Table 5.
Multivariate regression analysis of predictors of ICH recurrence within 90 days.
| Variables | P | OR | 95% CI | Variables | P | OR | 95% CI |
|---|---|---|---|---|---|---|---|
| Model 1Aa | Model 1Ba | ||||||
| Hypertension | 0.077 | 0.358 | 0.115–1.116 | Hypertension | 0.069 | 0.344 | 0.109–1.084 |
| Previous symptomatic ICH | 0.049 | 3.087 | 1.003–9.504 | Previous symptomatic ICH | 0.042 | 3.290 | 1.041–10.397 |
| IVH presence | 0.441 | 1.545 | 0.511–4.677 | IVH presence | 0.368 | 1.676 | 0.545–5.155 |
| cSAH presence | 0.013 | 5.473 | 1.425–21.028 | Adjacent cSAH | 0.007 | 6.562 | 1.680–25.631 |
| Remote cSAH | 0.722 | 1.554 | 0.137–17.608 | ||||
| Model 2Ab | Model 2Bb | ||||||
| Age | 0.287 | 1.041 | 0.968–1.121 | Age | 0.306 | 1.040 | 0.965–1.121 |
| Previous symptomatic ICH | 0.008 | 5.129 | 1.542–17.056 | Previous symptomatic ICH | 0.007 | 5.509 | 1.611–18.839 |
| Previous antiplatelet use | 0.387 | 1.906 | 0.442–8.222 | Previous antiplatelet use | 0.270 | 2.335 | 0.518–10.536 |
| Previous anticoagulants use | 0.037 | 32.126 | 1.242–830.844 | Previous anticoagulant use | 0.011 | 80.008 | 2.748–2329.227 |
| cSAH presence | 0.007 | 6.573 | 1.666–25.935 | Adjacent cSAH | 0.003 | 8.732 | 2.053–37.131 |
| Remote cSAH | 0.793 | 1.448 | 0.091–22.985 | ||||
| Model 3Ac | Model 3Bc | ||||||
| Previous symptomatic ICH | 0.018 | 3.989 | 1.266–12.576 | Previous symptomatic ICH | 0.016 | 4.347 | 1.321–14.299 |
| Hypertension | 0.056 | 0.326 | 0.103–1.027 | Previous anticoagulant use | 0.014 | 53.011 | 2.252–1248.103 |
| cSAH presence | 0.006 | 6.388 | 1.695–24.074 | Hypertension | 0.063 | 0.329 | 0.102–1.061 |
| Adjacent cSAH | 0.003 | 7.868 | 2.055–30.123 | ||||
Variables with a P-value < 0.1 in the univariate logistic analysis were included. Hypertension, Previous symptomatic ICH, IVH presence, cSAH presence were included in model 1A. The first 3 variables, adjacent cSAH and remote cSAH were included in model 1B. Eighteen patients were excluded because of missing follow-up data on recurrence.
Prespecified plausible predictors of recurrent ICH were included. Age, previous symptomatic ICH, previous antiplatelet use, previous anticoagulant use, cSAH presence were included in model 2A. The first 4 variables, adjacent cSAH and remote cSAH were included in model 2B. 30 patients were excluded because of missing data.
Prespecified plausible predictors as well as variables with a P-value < 0.1 in univariable regression were included using backward logistic regression. Age, hypertension, previous ICH, previous antiplatelet use and anticoagulant use, IVH, and cSAH were entered, among which 4 variables, including age, previous antiplatelet and anticoagulants use, and IVH were eliminated in the final model as backward logistic regression in model 3A. Age, previous antiplatelet use and anticoagulant use, IVH, hypertension, previous ICH, adjacent and remote cSAH were entered, among which 4 variables, including age, antiplatelet use, IVH and remote cSAH were eliminated in the final model as backward logistic regression in model 3B. Thirty patients were excluded because of missing data.
cSAH, convexity subarachnoid hemorrhage; ICH, intracerebral hemorrhage; IVH, intraventricular hemorrhage.