Blood pressure regulation to prevent progression of blunt traumatic intracranial hemorrhage in stable patients.

BACKGROUND

Target blood pressure (BP) in stable (non-hypotensive) patients with acute isolated blunt traumatic intracranial hemorrhage (TICH) is unknown. To address this issue, our study correlated BP with radiological volumetric progression (RP) and neurological deterioration (ND) in these patients.

METHODS

A retrospective review of hemodynamically stable adults (n = 184) with isolated TICH not requiring emergent surgery consecutively admitted to a Level I trauma center. BPs before admission computed tomography (CT) scan (CT1) and between CT1 and a follow-up CT (CT2) were correlated with TICH volume and Glasgow Coma Scale (GCS) during these time periods. Predictors for deterioration were studied. Primary outcomes were increased measured TICH and decreased GCS at the CT1-CT2 interval.

RESULTS

Age (57 years), % male (73), ISS (17), % falls (77), comorbidities (1.2/pt), and % anticoagulation (20) were similar in patients with or without RP (n = 107, 58%) or ND (n = 34, 18%). By univariate analysis, RP patients had an average systolic (SBP), diastolic (DBP), and mean BP (MAP) similar to non-RP patients; whereas ND patients compared to non-ND patients had a higher mean admission DBP (p < 0.02) and MAP (p < 0.04), a higher mean admission peak MAP (p < 0.01) and DBP (p < 0.01), a higher CT1-CT2 interval peak DBP (p < 0.01) and peak MAP (p < 0.01), and a lower CT1-CT2 nadir SBP (p < 0.04). Spearman rank correlation test did not show association among average SBP, MAP, DBP, absolute or % change in BPs, and absolute or % change in TICH volumes in any phase. Multivariate analysis identified higher nadir admission SBP [adjusted odds ratio (AOR) 1.29 per 10 mmHg increase] and lower peak MAP during the CT1-CT2 period (AOR 0.71 per 10 mmHg decrease) as independent predictors of RP, and a peak DBP in the CT1-CT2 interval (AOR 1.48) as an independent predictor of ND. Other predictors of ND included bilateral admission TICH (AOR 3.31) and increased injury volume (AOR 1.36), while the number of comorbidities/patient (AOR 4.34), bilateral injury (AOR 3.12), and midline shift (AOR 4.34) predicted RD.

CONCLUSIONS

A comprehensive dynamic analysis correlating repeated BP determinations with quantifiable repeated parameters of TICH deterioration (injury volume and GCS) did not demonstrate a clinically relevant protective target BP value. Current practices of BP control in this specific group of patients should be further investigated.

LEVEL OF EVIDENCE III

Prognostic, Level II study.