Abstracts

Rationale: Therapeutic hypothermia (TH) is the standard of care for patients after ventricular fibrillation cardiac arrest (CA). Continuous EEG monitoring (cEEG) is increasingly used to monitor for subclinical seizures during TH and rewarming and to provide prognostic information. The value of this practice in terms of clinical impact in relation to cost, versus a practice of selective use of clinically-directed routine 20-40 minute EEG (rEEG) has not been performed. We compared clinical outcome and charges in a TH population in which cEEG was performed routinely with an equivalent cohort in which it was not. Methods: In December 2005, our institution implemented a protocol for TH after CA which did not include routine use of cEEG (comprising the pre-TH-cEEG cohort). In 2009, due to concerns in the literature for subclinical seizures during TH, cEEG was initiated on all CA TH patients (TH-cEEG cohort). Clinical outcome, Cerebral Performance Category (CPC) and estimated EEG charges were calculated for the pre-TH-cEEG and TH-cEEG cohorts. Clinical data was gathered from the EMR system. EEG estimated unit charges were based on the National Charge Data 50th percentile charges expressed in USD. The source of these data is the CMS 2010 Standard Analytical File as reported in Code Correct by MedAssets, Inc. CPT code 95822 was used to estimate the unit charge for rEEG, and 95951 for cEEG. Results: 91 patients were in the pre-TH-cEEG and 47 in the TH-cEEG cohorts. There was no statistical difference for age, mortality, hospital duration or CPC between the cohorts. In the pre TH-cEEG cohort, 19 patients (21%) underwent rEEG and 3 (3%) were placed on cEEG. The total estimated EEG charges for the entire pre-TH-cEEG cohort was $29,860.25 for a mean of $1571.59/patient for the 19 patients undergoing EEG. In the TH-cEEG cohort, the total estimated EEG charge was $203,998.50 at a mean charge of $4340.39/patient (p <0.0001). The mean duration of cEEG in the TH-cEEG cohort = 33.1 hours (range 15-102.5). Two patients in the pre-TH-cEEG cohort had recorded seizures, compared to five in the TH-cEEG cohort. All seven had a poor outcome. Conclusions: The main impetus for using cEEG in TH is to detect subclinical seizures in order to maximize clinical outcome. However, cEEG adds to costs as reflected in the significantly higher estimated charges identified in our study. Our study did not demonstrate improved outcomes associated with these higher charges. Implementation of cEEG increased the rate of detection of subclinical seizures, but those with seizures detected and treated by this modality had poor outcomes, thus raising questions regarding the value of routine deployment of cEEG in this setting. Future research on cEEG in TH should include critical assessments of economic value as well as overall clinical outcome in order to determine its optimal place in the care of CA TH patients.