Adolescent Risk Behavior Screening and Interventions in Hospital Settings: A Scoping Review

Research on prediction model of adolescent suicide and self-injury behavior based on machine learning algorithm

Objective

To explore the risk factors that affect adolescents' suicidal and self-injurious behaviors and to construct a prediction model for adolescents' suicidal and self-injurious behaviors based on machine learning algorithms.

Methods

Stratified cluster sampling was used to select high school students in Chongqing, yielding 3,000 valid questionnaires. Based on whether students had engaged in suicide or self-injury, they were categorized into a suicide/self-injury group (n=78) and a non-suicide/self-injury group (n=2,922). Gender, age, insomnia, and mental illness data were compared between the two groups, and a logistic regression model was used to analyze independent risk factors for adolescent suicidal and self-injurious behavior. Six methods-multi-level perceptron, random forest, K-nearest neighbor, support vector machine, logistic regression, and extreme gradient boosting-were used to build predictive models. Various model indicators for suicidal and self-injurious behavior were compared across the six algorithms using a confusion matrix to identify the optimal model.

Result

In the self-injury and suicide groups, the proportions of male adolescents, late adolescence, insomnia, and mental illness were significantly higher than in the non-suicide and self-injury groups (p <0.05). Compared with the non-suicidal self-injury group, this group also showed significantly increased scores in cognitive subscales, impulsivity, psychoticism, introversion-extroversion, neuroticism, interpersonal sensitivity, depression, anxiety, hostility, terror, and paranoia (p <0.05). These statistically significant variables were analyzed in a logistic regression model, revealing that gender, impulsivity, psychoticism, neuroticism, interpersonal sensitivity, depression, and paranoia are independent risk factors for adolescent suicide and self-injury. The logistic regression model achieved the highest sensitivity and specificity in predicting adolescent suicide and self-injury behavior (0.9948 and 0.9981, respectively). Performance of the random forest, multi-level perceptron, and extreme gradient models was acceptable, while the K-nearest neighbor algorithm and support vector machine performed poorly.

Conclusion

The detection rate of suicidal and self-injurious behaviors is higher in women than in men. Adolescents displaying impulsiveness, psychoticism, neuroticism, interpersonal sensitivity, depression, and paranoia have a greater likelihood of engaging in such behaviors. The machine learning model for classifying and predicting adolescent suicide and self-injury risk effectively identifies these behaviors, enabling targeted interventions.

Attention-Deficit/Hyperactivity Disorder and Teen Self-Report on Health Behaviors and Social-Emotional Wellbeing: United States, July 2021-December 2022

OBJECTIVE

Promoting health during adolescence can support long-term well-being, especially for teens diagnosed with attention-deficit/hyperactivity disorder (ADHD), who face increased risks due to the disorder's impact on development and health behaviors. ADHD is often associated with difficulties in social interactions, a higher likelihood of bullying involvement, and co-occurring mental health conditions. These factors may also be influenced by health factors such as physical activity, sleep quality, and screen time usage. Nationally representative teen self-reports provide a novel perspective on ADHD-related health outcomes compared with relying on parent reports.

METHOD

We used nationally representative data from the National Health Interview Survey (NHIS) and NHIS-Teen from July 2021 to December 2022, to examine teen-reported health and well-being factors, stratified by parent-reported ADHD diagnoses among teens aged 12 to 17 years. Weighted prevalence estimates and adjusted prevalence ratios (aPR) adjusting for teen age, sex, and family income, all with 95% confidence intervals (CIs), were calculated.

RESULTS

Just over 10% of teens had ADHD and they reported higher prevalence of bullying victimization (aPR = 1.64, CI = 1.27-2.11), difficulties making friends (aPR = 1.83, CI = 1.15-2.90), difficulty getting out of bed (aPR = 1.29, CI = 1.02-1.64), irregular wake times (aPR = 2.17, CI = 1.45-3.25), and >4 hours daily screen time (aPR = 1.26, CI = 1.05-1.52) than teens without ADHD; teens with ADHD reported a lower prevalence of lacking peer support (aPR = 0.70, CI = 0.51-0.96).

