Identification of Patient Factors Associated with Malignant and Nonmalignant EEG Patterns in Adult Inpatients Receiving Stat Electroencephalography
Abstract number :
1.165
Submission category :
3. Neurophysiology / 3C. Other Clinical EEG
Year :
2022
Submission ID :
2203978
Source :
www.aesnet.org
Presentation date :
12/3/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:22 AM
Authors :
Paolo Angelo Balaga, DNP, ACNPC-AG, CCRN-CMC – University of Pittsburgh Medical Center Presbyterian - Shadyside; Joanna Fong-Isariyawongse, MD, FAES – Director of Critical Care Continuous EEG Monitoring, University of Pittsburgh Comprehensive Epilepsy Center, University of Pittsburgh Presbyterian - Shadyside; Dianxu Ren, MD, PhD – Director for Statistical and Data Support Services, Department of Health and Community Systems, University of Pittsburgh School of Nursing; Jane Guttendorf, DNP, CRNP, ACNP-BC – Assistant Professor, Department of Acute/Tertiary Care, University of Pittsburgh School of Nursing
Rationale: At least 170,000 individuals in the United States alone have a first seizure each year, and more than half do not have any previous diagnosis of epilepsy. Electroencephalography (EEG) is frequently used as an initial neurophysiological diagnostic test to diagnose seizures or status epilepticus, as well as to evaluate for any cerebral dysfunction of various clinical etiologies. EEG is often ordered by clinicians in the hospital setting to evaluate for patients with symptoms suspicious of seizures or status epilepticus, or with altered mental status of unknown etiologies.
Methods: In this study, stat electroencephalography (EEG) records were retrospectively reviewed to determine stat EEG ordering practices, incidence rate of malignant and nonmalignant EEG findings, and patient factors associated with malignant and nonmalignant findings on stat EEG. Univariate analysis was performed to determine factors associated with malignant EEG findings, followed by multivariate regression for demographic variables or discharge diagnoses potentially predictive of malignant EEG findings.
Results: Twenty-three percent of stat EEGs had malignant EEG readings. There were no significant differences between outcome groups (malignant vs. nonmalignant EEG findings) for age (p = .42), gender (p = .28) or race (p = .71). There were no significant differences between outcome groups for day and time of test (p = .94) or indication for the test (p = .62). No significant differences were noted between groups for any of the past medical histories included in this evaluation. In univariate analysis, statistically significant differences were seen between groups for discharge diagnoses of central nervous system (CNS) infection (p = 0.01) and encephalopathy (p = 0.03). In multivariate regression analysis, discharge diagnoses of head trauma (OR 24.872, 95% CI 1.647, 375.676; p = 0.020), CNS infection (OR 44.522, 95% CI 3.451, 574.452; p = 0.004), CNS tumor (OR 15.973, 95% CI 1.067, 239.094; p = 0.045) and encephalopathy (OR 6.423, 95% CI 1.146, 35.986; p = 0.034) were associated with malignant EEG findings.
Conclusions: In conclusion, almost a quarter of stat EEGs had malignant EEG readings. Most stat EEG orders were placed by the departments of Internal Medicine and Neurology. There was no significant difference between the patient’s age, gender classification, or race and malignant vs nonmalignant EEG findings. The incidence of nonmalignant findings is three times that of malignant EEG both during business hours and after hours. No significant differences were noted between outcome groups for day and time of test and indication for the test. Of note, past medical history (including history of epilepsy, head trauma, intracerebral hemorrhage, stroke, CNS tumor or infection) was not significantly associated with malignant vs. nonmalignant EEG findings. However, discharge diagnoses of CNS infection, head trauma, CNS tumor, and encephalopathy were identified as factors associated with malignant EEG findings on stat EEG.
Funding: None
Neurophysiology