Suicidal ideation and behavior screening in intractable focal epilepsy eligible for drug trials

Three suicidal ideation and suicidal behavior instruments were used to assess the prevalence of lifetime and recent suicidal ideation and suicidal behavior in patients with frequent treatment-resistant focal seizures who would be eligible for randomized clinical trials. This was done to determine which instrument was optimal for use in epilepsy.

In a cross-sectional study, we compared lifetime and recent suicidal ideation and suicide attempt on the MINI International Neuropsychiatric Interview (MINI), Columbia Suicide Severity Rating Scale (C-SSRS), and Interactive Voice Response System CSSRS (E-CSSRS). A safety algorithm determined treatment referral. Coordinators and participants evaluated experiences with the C-SSRS. The proportion of participants that baseline assessment would exclude from clinical trial enrollment was determined.

Among 208 participants, 1.6-3.9% had recent high risk suicidal ideation and 1.0-4.7% had a recent suicide attempt across all instruments. Lifetime high-risk suicidal ideation occurred in 12.1-14.1%. Lifetime suicide attempt occurred in 10.2-13.1% of participants. Of those with recent suicide attempt, 31.1% required referral to a health professional, and 3.9% needed urgent referral. Lifetime suicidal behavior (including aborted suicide attempt, interrupted suicide attempt, suicide attempt, preparatory acts or behavior, and nonsuicidal self-injurious behavior) was found in 21.1% on the E-CSSRS and 15.5% on the C-SSRS. Agreement (Kappa) was good to excellent for comparisons of all instruments. Fifty-two percent of subjects preferred either the CSSRS or E-CSSRS, whereas the rest had no preference; of those having a preference, 87.5% favored the CSSRS. Of the 18.9% of participants who might have been excluded from trials based on suicidal ideation and suicide attempt, the CSSRS identified high-risk suicidal ideation or suicide attempt in the preceding 2 years in only 4.4%.

Suicidality screening is feasible in people with epilepsy. Slightly more suicidal behavior is reported with the E-CSSRS than C-SSRS, suggesting the E-CSSRS may be optimal. The proportion of patients who may be excluded from clinical trials based on worrisome suicidal ideation or suicide attempt is small, suggesting that it is possible to enroll most eligible individuals.