)