PROGNOSTIC FACTORS AFTER ACUTE SUBDURAL-HEMATOMA

Factors that hare been shown to affect outcome after acute subdural he matoma (ASDH) include age, Injury Severity Score (ISS), intracranial p ressure (ICP), direct admission to a trauma center, presence of subara chnoid hemorrhage, score on the Glasgow Coma Scale (GCS), and timing o f operation. However, these data come from selected patient population s (e.g., operated, comatose, or minimally symptomatic patients, etc.). In an effort to evaluate factors that predict outcome for the entire spectrum of ASDH patients, we evaluated 211 patients with ASDH and GCS scores of 3 to 15. One hundred twenty-eight patients (61%) were manag ed nonoperatively (Nonop), whereas 83 (39%) were managed with cranioto my [operatively (Op)] Op patients had more severe brain injuries, as e videnced by their lower GCS scores (Op 7.8 vs. Nonop 10.7, p = 0.0001) , higher incidence of large ASDH with midline shift (Op 61% large ASDH , 83% midline shift vs. Nonop 16% large ASDH, 30% midline shift, p = 0 .001 for each comparison), and higher incidence of basilar cistern eff acement (Op 61% vs. Nonop 21%, p = 0.001). Thirty-five percent of the Op patients had their hematoma evacuated within 3 hours (early), where as 65% did not (delayed). Early Op patients had a significantly lower incidence of functional survival (early = 24% vs. delayed = 51%, p = 0 .02). The early patients seem to have had more significant head injuri es, as evidenced by their lower GCS scores (early 7.0 vs. delayed = 8. 4), higher incidence of associated intracranial injuries (early = 1.14 vs. delayed = 0.85), and higher incidence of cistern effacement (earl y = 76% vs. delayed 53%,p = 0.002). Among 97 patients with GCS scores of 3 to 8, 26% survived to a functional (F) outcome, whereas 74% were nonfunctional (NF), including 47% who died. The F survivors had signif icantly lower ISSs (F = 24.8 vs. NF = 31.3, p = 0.03), higher GCS scor es (F = 6.4 vs. NF = 4.6, p = 0.001), fewer associated intracranial in juries (F = 0.68 vs. NF = 1.19, p 0.02), lower maximal ICP (F = 25.8 v s. NF = 49.3, p = 0.03), and higher incidence of patent basilar cister ns (F = 41% vs. NF = 14%,p = 0.03). Among the 114 patients with GCS sc ores of 9 to 15, 79% went on to a F recovery, whereas 20% remained NF, including eight (7%) who died. F survivors had significantly lower IS Ss (F = 19.7 vs. NF = 26.8,p = 0.001), age (F = 36.4 vs. NF = 48.0, p = 0.005), higher GCS scores (F = 13.7 vs. NF = 12.0, p = 0.001), lower incidence of early craniotomy (F = 3% vs. NF = 30%,p = 0.02), and hig her incidence of patent basilar cisterns (F = 78% vs. NF = 44%,p = 0.0 2). Survivors were also significantly less likely to have been involve d in a motor vehicle crash (F = 29% vs. NF = 57%, p = 0.02). In a logi stic regression model of all ASDH patients, the factors that were foun d to predict outcome independently include age (p = 0.002), GCS scores (p = 0.002), ISSs (p = 0.003), and pupillary reactivity (p = 0.006). Using this model, early operation was nearly statistically significant for predicting NF outcome (p = 0.07). Early operation improved outcom e only in comatose patients with large ASDH. We conclude that, in eval uating the entire spectrum of patients with ASDH, factors that are ind ependently predictive of outcome include age, GCS scores, ISSs, and pu pillary reactivity. Among patients with GCS scores of 3 to 8, the fact ors that are most predictive of outcome seem to be those related to th e severity of the head injury, whereas in patients with GCS scores of 9 to 15, the factors that correlate with outcome are those that relate to the severity of multisystem injury.