Who and what retrospective risk assessments miss: Examining retrospective denial of momentary suicidal ideation in adolescents

Abstract

Background: Emerging evidence indicates that real-time assessments such as ecological momentary assessment (EMA) detect individuals experiencing suicidal ideation (SI) who go undetected by retrospective assessments. However, it remains unclear for whom and why such discrepancies occur. Few prior studies accounted for method variance (e.g., survey vs. interview), and no studies have comprehensively examined which individual characteristics differentiate those with and without discrepant reports across EMA and retrospective measures. The current study aimed to address these gaps by examining discrepancies in SI endorsement across EMA and two forms of retrospective measures (survey, interview), as well as comparing those with and without discrepant reports on demographic, clinical, and EMA-based variables. Methods: Eighty community-based adolescents completed a baseline online survey assessing demographic and clinical characteristics, followed by a 14-day EMA assessing momentary SI (5x/day). At the conclusion of the EMA period, adolescents retrospectively reported their SI during the past 2 weeks via a survey and an interview. Results: Of 48 adolescents who endorsed some SI during EMA, 35%–41% retrospectively denied having experienced SI during the EMA period on the follow-up survey and interview. Adolescents with retrospective denial had a lower likelihood of lifetime SI (ps < 0.05, φc = 0.53–0.55), lower SI and depression severity (ps < 0.05, ds = 0.52–0.64), and lower hopelessness at baseline (ps < 0.05, ds = 0.56–0.60), as well as less frequent and less severe SI (ps < 0.05, ds = 0.32–0.72) during EMA compared to those who endorsed SI across both EMA and retrospective measures. Conclusion: Real-time monitoring detects more adolescents experiencing SI than retrospective assessments, particularly those with lower clinical severity. Therefore, real-time monitoring has the potential to facilitate early interventions by identifying adolescents who are not yet as advanced in the suicide risk trajectory but are at risk of further escalation.