Josephson CB, et al. - Via this open cohort study, researchers seek to derive prediction models to estimate the risk of psychiatric adverse effects from levetiracetam use among patients with epilepsy. They created 2 prediction models: one for the overall population and one for those without a history of a psychiatric sign, symptom, or disorder during the study period. As per observations, levetiracetam has rapidly become a drug of first choice, and these models seem to have potential utility in guiding prescription in clinical practice.

• Researchers performed a retrospective open cohort study including all patients meeting the case definition for epilepsy after the Acceptable Mortality Reporting date in The Health Improvement Network (THIN) database based in the United Kingdom (inclusive January 1, 2000, to May 31, 2012) who received a first-ever prescription for levetiracetam.
• THIN had 11,194,182 registered patients, of these, 7400 were presumed to be incident cases (66.1 cases per 100 000 persons) over a maximum of 12 years’ follow-up.
• The index date was when patients received their first prescription code for levetiracetam; the cases were followed-up for 2 years or until an event, loss to follow-up, or censoring occurred.
• They performed the analyses on April 22, 2018.
• A Read code for any psychiatric sign, symptom, or disorder as reached through consensus by 2 authors was the outcome of interest.
• Using regression techniques, they derived 2 prediction models, one for the overall population and one for those without a history of a psychiatric sign, symptom, or disorder during the study period.

• Researchers identified 1173 patients with epilepsy who were receiving levetiracetam; 590 (50.3%) females; overall median age 39 (interquartile range, 25-56) years.
• A psychiatric symptom or disorder was observed in a total of 14.1% (165 of 1173) cases within 2 years of index prescription.
• Significantly higher odds of reporting a psychiatric symptom was noted for women (odds ratio [OR], 1.41; 95% CI, 0.99-2.01; P=.05) and those with a preexposure history of higher social deprivation (OR, 1.15; 95% CI, 1.01-1.31; P=.03), depression (OR, 2.20; 95% CI, 1.49-3.24; P < .001), anxiety (OR, 1.74; 95% CI, 1.11-2.72; P=.02), or recreational drug use (OR, 2.02; 95% CI, 1.20-3.37; P=.008).
• The model was noted to perform well after stratified k = 5-fold cross-validation (area under the curve [AUC], 0.68; 95% CI, 0.58-0.79).
• When all risk factors were present, they noted a gradient in risk, with probabilities increasing from 8% for 0 risk factors to 11% to 17% for 1, 17% to 31% for 2, 30% to 42% for 3, and 49%.
• A second model was observed performing comparably well after k = 5-fold cross-validation for those free of a preexposure psychiatric code, (AUC, 0.72; 95% CI, 0.54-0.90).
• Using threshold cutoffs of 0.10 (full model) and 0.14 (second model), specificity was maximized; a score below these thresholds suggests safety of prescription.