0343 Association between the real-world use of trazodone and daytime impairment in US patients with insomnia

Abstract

Abstract Introduction Trazodone is approved for the treatment of depression and is used off-label to treat insomnia despite Clinical Practice Guidelines recommending against this use. The objective of this real-world retrospective study is to explore whether trazodone use is associated with improvements in daytime impairment in US patients with insomnia. Methods Information on insomnia patients aged ≥ 18 years prescribed trazodone as first-line treatment between October 2015 and March 2020 was retrieved from HealthVerity. Insurance claims data were extracted for each patient alongside physician notes in free-text form. From the claims, patient demographics and prescription information were extracted. An ontology analogous to the Insomnia Daytime Symptoms and Impact Questionnaire (IDSIQ) was designed with three domains (i.e., alert/cognition, mood and sleepiness). The 14 IDSIQ terms were extended to 109 terms to reflect physicians’ language. Daytime impairment symptom frequency occurring from 182 days before treatment to ~3 months (mean) during treatment was measured as annualized rates normalized by one year as the duration before and during treatment is different. Each symptom noted in a visit is counted once in a specific domain. Results were stratified by dose: trazodone < 150mg (typically used for off-label insomnia treatment) and trazodone ≥150mg (approved dose for depression treatment). Results 1187 and 246 unique patients were identified on the < 150mg and ≥150 mg trazadone doses, respectively. For trazodone < 150mg, the annualized rate of daytime impairment symptoms before and during treatment was 139.4 and 147.5, respectively, with a statistically non-significant overall increase (6%, p=0.313) and statistically non-significant increases in each domain (alert/cognition: 4%,p=0.574; mood: 10%, p=0.173; sleepiness: 5%,p=0.401). For ≥150mg trazodone, the annualized rate of daytime impairment symptoms before and during treatment was 143.5 and 142.2, respectively, with a statistically non-significant overall decrease (1%,p=0.941), statistically non-significant decreases in the alert/cognition (19%,p=0.128) and sleepiness (5%,p=0.712) domains and statistically non-significant increase in the mood domain (14%,p=0.400). Conclusion Patients with insomnia prescribed trazodone show similar levels of daytime impairment before and during treatment. This study suggests that the use of trazodone in patients with insomnia is not associated with a decrease in daytime impairment. Support (if any) Idorsia Pharmaceuticals Ltd.