206 - Automating Depression and Suicide Screening in Pediatric Primary Care

Publication Number: 3198.206

Samantha White, Duke University School of Medicine, Durham, NC, United States; Elliott J. Sally, Duke University School of Medicine, Chapel Hill, NC, United States; Mya Sendak, Duke University School of Medicine, Department of Pediatrics, Durham, NC, United States; Jason Tatreau, Duke University School of Medicine, Durham, NC, United States

Background: Suicide is the 2nd leading cause of death among US adolescents, leading the AAP to recommend annual, universal suicide screening for patients ages 12 and older. One study found that 40% of teenage patients who died by suicide had seen primary care in the year before their death, demonstrating primary care providers' role in addressing suicidal ideation (SI). Universal suicide screening within primary care has been associated with increased detection, mental health referrals, and preparedness of staff to screen and address mental health concerns. While these results are encouraging, common barriers to universal screening include time restraints and lack of resources.

Objective: To address these concerns, we automated depression and suicide screening of 12-17-year-olds at their primary care visits across a large health system. We mobilized a remote Pediatric Suicide Prevention Response Team to address positive suicide screens via safety assessments. We predicted that this would lead to an increase in completion of Stanley-Brown Safety Plans (SBSP) for teens who screened positive for SI or self-harm (SH) on the Patient Health Questionnaire-9 (PHQ-9) and Ask Suicide-Screening Questions (ASQ) in primary care.

Design/Methods: We evaluated our quality improvement (QI) intervention by conducting a retrospective cohort study of adolescents ages 12-17 years who had a PHQ-9 positive for SI/SH at their primary care visit between 12/2/24-10/17/25. A PHQ-9 was considered positive for SI/SH if it scored >0 on question 9, which would automatically present the ASQ to the patient. We used descriptive statistics to compare completion rates of the ASQ and SBSP before and after the intervention went live in the EMR on 5/13/25.

Results: 1526 patients had a PHQ-9 positive for SI/SH, with 688 patients prior to automated screening and 838 afterwards. Of the 688 pre-intervention patients, only 20% (n=135) had an ASQ completed, compared to 86% (n=719) post-intervention (Table 1). Additionally, for patients with a non-acute positive ASQ, 19% (n=20) had a SBSP documented prior to the intervention, compared to 16% (n=82) afterwards (Table 2).

Conclusion(s): There was a dramatic increase in ASQs completed for teens who endorsed SI/SH on their PHQ-9 following initiation of automated screening in primary care. Despite this increase, only 16% of suicidal teens had SBSPs documented in their chart in the first 5 months of our QI intervention. In our second PDSA cycle, we aim to improve the percentage of suicidal adolescents with SBSPs by increasing utilization of the Pediatric Suicide Prevention Response Team.

Table 1: The number of ASQs completed for patients with PHQ-9s positive for SI/SH pre- and post-intervention


Table 2: The number of SBSPs completed for patients with a non-acute positive ASQ pre- and post-intervention