Table 2.
Outcomes at day 8 and day 90 after stroke according to prior benzodiazepine use in propensity-score-matched cohort
| Benzodiazepine users | Propensity-score-matched* | |||
| No (n=56) | Yes (n=56)† | OR (95% CI) | P value | |
| Outcome at day 8 | ||||
| All-cause death | 2/56 (4.3) | 4/56 (6.8) | 1.81 (0.23 to 13.90) | 0.56 |
| NIHSS ≥6 | 19/51 (37.3) | 20/51 (39.4) | 1.10 (0.46 to 2.60) | 0.83 |
| Poor outcome (mRS ≥2) | 33/54 (60.3) | 34/54 (63.2) | 1.13 (0.48 to 2.66) | 0.78 |
| Poor outcome (Barthel <95) | 29/52 (55.3) | 32/52 (61.2) | 1.28 (0.51 to 3.15) | 0.59 |
| Cognitive impairment (MMSE <24) | 13/37 (34.4) | 10/35 (28.0) | 0.75 (0.24 to 2.34) | 0.62 |
| Outcome at day 90 | ||||
| All-cause death | 4/53 (7.6) | 12/52 (23.5) | 3.93 (0.91 to 16.98) | 0.067 |
| Poor outcome (mRS ≥2) | 25/53 (46.4) | 31/52 (59.7) | 1.72 (0.75 to 3.91) | 0.19 |
| Poor outcome (Barthel <95) | 18/49 (36.0) | 16/40 (39.6) | 1.18 (0.47 to 2.94) | 0.72 |
| Cognitive impairment (MMSE <24) | 6/37 (14.9) | 5/33 (15.0) | 1.04 (0.23 to 4.51) | 0.96 |
Values are n/N (%) unless otherwise indicated, calculated after handling missing baseline data including in propensity-score calculation by multiple imputation procedure (m=20 imputed data sets; n were estimated using the combined rates and the mean number of patients without missing outcome values).
*Calculated using a generalised estimating equations (GEE) model (binomial distribution, logit function) take into account the propensity-score matched design.
†Mean numbers of matched pairs among the 20-imputed datasets.
MMSE, Mini Mental State Examination; mRS, modified Rankin score; NIHSS, National Institutes of Health Stroke Scale.