CONCLUSION

Teens with ADHD face distinct challenges related to social-emotional well-being and health behaviors that support overall wellness. Findings may inform opportunities for health promotion among teens with ADHD.

Increasing Adolescent Sexual Activity Screening Through a Provider-Based Intervention

Screening for adolescent sexual activity is a vital aspect of comprehensive pediatric care. Adolescents engage in risky sexual behaviors. Thus, a complete and accurate sexual health history can assist in the prevention and treatment of disease, prevention of unwanted pregnancy, treatment of existing diseases, and optimal planning of future healthcare for adolescents. Current evidence shows that provider-focused strategies improve the delivery of preventive services, including sexual health screenings. In this initiative, we assessed and examined pre- and post-screening rates for sexual activity among adolescents by advanced practice providers. This multi-site initiative was implemented in four school-based health centers and a school-linked center that included 2,102 unique patients ages 9 to 24 years. Our biphasic intervention included education for advanced practice providers and electronic health record modifications. Pre- and post-data collection was conducted to determine changes in the rate of screening for sexual activity during a primary care adolescent health visit over a 3-year period. Data were collected via retrospective medical chart review and analyzed in three time periods for comparison from 2018 to 2021. Screening rates for sexual activity increased significantly after the intervention (all p < .001) with the likelihood more than double that of the year before it was implemented. The intervention was deemed to be a feasible and cost-effective strategy to improve the provider's willingness and ability to provide more adolescent sexual health screenings.

Social Navigation for Adolescent Emergency Department Patients: A Randomized Clinical Trial

PURPOSE

Adolescent emergency department (ED) patients have unmet social needs that contribute to ED use. This study aimed to evaluate the effect of social needs navigation for adolescents on subsequent ED visits and community resource use and to identify characteristics associated with elevated social risk.

METHODS

Between July 2017 and August 2019, we used a random date generator to establish intervention and control group enrollment dates. All adolescents completed a social needs survey. Adolescents enrolled on intervention dates received in-person, risk-tailored social needs navigation. Those enrolled on control dates received a preprinted resource guide. We used chart review and follow-up calls to assess 12-month ED revisits and community resource use. Logistic regression was used to compare these outcomes between groups. We measured the association between ≥3 reported unmet needs and characteristics hypothesized a priori to be associated with elevated social risk (nonurgent visits, obesity, or any of nine "socially sensitive" chief complaints) using logistic regression.

RESULTS

A total of 399 adolescents were randomized. There was no difference between groups in the number of ED revisits. There was increased community resource use in the intervention group (adjusted odds ratio [aOR]: 3.5 [95% confidence interval {CI}: 1.5, 8.2]). Adolescents with a socially sensitive chief complaint had increased odds of ≥3 unmet needs (aOR: 2.2 [95% CI: 1.3, 3.6]), as did those with food insecurity in a post hoc analysis (aOR: 9.9 [95% CI: 4.0, 24.6]).

DISCUSSION

Social needs navigation increased community resource use but not subsequent ED visits. Adolescents with socially sensitive chief complaints or food insecurity reported increased unmet needs.

An Electronic Teen Questionnaire, the eTeenQ, for Risk Behavior Screening During Adolescent Well Visits in an Integrated Health System: Development and Pilot Implementation

Background

Screening for risk behaviors is a routine and essential component of adolescent preventive health visits. Early identification of risks can inform targeted counseling and care. If stored in discrete fields in the electronic health record (EHR), adolescent screening data can also be used to understand risk behaviors across a clinic or health system or to support quality improvement projects.

Objective

Goals of this pilot study were to adapt and implement an existing paper adolescent risk behavior screening tool for use as an electronic data capture tool (the eTeenQ), to evaluate acceptance of the eTeenQ, and to describe the prevalence of the selected risk behaviors reported through the eTeenQ.

