Among patients with epilepsy, a simple model that incorporates factors such as a patient’s sex and history of depression, anxiety, and recreational drug use may help predict the risk of a psychiatric adverse effect from levetiracetam, according to a study published in JAMA Neurology.
“This study derived 2 simple models that predict the risk of a psychiatric adverse effect from levetiracetam” and can “guide prescription in clinical practice,” said Colin B. Josephson, MD, of the department of clinical neurosciences at the University of Calgary (Canada) and his research colleagues.
Levetiracetam is a commonly used first-line treatment for epilepsy because of its ease of use, broad spectrum of action, and safety profile, the researchers said. Still, psychiatric adverse reactions occur in as many as 16% of patients and frequently require treatment discontinuation.
To evaluate whether routine clinical data can predict which patients with epilepsy will experience a psychiatric adverse event from levetiracetam, the investigators analyzed data from The Health Improvement Network (THIN) database, which includes anonymized patient records from general practices in the United Kingdom. They assessed 21 variables for possible inclusion in prediction models. They identified these variables by searching the literature and weighing input from a panel of experts.
Their analysis included data from Jan. 1, 2000–May 31, 2012. Among the more than 11 million patients in THIN, the researchers identified 7,300 incident cases of epilepsy. The researchers examined when patients received a first prescription for levetiracetam and whether patients experienced a psychiatric symptom or disorder within 2 years of the prescription.
Among 1,173 patients with epilepsy receiving levetiracetam, the median age was 39 years; about half were women. In all, 14.1% experienced a psychiatric symptom or disorder within 2 years of prescription. Women were more likely to report a psychiatric symptom (odds ratio, 1.41), as were patients with a history of social deprivation (OR, 1.15), anxiety (OR, 1.74), recreational drug use (OR, 2.02), or depression (OR, 2.20).
The final model included female sex, history of depression, history of anxiety, and history of recreational drug use. Low socioeconomic status was not included because “it would be challenging to assign this score in clinic,” the authors said.
“There was a gradient in risk probabilities increasing from 8% for 0 risk factors to 11%-17% for 1, 17% to 31% for 2, 30%-42% for 3, and 49% when all risk factors were present,” Dr. Josephson and his colleagues indicated. “The discovered incremental probability of reporting a psychiatric sign can help generate an index of suspicion to counsel patients.”
Using the example of a woman patient with depression, the model “suggests she would be at risk,” with a 22% chance of a psychiatric adverse event in the 2 years after receiving a levetiracetam prescription.
The researchers created a second prediction algorithm based on data from patients without documentation of a mental health sign, symptom, or disorder prior to their levetiracetam prescription. This model incorporated age, sex, recreational drug use, and levetiracetam daily dose; it performed comparably well and might be used to determine safety of prescription, according to Dr. Josephson and his colleagues.
The authors noted that the study was limited by an inability to evaluate medication adherence and seizure type and frequency. One advantage of the study’s design is that it may have circumvented expectation bias because general practitioners were not prone to anticipating psychiatric adverse events or to have a lower threshold for diagnosing them.
The authors disclosed research fellowships and support from foundations and federal agencies.
SOURCE: Josephson CB et al. JAMA Neurol. 2019 Jan 28. doi: 10.1001/jamaneurol.2018.4561.