Shekunov, Julia

Importance The potential utility of real-time digital therapeutics that identify impending behavioral outbursts to proactively apply evidence-based interventions for managing children with disruptive behavior disorder has not been tested. Objectives To evaluate feasibility and adherence of a digitally augmented, real-time, proactive behavioral intervention (smartwatch) for children with disruptive behaviors. Design, Setting, and Participants This randomized clinical trial was conducted between March 1, 2022, and December 31, 2023, at Mayo Clinic in Rochester, Minnesota. The study enrolled children aged 3 to 7 years with externalizing behavior problems rated above the clinically significant range (I score ≥120; T score ≥60) on the Eyberg Child Behavior Inventory (ECBI). Interventions Twelve weekly sessions of parent-child interaction therapy (PCIT) (treatment as usual [TAU] arm) vs artificial intelligence (AI)–enhanced PCIT augmented with real-time behavioral outburst alerts (PCIT-AI arm). Main Outcomes and Measures The primary outcome was achieving the a priori feasibility benchmark set at greater than 70% for both completion of PCIT sessions and daily smartwatch wear. Parental reaction time to behavior prompts was measured in seconds. Secondary outcomes were percentage changes in ECBI and Pediatric Sleep Questionnaire scores. Behavioral outburst duration was examined based on parental logs of start and end times of observed tantrums using a smartphone application. Result This study enrolled 50 children (median [IQR] age, 5.0 [4.0-6.0] years; 34 [68%] male and 16 [32%] female; 28 randomized to the PCIT-AI arm and 22 to the PCIT-TAU arm). Of these children, 37 (74%; 21 in the PCIT-AI arm and 16 in the PCIT-TAU arm) completed PCIT, with children wearing the watch a median (IQR) 75.7% (67.8%-82.9%) of the time, thus achieving the feasibility benchmark (primary outcome). Median (IQR) response time to behavior prompts in the PCIT-AI arm was 3.65 (2.07-7.08) seconds. Assessing between-group differences, there was no statistically significantly greater numerical improvement in ECBI-intensity percentage change (13.86%; 95% CI, −1.41% to 29.14%), ECBI-problem percentage change subscale scores (38.59%; 95% CI, −1.22% to 78.40%), or absolute change in Pediatric Sleep Questionnaire scores (−0.16; 95% CI, −1.85 to 1.53) with the PCIT-AI arm than with the PCIT-TAU arm. Children in the PCIT-AI arm had significantly shorter mean (SD) tantrum durations (10.4 [20.8] minutes vs 22.1 [30.0] minutes; P < .001) and lower odds of tantrums lasting for 15 minutes or more (odds ratio, 3.66; 95% CI, 2.72-4.95). Conclusions and Relevance In this randomized clinical trial, both recruitment and technology feasibility benchmarks were achieved. The findings inform the design of fully powered efficacy study of wearable, digitally augmented PCIT for children with disruptive behaviors. Trial Registration ClinicalTrials.gov Identifier: NCT05077722

Objective: Parents frequently purchase and inquire about smartwatch devices to monitor child behaviors and functioning. This pilot study examined the feasibility and accuracy of using smartwatch monitoring for the prediction of disruptive behaviors. Methods: The study enrolled children ( N = 10) aged 7–10 years hospitalized for the treatment of disruptive behaviors. The study team completed continuous behavioral phenotyping during study participation. The machine learning protocol examined severe behavioral outbursts (operationalized as episodes that preceded physical restraint) for preparing the training data. Supervised machine learning methods were trained with cross-validation to predict three behavior states—calm, playful, and disruptive. Results: The participants had a 90% adherence rate for per protocol smartwatch use. Decision trees derived conditional dependencies of heart rate, sleep, and motor activity to predict behavior. A cross-validation demonstrated 80.89% accuracy of predicting the child's behavior state using these conditional dependencies. Conclusion: This study demonstrated the feasibility of 7-day continuous smartwatch monitoring for children with severe disruptive behaviors. A machine learning approach characterized predictive biomarkers of impending disruptive behaviors. Future validation studies will examine smartwatch physiological biomarkers to enhance behavioral interventions, increase parental engagement in treatment, and demonstrate target engagement in clinical trials of pharmacological agents for young children.

