Use of the Patient Health Questionnaire-2 to predict suicidal ideations in patients taking varenicline

Abstract

Varenicline, when taken for smoking cessation, has been associated with suicidal ideations (SI), but the incidence and risk factors for SI are largely unknown. The aim of this paper was to describe the incidence and pretreatment patient factors that are associated with SI when taking varenicline. We conducted a smoking cessation trial of 217 patients at a single center, all of whom took varenicline. Between 6 and 18 months after study completion (when the risk for SI became known), we successfully contacted 72 (response rate 33%) and queried them for SI during varenicline treatment. Of these, 4 (6%) had SI. Self-reported history of depression was not associated with SI (p = 0.11), but depressive symptoms at the time of varenicline initiation as measured by the Patient Health Questionnaire-8 (p = 0.004) and Patient Health Questionnaire-2 (p = 0.007) were associated with SI. The Patient Health Questionnaire-2 (which is only two questions) had a sensitivity of 75% and a specificity of 90% to predict SI. We conclude that current depressive symptoms, not necessarily a history of depression, are associated with SI when taking varenicline for smoking cessation. Providers should consider screening for depressive symptoms before treatment with varenicline. (Am J Addict 2012;00:1-7).