Methods

The multidisciplinary project team applied an iterative process to develop the 29-item eTeenQ. Two unique data entry forms were created with attention to (1) user interface and user experience, (2) the need to maintain patient privacy, and (3) the potential to transmit and store data for future use in clinical care and research. Three primary care clinics within a large health system piloted the eTeenQ from August 17, 2020, to August 27, 2021. During preventive health visits for adolescents aged 12 to 18 years, the eTeenQ was completed on tablets and responses were converted to a provider display for teens and providers to review together. Responses to the eTeenQ were stored in a REDCap (Research Electronic Data Capture; Vanderbilt University) database, and for patients who agreed, responses were transferred to an EHR flowsheet. Responses to selected eTeenQ questions are reported for those consenting to research. At the conclusion of the pilot, the study team conducted semistructured interviews with providers and staff regarding their experience using the eTeenQ.

Results

Among 2816 adolescents with well visits, 2098 (74.5%) completed the eTeenQ. Of these, 1811 (86.3%) agreed to store responses in the EHR. Of 1632 adolescents (77.8% of those completing the eTeenQ) who consented for research and remained eligible, 1472 (90.2%) reported having an adult they can really talk to and 1510 (92.5%) reported feeling safe in their community, yet 401 (24.6%) reported someone they lived with had a gun and 172 (10.5%) reported having had a stressful or scary event that still bothered them. In addition, 157 (9.6%) adolescents reported they were or wondered if they were gay, lesbian, bisexual, pansexual, asexual, or other, and 43 (2.6%) reported they were or wondered if they were transgender or gender diverse. Of 11 staff and 7 providers completing interviews, all felt that the eTeenQ improved confidentiality and willingness among adolescents to answer sensitive questions. All 7 providers preferred the eTeenQ over the paper screening tool.

Conclusions

Electronic capture of adolescent risk behaviors is feasible in a busy clinic setting and well accepted among staff and clinicians. Most adolescents agreed for their responses to risk behavior screening to be stored in the EHR.

Quality Initiative to Increase Delivery of Adolescent Hospital-Based Reproductive Health Care

OBJECTIVES

Hospitals are an important nontraditional setting in which to address adolescent reproductive health. However, opportunities for intervention are frequently missed, especially for boys and patients hospitalized for noningestion complaints. Our global aim was to increase delivery of reproductive health care to adolescents hospitalized through our children's hospital Pediatric Hospital Medicine service.

METHODS

We performed 2 quality improvement intervention cycles: (1) provider education and monthly reminder e-mails and (2) an automated electronic health record (EHR) adolescent history and physical note template with social history prompts while discontinuing reminder e-mails. The primary outcome measure was sexual history documentation (SHD). Secondary measures were sexually transmitted infection (STI) testing and contraception provision. Statistical process control charts were used to analyze effectiveness of interventions.

RESULTS

From July 2018 through June 2019, 528 Primary Hospital Medicine encounters were included in this study and compared with published baseline data on 150 encounters. Control charts revealed a special cause increase in SHD from 60% to 82% overall, along with 37% to 73% for boys and 57% to 80% for noningestion hospitalizations. Increased SHD correlated with cycle 1 and was maintained through cycle 2. Percent STI testing significantly increased but did not shift or trend toward special cause variation. Contraception provision, length of stay, and patient relations consultations were not affected.

CONCLUSIONS

The interventions were successful in increasing SHD, including among boys and noningestion hospitalizations. The EHR enhancement maintained these increases after reminder emails were discontinued. Future interventions should specifically address STI testing and provision of contraception.

Emergency department strategies to combat the opioid crisis in children and adolescents

The opioid crisis has greatly affected not only adults but also children as well. As clinicians develop effective approaches to minimize pain and distress in children, the risks and benefits of opioids must be carefully considered. Children of parents with opioid use disorder are also at risk of living in unstable environments, performing poorly academically, engaging in future drug use, and having increased stress, which affects their development before entering adulthood. This statement focuses on the effects of the opioid crisis on children and adolescents and is intended to inform institutional policies, improve education, advocate for evidence-informed guidelines, and improve the care of children affected by the opioid epidemic who are seen in the emergency department.