Predicting Adolescent Intervention Non-responsiveness for Precision HIV Prevention Using Machine Learning.

Interventions to teach protective behaviors may be differentially effective within an adolescent population. Identifying the characteristics of youth who are less likely to respond to an intervention can guide program modifications to improve its effectiveness. Using comprehensive longitudinal data on adolescent risk behaviors, perceptions, sensation-seeking, peer and family influence, and neighborhood risk factors from 2564 grade 10-12 students in The Bahamas, this study employs machine learning approaches (support vector machines, logistic regression, decision tree, and random forest) to identify important predictors of non-responsiveness for precision prevention. We used 80% of the data to train the models and the rest for model testing. Among different machine learning algorithms, the random forest model using longitudinal data and the Boruta feature selection approach predicted intervention non-responsiveness best, achieving sensitivity of 85.4%, specificity of 78.4% and AUROC of 0.93 on the training data, and sensitivity of 84.3%, specificity of 67.1%, and AUROC of 0.85 on the test data. Key predictors include self-efficacy, perceived response cost, parent monitoring, vulnerability, response efficacy, HIV/AIDS knowledge, communication about condom use, and severity of HIV/STI. Machine learning can yield powerful predictive models to identify adolescents who are unlikely to respond to an intervention. Such models can guide the development of alternative strategies that may be more effective with intervention non-responders.

Comparing standard versus enhanced implementation of an evidence-based HIV prevention program among Bahamian sixth grade students: findings from nationwide implementation trials.

BACKGROUND: Effective implementation strategies are needed to address the challenges encountered by teachers in implementation of evidence-based HV prevention programs in schools. The current study: 1) compares implementation fidelity of Focus on Youth in the Caribbean (FOYC) plus Caribbean Informed Parents and Children Together (CImPACT) intervention using enhanced implementation strategies (including biweekly monitoring/feedback and site-based mentoring) to those using more traditional approach (teacher training only); and 2) evaluates the impact of school coordinators' and mentors' performance on teachers' implementation fidelity and student outcomes. METHODS: Data from an enhanced implementation trial in 2019-2020, involving 24 government primary schools, 79 teachers, and 2252 students, were compared to data from a standard implementation trial in 2011-2012, involving 35 government primary schools, 110 teachers and 2811 students using mixed-effects modeling and structural equation modeling. FINDINGS: Teachers in the 2019-2020 trial taught more core activities (28.3 vs. 16.3, t = 10.80, P < 0.001) and sessions (7.2 vs. 4.4, t = 9.14, P < 0.001) than those participating in the 2011-2012 trial. Teachers taught > 80% of the intervention curriculum in 2019-2020 compared to 50% curriculum delivery in 2011-2012. Teachers who had a "very good" or "excellent" school coordinator in their schools taught more core activities than those who had a "satisfactory" school coordinator (30.4 vs. 29.6 vs. 22.3, F = 18.54, P < 0.001). Teachers who worked in a school which had a "very good" mentor, taught more core activities than those teachers who did not have a mentor or had only a "satisfactory" mentor (30.4 vs. 27.6; t = 2.96; p = 0.004). Teachers' confidence in implementing core activities, comfort level with the curriculum, attitudes towards sex education in schools, and perceived principal support were significantly related to increased self-efficacy, which in turn was related to teachers' implementation fidelity. The degree of implementation was significantly associated with improved student outcomes. IMPLICATIONS/CONCLUSION: An evidence-based HIV prevention intervention can achieve a high degree of implementation when delivered with enhanced implementation strategies and implementation monitoring. Future program implementers should consider the purposeful selection and training of school coordinators and mentors to support low-implementing teachers as a potentially important strategy when attempting to achieve high-quality implementation of school-based interventions.

The Effectiveness of Two Implementation Strategies for Improving Teachers' Delivery of an Evidenced-based HIV Prevention Program.

BACKGROUND: Effective implementation strategies are needed to enhance the success of evidence-based prevention programs. The current study evaluates the effects of two implementation strategies on teachers' implementation of an evidenced-based HIV intervention. METHODS: Using our 7-item pre-implementation school screening tool, we identified teachers who were at-risk for not implementing the Focus on Youth HIV-risk reduction intervention curriculum which targets grade six through grade 8 students. After completing a two-day curriculum workshop, 81 low- and moderate-performing teachers were randomly assigned to one of four experimental conditions and were asked to teach the two-month intervention curriculum. This optimization trial examines the impact of two implementation strategies: biweekly monitoring/feedbacks (BMF) and site-based assistance/mentorship (SAM). The primary outcome is implementation fidelity defined as number of core activities taught. Linear mixed-effects model was used to examine the association of the implementation strategies with implementation fidelity. RESULTS: BMF and SAM were significantly associated with teachers' implementation fidelity. Teachers who received both BFM and SAM taught the greatest numbers of core activities (15 core activities on average), followed by teachers who received either BMF (6.9 activities) or SAM (7.9 activities). Teachers who did not receive BMF or SAM taught the lowest numbers (4.1 activities). Teachers' sustained implementation of FOYC in the prior school year was related to increased implementation fidelity during the optimization trial. Teachers' confidence in implementing five core activities, attitudes toward sex education in schools, and perceived principal support were significantly related to increased self-efficacy, which in turn was related to teachers' fidelity of implementation before the optimization trial. CONCLUSION: BMF and SAM are effective in promoting teachers' implementation of youth evidence-based interventions. Researchers and future program implementers should consider teacher training, teachers' attitudes toward sex education, perceived principal support, and self-efficacy when attempting to maintain the effects of teacher-delivered interventions in schools.

