Objective:Practice guidelines emphasize frequent clinical monitoring of patients at high risk for psychosis. No brief instrument assessing attenuated psychotic symptoms has been validated for this purpose. This study examined use of three self-report questionnaires, which were developed as psychosis risk screeners, for monitoring symptom severity in a naturalistic clinical sample of 54 adolescents.Methods:Self-report measures (Prime Screen–Revised, Prodromal Questionnaire–Brief Version [PQ-B], and Youth Psychosis At-Risk Questionnaire–Brief) and clinician assessments (Structured Interview for Psychosis Risk Syndromes) were administered to participants at baseline and six months.Results:Changes in self-report scores were moderately correlated with changes in clinician ratings. The PQ-B demonstrated slightly better agreement with changes in clinician ratings than the other two measures.Conclusions:Questionnaires developed as psychosis risk screeners could be used for symptom monitoring. Further validation o...

Background: Youth treated with antipsychotic medications are high risk for weight gain, increased lipids/glucose, and development of metabolic syndrome. Little is known about the dietary intake/nutritional adequacy in this vulnerable population, and effect on weight gain. This secondary data analysis describes the baseline intake and changes in diet after receiving healthy lifestyle education/counseling over 6 months, in a sample of youth with antipsychotic-induced weight gain. Methods: The U.S. Department of Agriculture (USDA) Automated Multiple-Pass Method 24-hour dietary recall was administered to 117 youth at baseline, 3 months, and 6 months. Parent/child received personalized healthy lifestyle education sessions over 6 months. Baseline intake was compared with the USDA Recommended Daily Allowance using independent samples t -tests. Individual dietary covariates were examined for change over 6 months using longitudinal linear mixed modeling. Influence of each on body mass index (BMI) z -score change was tested in a pooled group analysis and then compared by treatment group. Results: Pooled analysis revealed baseline consumption high in carbohydrates, fat, protein, sugar, and refined grains, while low in fruit/vegetables, whole grains, fiber, and water. Change over 6 months demonstrated a statistically significant decrease in daily calories ( p = 0.002), carbohydrates ( p = 0.003), fat ( p = 0.012), protein ( p = 0.025), sugar ( p = 0.008), refined grains ( p = 0.008), total dairy ( p = 0.049), and cheese ( p = 0.027). Small increases in fruits/vegetables were not statistically significant, although the Healthy Eating Index subscores for total vegetables ( p = 0.013) and dark green/orange vegetables ( p = 0.034) were. No dietary covariates were predictors of change in BMI z -score. Nondietary predictors were parent weight/BMI and treatment group, with the metformin and switch groups experiencing significant decreases in BMI z -score. Conclusions: Further pediatric studies are necessary to assess the effects of antipsychotic medications on dietary intake, and test efficacy of healthy lifestyle interventions on change in nutrition. The relationship of nutrition to cardiometabolic health in this population must be further investigated. Clinical Trial Registration number: NCT 02877823.

Background: Patient engagement is a critical but challenging public health priority in behavioral health care. During telehealth sessions, health care providers need to rely predominantly on verbal strategies rather than typical nonverbal cues to effectively engage patients. Hence, the typical patient engagement behaviors are now different, and health care provider training on telehealth patient engagement is unavailable or quite limited. Therefore, we explore the application of machine learning for estimating patient engagement. This can assist psychotherapists in the development of a therapeutic relationship with the patient and enhance patient engagement in the treatment of mental health conditions during tele–mental health sessions. Objective: This study aimed to examine the ability of machine learning models to estimate patient engagement levels during a tele–mental health session and understand whether the machine learning approach could support therapeutic engagement between the client and psychotherapist. Methods: We proposed a multimodal learning-based approach. We uniquely leveraged latent vectors corresponding to affective and cognitive features frequently used in psychology literature to understand a person’s level of engagement. Given the labeled data constraints that exist in health care, we explored a semisupervised learning solution. To support the development of similar technologies for telehealth, we also plan to release a dataset called Multimodal Engagement Detection in Clinical Analysis (MEDICA). This dataset includes 1229 video clips, each lasting 3 seconds. In addition, we present experiments conducted on this dataset, along with real-world tests that demonstrate the effectiveness of our method. Results: Our algorithm reports a 40% improvement in root mean square error over state-of-the-art methods for engagement estimation. In our real-world tests on 438 video clips from psychotherapy sessions with 20 patients, in comparison to prior methods, positive correlations were observed between psychotherapists’ Working Alliance Inventory scores and our mean and median engagement level estimates. This indicates the potential of the proposed model to present patient engagement estimations that align well with the engagement measures used by psychotherapists. Conclusions: Patient engagement has been identified as being important to improve therapeutic alliance. However, limited research has been conducted to measure this in a telehealth setting, where the therapist lacks conventional cues to make a confident assessment. The algorithm developed is an attempt to model person-oriented engagement modeling theories within machine learning frameworks to estimate the level of engagement of the patient accurately and reliably in telehealth. The results are encouraging and emphasize the value of combining psychology and machine learning to understand patient engagement. Further testing in the real-world setting is necessary to fully assess its usefulness in helping therapists gauge patient engagement during online sessions. However, the proposed approach and the creation of the new dataset, MEDICA, open avenues for future research and the development of impactful tools for telehealth.

