0412 Evaluation of PROMIS Sleep Disturbance Item Bank for Postpartum Insomnia Disorder Screening

Abstract Introduction Impaired sleep experience in the postpartum period is associated with negative maternal and infant outcomes. Insomnia is common yet underdiagnosed following childbirth. Prevalence varies widely depending on the insomnia measurement tool used, and few have been validated in the postpartum population. We aimed to assess the diagnostic accuracy of sleep surveys (Bergen Insomnia Scale (BIS), PROMIS Sleep Disturbance (SD) and Sleep-Related Impairment (SRI)) compared to diagnostic clinical interviews for sleep disorders (SCISD-R). Methods After IRB approval, adults who delivered a live infant were recruited in a longitudinal prospective cohort study. Patients were approached during delivery admission at two hospitals (CA and AR) and invited to complete BIS, PROMIS SD and SRI at 6 weeks postpartum. Participants were also invited to complete virtual 1-hour SCISD-R interviews, the criterion standard for insomnia diagnosis, until 150 interviews were completed. Area under the ROC curve were analyzed for BIS, PROMIS SD and SRI scores against insomnia disorder. Sensitivity and specificity of PROMIS SD and SRI against SCISD-R-diagnosed insomnia disorder was evaluated with a cut-off T score of 55 (a recommended threshold) to distinguish abnormal SD and SRI cases. Results 150 of the 413 (36.3%) invited participants completed SCISD-R at mean 43.6±6.6 days postpartum, with mean interview duration 24.6±9.7 minutes. Of these, 144 (96.0%) participants completed PROMIS SD, SRI and BIS. 20 individuals (13.3%) met criteria for insomnia disorder. AUC for BIS raw score, PROMIS SD and SRI T scores were 0.73, 0.85 and 0.70 respectively. At a PROMIS SD cut-off T score of 55, sensitivity was 0.75 (95.0% CI 0.509, 0.913) and specificity was 0.758 (0.673, 0.830). For PROMIS SRI, sensitivity at cut-off T score of 55 was 0.75 (0.509, 0.913), and specificity was 0.45 (0.36, 0.54). Conclusion The ranked order of discrimination for insomnia disorder compared to criterion standard diagnostic interviews was PROMIS SD, BIS and SRI (best to worst). PROMIS SD should be considered for screening for insomnia disorder following childbirth. Support (if any) This study was funded by the National Heart, Lung, and Blood Institute (R01HL166253-01A1).