Seizure vs. syncope: measuring serum creatine kinase in the emergency department

To determine the utility of serum creatine kinase measurement for diagnosing generalized tonic-clonic seizures in patients presenting to an emergency department with transient loss of consciousness.

Prospective evaluation of a diagnostic parameter. Retrospective data collection with blinded assignment to diagnostic groups.

University teaching hospital.

Sequential sample of 205 patients with transient loss of consciousness. The study group consisted of 96 patients who had creatine kinase measurements in the emergency department.

An investigator blinded to the results of creatine kinase measurements retrospectively classified events into seizure and nonseizure groups on the basis of clinical presentation, prior history, and follow-up investigations. Mean (+/- SE) creatine kinase level was significantly higher in the seizure group (231.1 +/- 34.8 U/L vs. 70.5 +/- 5.6 U/L, p less than 0.001). Elevated creatine kinase had a test specificity of 0.98 (95% CI 0.90-1.00) and a sensitivity of 0.43 (95% CI 0.28-0.59). The discriminating power of creatine kinase elevation was directly related to the time interval between the event and testing (p less than 0.0001). Among samples taken more than three hours after the event, test sensitivity was 0.80 (0.52-1.00) and specificity was 0.94 (0.71-1.00).

Creatine kinase may be a useful test for evaluating patients with transient loss of consciousness. The test is highly specific for diagnosing generalized seizures in the emergency department. Test sensitivity improves by sampling serum at least three hours after the event.