Effect of icu antihypertensive choice on discharge hypertension in patients admitted to neuro-icu

Introduction: Blood pressure control in Neurointensive Care (NeuroICU) patients is important for preventing secondary brain injury and recurrent stroke. While the strongest predictor of discharge hypertension (HTN) is residual HTN when transferring from the NeuroICU to general care (NeuroICU transfer), optimal NeuroICU anti-hypertensive medication (AHM) regimens are unknown. We therefore sought to compare effectiveness by AHM class and NeuroICU diagnosis. Methods: We included NeuroICU patients admitted (4/2016-7/2022) receiving an AHM at NeuroICU transfer, stratifying analyses by vascular (ischemic/hemorrhagic) or non-vascular diagnosis and excluding patients with in-hospital mortality or direct NeuroICU discharge. We defined HTN as systolic (SBP) ≥140mmHg or diastolic pressure (DBP) ≥90mmHg. 24-hour AHM intensity was normalized to its cohort maximum. We identified diagnosis-specific factors independently predicting HTN at NeuroICU transfer and separately at hospital discharge, adjusting for age, sex, ICU length of stay, goal blood pressure, and AHM intensity. Results: 3,537 patients were included (1,162 vascular, 2,375 non-vascular); 45% with HTN at NeuroICU transfer, 35.7% at discharge. Predictors of HTN differed by diagnosis and timepoint. The most frequent AHM classes were ACE inhibitors (ACEi)/ARBs, beta blockers, and hydralazine in vascular patients, and calcium channel blockers (CCBs) in non-vascular patients. In vascular patients, ACEi/ARBs were independently associated NeuroICU transfer HTN (OR 1.59, 95%CI 1.15-2.20), but no AHMs were associated with hospital discharge HTN. In non-vascular patients, ACEi/ARBs and hydralazine were independently associated with NeuroICU transfer HTN (OR 1.32, 1.02-1.70; OR 1.77, 1.18-2.67) while aldosterone antagonists (AAs) were associated with decreased odds of ICU transfer HTN (OR 0.54, 0.27-1.04). Hydralazine and CCBs were independently associated with discharge HTN in non-vascular patients (OR 1.55, 1.04-2.30; OR 1.36, 1.07-1.72), while AAs decreased the risk (OR 0.49, 0.23-0.96). Conclusions: The ideal NeuroICU AHM regimen may be disease-specific. AHM classes differed in HTN control at NeuroICU transfer and hospital discharge based on diagnosis, independent of dose intensity. Further investigation should clarify if these findings vary by institution.