Adolescent HIV-related behavioural prediction using machine learning: a foundation for precision HIV prevention.

BACKGROUND: Precision prevention is increasingly important in HIV prevention research to move beyond universal interventions to those tailored for high-risk individuals. The current study was designed to develop machine learning algorithms for predicting adolescent HIV risk behaviours. METHODS: Comprehensive longitudinal data on adolescent risk behaviours, perceptions, peer and family influence, and neighbourhood risk factors were collected from 2564 grade-10 students at baseline followed for 24 months over 2008-2012. Machine learning techniques [support vector machine (SVM) and random forests] were applied to innovatively leverage longitudinal data for robust HIV risk behaviour prediction. In this study, we focused on two adolescent risk behaviours: had ever had sex and had multiple sex partners. Twenty percent of the data were withheld for model testing. RESULTS: The SVM model with cost-sensitive learning achieved the highest sensitivity, at 79.1%, specificity of 75.4% with AUC of 0.86 in predicting multiple sex partners on the training data (10-fold cross-validation), and sensitivity of 79.7%, specificity of 76.5% with AUC of 0.86 on the testing data. The random forest model obtained the best performance in predicting had ever had sex, yielding the sensitivity of 78.5%, specificity of 73.1% with AUC of 0.84 on the training data and sensitivity of 82.7%, specificity of 75.3% with AUC of 0.87 on the testing data. CONCLUSION: Machine learning methods can be used to build effective prediction model(s) to identify adolescents who are likely to engage in HIV risk behaviours. This study builds a foundation for targeted intervention strategies and informs precision prevention efforts in school-setting.

Global Burden of Violence.

This article documents the increasing numbers of children impacted annually by 1 or more types of violence against children and describes the range of types of injuries and their immediate and long-term impacts on child outcomes. The article describes the growing number of international collaborations to decrease the numbers of children impacted by violence and to mitigate the consequences thereof, with a particular emphasis on children living in war zones.

Medical Students on the Virtual Front Line: A Literature Review Elective to Provide COVID-19 Clinical Teams With Essential Information.

PROBLEM: At the Hackensack Meridian School of Medicine (HMSOM) in New Jersey, clinical activities for students were suspended on March 15, 2020, due to the COVID-19 pandemic. Clinical teams at Hackensack Meridian Health (HMH) needed resources for identifying and assimilating the medical literature regarding COVID-19, which was expanding and evolving daily. HMH leaders reached out to HMSOM leaders for assistance. The HMSOM leadership and faculty quickly organized a literature review elective. APPROACH: Eight second-year medical students participated in a literature review elective course to research and synthesize the COVID-19 clinical literature to provide synopses of best practices for various clinical teams. By March 23, students were searching the literature and writing reports independently, mentored by a senior dean (an infectious diseases specialist) and supported by the associate dean of libraries and library team. The library team updated and categorized student reports daily on a website dedicated to the elective. OUTCOMES: During the 6-week elective, 8 students produced 70 reports synthesizing the emerging COVID-19 literature to help answer practitioners' clinical questions in real time. One student report was posted on the American Academy of Ophthalmology website. All 70 were published online in Elsevier's health education faculty hub. On course evaluations, students expressed regret about not being directly involved in patient care but articulated their gratitude to be able to contribute to the clinical teams. NEXT STEPS: In June 2020, the students returned to their clinical clerkships as COVID-19 clinical volumes declined and personal protective equipment became more available. Students continued to be available to the clinical teams to assist with COVID-19 questions. This literature review elective can serve as a model for other medical schools to use to deploy students to help synthesize the evolving literature on COVID-19 or other rapidly emerging research topics.

Bahamas National Implementation Project: Proposal for Sustainability of an Evidence-based HIV Prevention Intervention in a School Setting.

BACKGROUND: Sustained implementation of school-based prevention programs is low. Effective strategies are needed to enhance both high-level implementation fidelity and sustainability of prevention programs. OBJECTIVE: This proposed study aims to determine if the provision of either biweekly monitoring and feedback and site-based assistance and mentorship or both to at-risk and moderate-performing teachers with monitoring through an enhanced decision-making platform by the Ministry of Education (MOE) and Ministry of Health (MOH) based on the real-time implementation data will increase national implementation fidelity and result in sustained implementation over time. METHODS: This study will target government schools including 200 grade 6 teachers in 80 primary schools and 100 junior/middle high school teachers (and their classes) on 12 Bahamian islands. Teacher and school coordinator training will be conducted by the MOE in year 1, followed by an optimization trial among teachers in the capital island. Informed by these results, an implementation intervention will be conducted to train using different levels of educational intensity all at-risk and moderate-performing teachers. Subsequently selected training and implementation strategies will be evaluated for the national implementation of Focus on Youth in the Caribbean and Caribbean Informed Parents and Children Together in years 2 to 5. RESULTS: It is hypothesized that a more intensive training and supervision program for at-risk and moderate-performing teachers will enhance their implementation fidelity to the average level of the high-performing group (85%), an HIV prevention program delivered at the national level can be implemented with fidelity in grade 6 and sustained over time (monitored annually), and student outcomes will continue to be highly correlated with implementation fidelity and be sustained over time (assessed annually through grade 9). The proposed study is funded by the National Institute of Child Health and Human Development from August 1, 2018, through May 31, 2023. CONCLUSIONS: The study will explore several theory-driven implementation strategies to increase sustained teacher implementation fidelity and thereby increase the general public health impact of evidence-based interventions. The proposed project has potential to make significant contributions to advancing school-based HIV prevention research and implementation science and serve as a global model for the Fast Track strategy. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/14816.