Abstract
In this observational study, we analysed a cohort of 164 subarachnoid haemorrhage survivors (until discharge from intensive care) with the aim to detect factors that influence the length of stay (LOS) in intensive care with multiple linear regression methods. Moreover, binary logistic regression methods were used to examine whether the time in intensive care is a predictor of outcome after 1 year. The clinical 1-year outcome was measured prospectively in a 12-month follow-up by telephone interview and categorised by the modified Rankin Scale (mRS). Patients who died during their stay in intensive care were excluded. Complications like pneumonia (β = 5.11; 95% CI = 1.75-8.46; p = 0.0031), sepsis (β = 9.54; 95% CI = 3.27-15.82; p = 0.0031), hydrocephalus (β = 4.63; 95% CI = 1.82-7.45; p = 0.0014), and delayed cerebral ischemia (DCI) (β = 3.38; 95% CI = 0.19-6.56; p = 0.038) were critical factors depending the LOS in intensive care as well as decompressive craniectomy (β = 5.02; 95% CI = 1.35-8.70; p = 0.0077). All analysed comorbidities such as hypertension, diabetes, hypothyroidism, cholesterinemia, and smoking history had no significant impact on the LOS in intensive care. LOS in intensive care (OR = 1.09; 95% CI = 1.03-1.15; p = 0.0023) as well as WFNS grade (OR = 3.72; 95% CI = 2.23-6.21; p < 0.0001) and age (OR = 1.06; 95% CI = 1.02-1.10; p = 0.0061) were significant factors that had an impact on the outcome after 1 year. Complications in intensive care but not comorbidities are associated with higher LOS in intensive care. LOS in intensive care is a modest but significant predictor of outcomes after subarachnoid haemorrhage.