Abstract Background Implicit cognitive markers may assist with the prediction of suicidality beyond clinical risk factors. The aim of this study was to investigate neural correlates associated with the Death/Suicide Implicit Association Test (DS-IAT) via event-related potentials (ERP) in suicidal adolescents. Methods Thirty inpatient adolescents with suicidal ideations and behaviors (SIBS) and 30 healthy controls from the community were recruited. All participants underwent 64-channel electroencephalography, DS-IAT, and clinical assessments. Hierarchical generalized linear models with spatiotemporal clustering were used to identify significant ERPs associated with the behavioral outcome of DS-IAT (D scores) and group differences. Results Behavioral results (D scores) showed that the adolescents with SIBS had stronger implicit associations between “death” and “self” than the healthy group (P = .02). Within adolescents with SIBS, participants with stronger implicit associations between “death” and “self” reported more difficulty in controllability of suicidal ideation in the past 2 weeks based on the Columbia-Suicide Severity Rating Scale (P = .03). For the ERP data, the D scores and N100 component over the left parieto-occipital cortex had significant correlations. Significant group differences without behavioral correlation were observed for a second N100 cluster (P = .01), P200 (P = .02), and late positive potential (5 clusters, all P ≤ .02). Exploratory predictive models combining both neurophysiological and clinical measures distinguished adolescents with SIBS from healthy adolescents. Conclusions Our results suggest that N100 may be a marker of attentional resources involved in the distinction of stimuli that are congruent or incongruent to associations between death and self. Combined clinical and ERP measures may have utility in future refinements of assessment and treatment approaches for adolescents with suicidality.

Guanfacine is a selective alpha-2a adrenoreceptor agonist that with overdose can cause symptoms ranging from mild sedation to coma, respiratory depression, hyporeflexia, hypotonia, bradycardia, and hypotension. Despite a well-defined and predictable toxidrome, variations can be seen based on multiple factors including age, quantity ingested, organ functions, coingestions, time since ingestion, and specific dosage form. Here, we describe two cases of delayed presentation of extended release guanfacine toxicity and highlight the variations encountered in the toxidrome presentation. These cases bring to attention the importance of maintaining a high suspicion for such atypical presentations, keeping in mind the limitations of managing these complications on an inpatient psychiatric unit.

Background The treatment of depression in children and adolescents is a substantial public health challenge. This study examined artificial intelligence tools for the prediction of early outcomes in depressed children and adolescents treated with fluoxetine, duloxetine, or placebo. Methods The study samples included training datasets ( N = 271) from patients with major depressive disorder (MDD) treated with fluoxetine and testing datasets from patients with MDD treated with duloxetine ( N = 255) or placebo ( N = 265). Treatment trajectories were generated using probabilistic graphical models (PGMs). Unsupervised machine learning identified specific depressive symptom profiles and related thresholds of improvement during acute treatment. Results Variation in six depressive symptoms (difficulty having fun, social withdrawal, excessive fatigue, irritability, low self‐esteem, and depressed feelings) assessed with the Children’s Depression Rating Scale‐Revised at 4–6 weeks predicted treatment outcomes with fluoxetine at 10–12 weeks with an average accuracy of 73% in the training dataset. The same six symptoms predicted 10–12 week outcomes at 4–6 weeks in (a) duloxetine testing datasets with an average accuracy of 76% and (b) placebo‐treated patients with accuracies of 67%. In placebo‐treated patients, the accuracies of predicting response and remission were similar to antidepressants. Accuracies for predicting nonresponse to placebo treatment were significantly lower than antidepressants. Conclusions PGMs provided clinically meaningful predictions in samples of depressed children and adolescents treated with fluoxetine or duloxetine. Future work should augment PGMs with biological data for refined predictions to guide the selection of pharmacological and psychotherapeutic treatment in children and adolescents with depression.