P4.009: Mental Health Algorithms for Pediatric Residents

Background: The United States faces a growing crisis in pediatric mental health care delivery. There is a growing burden on pediatric primary care providers to manage mental health concerns, due to the dual mental health provider shortage and increased care demand. Pediatricians, especially trainees, often report feeling underprepared to meet this need, due to limited training on mental health care topics.

To address this training deficit, we developed a mental health care handbook for pediatric residents, integrating diagnostic and treatment algorithms to enhance real-time decision-making. The algorithms include evidence-based guidelines, culturally informed perspectives, and local resource content. We then evaluated the handbook’s implementation and early impact.

Methods: Algorithms were developed by synthesizing information across publicly available practice guidelines. Following user-centered design principles, algorithms were revised with pediatric attending physicians, child psychiatrists, child psychologists, and trainees. The algorithms were then uploaded to a shared website for the pediatric residency program. Algorithms were introduced to trainees through focused educational lectures designed on each mental health topic. Reminders for use are implemented through real-time clinical feedback from supervising pediatrics and child psychiatry faculty, as well as through scheduled emails and meetings. To date, specific algorithms have been built for diagnosis and management of: Depression, Anxiety, SSRI prescribing, and Suicidality.

Algorithms are live documents and continue to be updated with ongoing feedback. A Google feedback form structured with Likert-scale and free-text survey questions is linked on every algorithm, querying usability and clinical efficacy. Quantitative utilization data were collected based on click-tracking. Click counts for the first four months of use were summarized using descriptive statistics. Pearson coefficients were leveraged to determine possible relationships between usage of individual algorithms.

Results: The total number of clicks per week for each individual algorithm was highest for Anxiety (mean = 2.37, SD = 3.17), followed by Suicidality (mean = 1.63, SD = 1.71), Depression (mean = 0.89, SD = 1.52), and SSRI (mean = 0.58, SD = 1.64).

Anxiety was positively correlated with each of the other algorithms, suggesting concurrent usage patterns. The strongest correlation was between Anxiety and Suicidality (r = 0.7771, p = 0.0001).

Preliminary qualitative feedback from the Google form endorsed ease of use and improved clinical decision-making.

Conclusion: The handbook demonstrates early adoption and sustained engagement, particularly for managing anxiety and suicidality. Limitations include the inability to confirm depth of tool use and a short evaluation timeframe.

Ongoing efforts will seek to: (1) increase visibility of the workflow to encourage utilization, (2) gather descriptive feedback to enable continuous improvement, and (3) expand the set of workflows to include ADHD and Autism Spectrum Disorders.

This initiative provides a scalable model for integrating mental health education into pediatric residency training.