How Safe Is Prolonged Video-EEG Monitoring?

Evidence from a large retrospective cohort analysis indicates that video-EEG monitoring is a safe procedure with a low risk of complications.



NEW ORLEANS—Prolonged video-EEG monitoring is a safe procedure with rare complications, according to results from a retrospective study that drew data from the National Inpatient Sample, the largest publicly available all-payer inpatient care database in the US.

At the 64th Annual Meeting of the American Academy of Neurology, Nancy E. Hammond, MD, reported that an examination of 19 years of medical records revealed that prolonged video-EEG was “quite safe and that rates of injuries and complications were lower than previously reported.” Dr. Hammond is an Assistant Professor in the Department of Neurology at the University of Kansas Hospital and Director of the Comprehensive Epilepsy Center in Kansas City.

Video-EEG monitoring is useful for the diagnosis and classification of epilepsy, diagnosis of psychogenic nonepileptic seizures, and for presurgical evaluation of patients for epilepsy surgery. “This is a unique procedure that carries with it some potential risks as we actually bring our patients into the hospital and try to induce them into having seizures,” Dr. Hammond said. “In theory, patients could suffer from status epilepticus, prolonged seizures, or injuries due to the seizures themselves. In addition, there have been several recent, well-publicized cases of death in video-EEG monitoring units, and that got me curious to see how big of a problem this really was.”

Safety Analysis From the National Inpatient Sample
When a review of the literature yielded only a few studies on the risks of video-EEG monitoring, Dr. Hammond and her coinvestigator, Richard Dubinsky, MD, turned to the National Inpatient Sample, which is a 20% stratified sample of all US acute hospital admissions (excluding federal hospitals). The database includes demographics, primary and secondary diagnoses, primary and secondary procedures, length of stay, hospital costs, and discharge disposition. This involves information on six to eight million hospital admissions per year. The researchers used this database to determine the frequency of injuries that occurred during video-EEG admissions.

Their retrospective cohort analysis encompassed 1990 to 2007. The investigators searched the database for the ICD-9 procedure code for video-EEG monitoring, then used the Clinical Classifications Software (CCS) for ICD-9-CM to redact diagnostic and procedure codes. Adverse events were identified using complication codes. To minimize nonstandard admissions and get a clearer picture of patients who were in the hospital primarily for video-EEG monitoring, the researchers excluded the top fifth and bottom fifth percentiles of age; patients who were admitted from the emergency room; records that were missing data on sex, length of stay, and other demographic information; records that came from small and rural hospitals; and patients who had a length of stay of more than 13 days.

Close to 37,000 different video-EEG admissions were identified. The large majority of patients came from large teaching hospitals; a little over half of the patients were female. A final diagnosis of epilepsy was made in 53% of patients and nonepilepsy in 13% of patients; 33% of patients had codes for both epilepsy and nonepileptic events.

A Low Rate of Complications
The overall rate of complications was very low—less than 2%. The most common complication was intracranial injury, occurring in 1.5% of patients. Rounding out the top five most common complications were urinary tract infections, status epilepticus, DVT, and complications of care, all of which occurred in fewer than 1% of patients. Orthopedic injuries were also quite rare. More serious complications, including pneumonia, falls, cardiac and respiratory arrest, and pulmonary embolus, were also all very rare in the patient sample.

A retrospective cohort study such as this has some limitations. First, the integrity of the data depends on correct coding. Also, the researchers were not able to review the clinical data associated with these cases. “We just had the diagnosis and the demographics, but we didn’t know the medications or their seizure status or if they were in the hospital for presurgical evaluation or what they were initially admitted to the hospital for. And, in addition, we could not exclude the pre-existing conditions, so some of the complications that we saw may have actually pre-existed prior to the hospitalization.”

Overall, however, the researchers concluded that “video-EEG monitoring is a safe procedure with rare complications.”

—Glenn S. Williams

Next Article: