Abstracts

Rationale: Cephalosporins are among the most widely prescribed class of antibiotics in hospitals. Cefepime, with its broad spectrum activity, is one of the mostly used drugs in empirical therapy for severe infections. Studies have suggested that mortality was higher in patients treated with cefepime than in those treated with other beta-lactams. There was no good explanation for this. Several case reports of non-convulsive status epilepticus associated with cefepime in patients with renal failure suggested that neurological complications might be involved. We compared the proportion of patients associating continuous epileptiform discharges on EEG while receiving cefepime with those receiving meropenem, another broad spectrum antibiotic commonly used. The aim was to determine whether non-convulsive status was more frequent with cefepime than with meropenem. This might explain the higher mortality reported in literature. Methods: We reviewed the databases of patients who had an EEG performed during treatment with cefepime or meropenem during the last three years. EEGs were either routine 20 to 30 minutes EEGs or continuous EEG monitoring. Clinical and biological characteristics of patients associating continuous epileptiform discharges on EEG while taking cefepime or meropenem were studied, as well as brain iconography. Proportions were compared using Fisher s exact test. Results: We studied 933 patients treated with cefepime and 1005 patients treated with meropenem. Continuous epileptiform discharges on EEG were found in 16 patients in the cefepime group and in none in the meropenem group (1.7 vs 0%, p< 0.001). Fifteen patients had generalized periodic epileptiform discharges (GPEDs) and 1 continuous multifocal spikes. Among the 16 patients, 78 % were hospitalized in the intensive care unit. Mean age was 67 years. No patients were known to have epilepsy. Brain iconography showed a sequellae of a left frontal stroke in one patient and aspecific leucoencephalopathy in two others. Blood creatinine concentration was over 1 mg/dL in 6 patients and elevated liver enzymes in 5. No patient had major electrolyte disturbances. Mean time to diagnosis was 5 days (range 1 to 10) after antibiotic was started. 30 day mortality was 21%.