[Existential suffering and autonomy: Contributions to the debates on euthanasia and medically assisted suicide in Argentina].

In the United States (US), one in five youth has obesity and youth from historically underserved communities are disproportionately affected []. Paediatric obesity has both short- and long-term physical and mental health risks [–] and poses a significant burden on the healthcare system [, ]. Lifestyle behaviour interventions are the cornerstone of treatment, but this approach only addresses some of the factors that contribute to this complex condition []. For many, successful weight loss may depend on a combination of lifestyle behaviour treatment and pharmacological or surgical treatment. Since 2020, three obesity medications (OMs) have been approved by the FDA for youth 12 years and older, including Liraglutide, Semaglutide and Phentermine/Topiramate. While previous OMs, such as Orlistat and Naltrexone/Buproprion, raised concerns about safety and effectiveness, these newer OMs have shown early success, with clinical trials demonstrating an average BMI reduction of 7% with Liraglutide (3.0 mg dose) over 56 weeks, 16% with Semaglutide (2.4 mg dose) over 68 weeks and 14% with Phentermine/Topiramate (15 mg/92 mg dose) over 56 weeks []. Despite their effectiveness, there is documented variability in outcomes from OMs in clinical trials and especially in real-world settings. In clinical trials, 73% of adolescents receiving Semaglutide [], 43% receiving Liraglutide [] and 47% receiving Phentermine/Topiramate had a BMI loss of greater than 5% []. One real-world study of EHR data for 3411 adults found that only Phentermine and Phentermine/Topiramate resulted in weight loss, and weight loss was only 2%–4% over 12 weeks []. A similar real-world study of 1720 paediatric patients on OMs for variable durations found a statistically significant but clinically modest (1.5%) weight loss only with Phentermine []. However, neither study included newer OMs, such as the GLP-1 receptor agonists Semaglutide and Liraglutide. One study of EHR data for 3555 adult patients demonstrated differential responses to Semaglutide over 52 weeks based on diagnoses, medications and sex using machine learning [], with the main findings showing that females, patients with prediabetes and patients using linaclotide had more pronounced weight loss. Despite the fact that many OMs may not be cost-effective at their current US pricing [, ], the demand for OMs continues to increase [] and disparities in access exist [, ]. Therefore, it is crucial to identify patient and treatment factors that contribute to successful outcomes with different OMs to assist providers in choosing the appropriate OM for patients and to ensure that OM treatment is delivered effectively. In developing tools to predict outcomes from OMs, machine learning methods can play a critical role. As shown in the Semaglutide study described above, machine learning methods can help us learn from large amounts of data, identify complex patterns and interactions and develop algorithms for prediction. Machine learning methods are also well-suited for non-linear multi-variable relationships and handling missing or high-dimensional data often present in EHRs.