Background: The fate of the bone flap is a significant decision during surgical treatment of acute subdural hematoma (SDH). A general guideline revolves around the surgeon's concern for brain edema. Limited studies, however, have focused on the factors that contribute to perioperative brain edema.

Methods: From 2012 to 2015, 38 patients who underwent decompressive craniectomy for acute SDH were reviewed. Clinical data were extracted (age, sex, initial Glasgow Coma Scale (GCS) score, sodium level, hematocrit, and intraoperative blood loss). From the preoperative scan, SDH volume, midline shift (MLS), and volume within the skull (to estimate baseline brain volume) were measured. From the postoperative scan, brain volume (including any herniating regions) was measured. Δ% was defined as the percentage change in postoperative brain volume compared with preoperative volume. Evident contralateral injury, contusions, and intraventricular hemorrhage (IVH) were noted.

Results: Fifteen patients demonstrated negative Δ%. Univariate analysis found significant correlations between Δ% and preoperative MLS, initial GCS, presence of IVH, and presence of contralateral injury (P < 0.05). A multiple regression for Δ% elicited a significant model (F [3, 34] = 17.387, P < 0.01) with R2 0.605, where Δ% = 16.197 - 1.246*GCS - 0.986 * MLS + 3.292 * IVH (with 0 = no IVH, 1 = presence of IVH).

Conclusions: A high proportion of patients can exhibit negative Δ%, or relative brain compression after decompression of SDH. For these patients, replacement of the bone flap may be reasonable to avoid obligatory interval cranioplasty. Preoperative MLS, initial GCS, and presence of IVH can help predict whether overall brain volume will swell or compress within the normal confines of the skull. This can guide the decision to retain or remove the bone flap.