Factors influencing the presence of hemiparesis in chronic subdural hematoma.

Objective: Chronic subdural hematoma (CSDH) has a variety of clinical presentations, including hemiparesis. Hemiparesis is of the utmost importance because it is one of the major indications for surgical intervention and influences outcome. In the current study, the authors intended to identify factors influencing the presence of hemiparesis in CSDH patients and to determine the threshold value of hematoma thickness and midline shift for development of hemiparesis.

Methods: The authors retrospectively reviewed 325 patients (266 with unilateral and 59 with bilateral hematomas) with CSDH who underwent surgical evacuation, regardless of presence or absence of hemiparesis.

Results: In univariate analysis, hematoma loculation, age, hematoma maximal thickness, and midline shift were significantly associated with hemiparesis. Moreover, patients with unilateral hematomas had a higher rate of hemiparesis than patients with bilateral hematomas. Sex, trauma history, anticoagulant and antiplatelet drug use, presence of comorbidities, Glasgow Coma Scale score, hematoma density characteristics on CT scan, and hematoma signal intensity on T1- and T2-weighted MRI were not associated with hemiparesis. In multivariate analysis, the presence of loculation and hematoma laterality (unilateral vs bilateral) influenced hemiparesis. For unilateral hematomas, maximal hematoma thickness of 19.8 mm and midline shift of 6.4 mm were associated with a 50% probability of hemiparesis. For bilateral hematomas, 29.0 mm of maximal hematoma thickness and 6.8 mm of shift were associated with a 50% probability of hemiparesis.

Conclusions: Presence of loculations, unilateral hematomas, older patient age, hematoma maximal thickness, and midline shift were associated with a higher rate of hemiparesis in CSDH patients. Moreover, 19.8 mm of hematoma thickness and 6.4 mm of midline shift were associated with a 50% probability of hemiparesis in patients with unilateral hematomas.