Clinical Reasoning of a Generative Artificial Intelligence Model Compared With Physicians.

This cross-sectional study assesses the ability of a large language model to process medical data and display clinical reasoning compared with the ability of attending physicians and residents.

Accuracy of a Generative Artificial Intelligence Model in a Complex Diagnostic Challenge.

This study assesses the diagnostic accuracy of the Generative Pre-trained Transformer 4 (GPT-4) artificial intelligence (AI) model in a series of challenging cases.

How Design Thinking and Quality Improvement Can Be Integrated into a "Human-Centered Quality Improvement" Approach to Solve Problems in Perinatology.

Both quality improvement (QI) and design thinking (DT) methodologies have their unique strengths and weaknesses. Although QI sees problems through a process-centered lens, DT leverages a human-centered approach to understand how people think, behave, and act when encountering a problem. By integrating these 2 frameworks, clinicians have a unique opportunity to rethink how to solve problems in health care by elevating the human experience and putting empathy back at the center of medicine.

Reach, engagement and effectiveness of in-person and online lifestyle change programs to prevent diabetes.

BACKGROUND: COVID-19 has accelerated interest in and need for online delivery of healthcare. We examined the reach, engagement and effectiveness of online delivery of lifestyle change programs (LCP) modelled after the Diabetes Prevention Program (DPP) in a multistate, real-world setting. METHODS: Longitudinal, non-randomized study comparing online and in-person LCP in a large multistate sample delivered over 1 year. Sample included at-risk adults (n = 26,743) referred to online (n = 9) and in-person (n = 11) CDC-recognized LCP from a multi-state registry (California, Florida and Colorado) between 2015 and 2018. The main outcome was effectiveness (proportion achieving > 5% weight loss) at one-year. Our secondary outcomes included reach (proportion enrolled among referred) and engagement (proportion ≥ 9 sessions by week 26). We used logistic regression modelling to assess the association between participant- and setting -level characteristics with meaningful weight loss. RESULTS: Online LCP effectiveness was lower, with 23% of online participants achieving > 5% weight loss, compared with 35% of in-person participants (p < 0.001). More adults referred to online programs enrolled (56% vs 51%, p < 0.001), but fewer engaged at 6-months (attendance at ≥9 sessions 46% vs 66%, p < 0.001) compared to in-person participants. CONCLUSIONS: Compared to adults referred to in-person LCP, those referred to online LCP were more likely to enroll and less likely to engage. Online participants achieved modest meaningful weight loss. Online delivery of LCP is an attractive strategy to deliver and scale DPP, particularly with social distancing measures currently in place. However, it is unclear how to optimize delivery models for maximal impact given trade-offs in reach and effectiveness.

Participation and weight loss in online National Diabetes Prevention Programs: a comparison of age and gender subgroups.

Early onset diabetes has adverse transgenerational effects, yet in-person National Diabetes Prevention Programs (NDPPs) have low reach among adults of peak reproductive age. We examined participation and weight loss with online NDPPs for younger versus older adults. Solera Health, Inc., collected data from 12,966 adults who enrolled in a yearlong online NDPP from 2015 to 2018. We used general linear models and logistic regression to assess differences between younger and older adults (<45 vs. ≥45 years) in session initiation (logging in), session completion (activities approximating intensity of in-person classes), and weight loss, overall and according to engagement thresholds. Almost all (N = 12,497, 96%) individuals who enrolled initiated ≥1 session(s), but fewer (N = 2,408, 19%) completed ≥4 sessions over ≥9 months, achieving 4.5% weight loss on average. Among all enrollees with ≥2 weights (N = 10,161), younger men and women lost less weight (1.8% and 1.7%, respectively) than older men (3.3%) and women (2.7%; all p < .05). Among all enrollees who completed ≥4 sessions over ≥9 months, weight loss did not differ between older men (4.3%), older women (4.0%), and younger men (3.5%), but younger women achieved less weight loss (3.0%) than older adults (all p < .001). Online programming supports NDPP reach and weight loss, although younger adults completed fewer sessions and young women achieved less weight loss than older adults. Efforts to increase ongoing engagement among younger adults are needed to prevent early onset of diabetes and adverse transgenerational effects.

New, combined, and reduced dosing treatment protocols cure Trypanosoma cruzi infection in mice.

The development of treatment protocols with reduced toxicity and equivalent or improved efficacy for Trypanosoma cruzi infection is a priority. We tested the effectiveness of benznidazole (BZ), nifurtimox (NFX), other prospective drugs in intermittent and combined treatment protocols to cure T. cruzi infection initiated with susceptible and drug-resistant parasite strains. A 40-day course of BZ, NFX, or the oxaborale AN4169 cured 100% of mice, whereas posaconazole (POS), and NTLA-1 (a nitro-triazole) cured approximately 90% and 20% of mice, respectively. Reducing the overall dosage of BZ or NFX by using an intermittent (once every 5 days) schedule or combining 5 daily doses of POS with 7 intermittent doses of BZ also provided approximately 100% cure. T. cruzi strains resistant to BZ were also found to be resistant to other drugs (POS), and extending the time of treatment or combining drugs did not increase cure rates with these isolates. Thus, dosing schedules for anti-T. cruzi compounds should be determined empirically, and compounds targeting different pathways may be combined to yield effective therapies with reduced toxicity. This work also suggests that standard treatment protocols using BZ and NFX may be significantly overdosing patients, perhaps contributing to the adverse events.