SUMMARY The case reports of 102 patients with subdural empyema, diagnosed in the years 1935-83, were reviewed to determine the factors affecting the outcome. Statistical analysis (likelihood ratio tests with chi square approximation and logistic regression) showed that year of diagnosis (p < 0.01) and level of consciousness at the moment of diagnosis (p < 0 01) had a significant bearing on the chance to survive and that these same two factors (each factor p < 0O01) and extent of subdural pus accumulation at the moment of diagnosis (p < 0-05) had a significant bearing on the chance of survival without severe disability. Among others the duration of the disease up to the moment of diagnosis and the mode of the first surgical procedure had no significant bearing on the outcome. These results together with those in the literature are discussed and it is concluded that diagnosis and treatment before the patient lapses into stupor or coma, increases the chance of survival and that with adequate management a mortality rate of 10% or lower is to be expected.Subdural empyema is a rare intracranial infection with high mortality. A review in 1975 indicated a mortality rate of about 40%.1 In more recent series the mortality rate ranged from 12% to 29%.2 11 Bannister in 1981 stated that a mortality rate of about 10% is to be expected nowadays with early diagnosis and prompt treatment. He also stressed the importance of performing a craniotomy as the primary surgical procedure.4 To determine which factors affect the outcome 102 cases of subdural empyema, in which the diagnosis was established in the years 1935-83, were reviewed.
Patients and methodsFrom the files of five neurosurgical centres in the Netherlands 102 patients were collected, in whom a subdural empyema was diagnosed between 1935 and 1983. In 97 cases the diagnosis was confirmed at operation. In four patients the diagnosis was made at necropsy and in one patient, who was treated non-surgically, CT confirmed the clinical diagnosis. The clinical findings of the patients have been reported in detail elsewhere. 11The outcome for each patient, mentioned in the clinical or outpatient chart, was graded according to a scale that was designed for this study (table 7). Grade A comprises patients who survived without or with a minor, not disabling, focal