Background: Antipsychotic medication (APM) can cause weight gain, insulin resistance, dyslipidemias, and an increased risk of developing type-2 diabetes and cardiovascular disease among youth. The study sought to increase water consumption, reduce sugar-sweetened beverage (SSB) intake, and prevent unhealthy weight gain via a healthy lifestyle intervention among youth newly started on a second-generation APM for psychiatric treatment. Methods: This randomized controlled trial enrolled 148 Medicaid-insured youth (ages 8–17) recently starting APM. The treatment group received both a biweekly home-delivery of bottled water and parental phone support from a family navigator. In-home visits conducted at baseline, three months, and six months assessed the participants’ height/weight and dietary intake. All participants received basic healthy lifestyle education emphasizing increased water intake and decreased SSB consumption. Longitudinal linear mixed models were conducted to examine between-group and within-group changes in BMI z-scores, and water/SSB intake. Results: No significant between-group differences in BMI z-score were found at three (p = 0.908) and six months (p = 0.919). However, the within-group increase in BMI z-score in the control group was significant from baseline to three months (p = 0.029). A between-group comparison found the treatment group significantly increased their water intake at three (p = 0.006) and six months (p = 0.002). No between-group differences were identified at three and six months for the reduction in SSB, although the treatment group did demonstrate a decrease from baseline to three months (p = 0.004). Conclusions: Neither group experienced unhealthy increases (>0.5%) in BMI z-score over the six months. Providing a safe/free water supply showed a superior improvement in water consumption in the treatment group, and an initial decrease in SSB. Further studies are needed to identify feasible, effective, and sustainable lifestyle interventions tailored to this at-risk population.

BACKGROUND Patient engagement is a critical but challenging public health priority in behavioral health care. During telehealth sessions, health care providers need to rely predominantly on verbal strategies rather than typical nonverbal cues to effectively engage patients. Hence, the typical patient engagement behaviors are now different, and health care provider training on telehealth patient engagement is unavailable or quite limited. Therefore, we explore the application of machine learning for estimating patient engagement. This can assist psychotherapists in the development of a therapeutic relationship with the patient and enhance patient engagement in the treatment of mental health conditions during tele–mental health sessions. OBJECTIVE This study aimed to examine the ability of machine learning models to estimate patient engagement levels during a tele–mental health session and understand whether the machine learning approach could support therapeutic engagement between the client and psychotherapist. METHODS We proposed a multimodal learning-based approach. We uniquely leveraged latent vectors corresponding to affective and cognitive features frequently used in psychology literature to understand a person’s level of engagement. Given the labeled data constraints that exist in health care, we explored a semisupervised learning solution. To support the development of similar technologies for telehealth, we also plan to release a dataset called Multimodal Engagement Detection in Clinical Analysis (MEDICA). This dataset includes 1229 video clips, each lasting 3 seconds. In addition, we present experiments conducted on this dataset, along with real-world tests that demonstrate the effectiveness of our method. RESULTS Our algorithm reports a 40% improvement in root mean square error over state-of-the-art methods for engagement estimation. In our real-world tests on 438 video clips from psychotherapy sessions with 20 patients, in comparison to prior methods, positive correlations were observed between psychotherapists’ Working Alliance Inventory scores and our mean and median engagement level estimates. This indicates the potential of the proposed model to present patient engagement estimations that align well with the engagement measures used by psychotherapists. CONCLUSIONS Patient engagement has been identified as being important to improve therapeutic alliance. However, limited research has been conducted to measure this in a telehealth setting, where the therapist lacks conventional cues to make a confident assessment. The algorithm developed is an attempt to model person-oriented engagement modeling theories within machine learning frameworks to estimate the level of engagement of the patient accurately and reliably in telehealth. The results are encouraging and emphasize the value of combining psychology and machine learning to understand patient engagement. Further testing in the real-world setting is necessary to fully assess its usefulness in helping therapists gauge patient engagement during online sessions. However, the proposed approach and the creation of the new dataset, MEDICA, open avenues for future research and the development of impactful tools for telehealth.