Patients with obstructive sleep apnea (OSA) frequently seek information online, yet the comparative quality of content delivered by web search engines versus generative AI systems is unclear. This study evaluated how different digital information sources perform in answering common patient questions about OSA. Thirty high-volume, patient-facing OSA questions were identified using Google Trends. Each question was submitted verbatim to four general-purpose large language models (GPT-4, GPT-5, DeepSeek, Mistral), a medically specialized retrieval-augmented model (OpenEvidence), and Google Search. Seven otolaryngologists with clinical experience in OSA independently rated each response for accuracy, clarity, completeness, relevance, and usefulness using a five-point rubric. Composite and domain scores were analyzed using one-way analysis of variance with multiple-comparison correction; inter-rater reliability was assessed with two-way random-effects intraclass correlation coefficients. A total of 180 question–system pairs received 6295 domain-level ratings. OpenEvidence achieved the highest mean composite score (4.33), followed by a tightly clustered group of LLMs (means 4.00–4.04). Google Search scored significantly lower (3.15). Differences among systems were statistically significant across all domains (p < 0.001), with large effect sizes for comparisons of OpenEvidence and general LLMs versus Google. Composite average-rater reliability was good (ICC = 0.70). For common OSA questions, generative AI systems—particularly a retrieval-augmented medical model—produced higher-quality patient-facing information than standard web search. These findings support cautious consideration of GenAI tools to supplement patient education in OSA, while underscoring the need for ongoing evaluation across diseases, disciplines, and patient populations. Patients with obstructive sleep apnea (OSA) frequently rely on online sources such as Google Search to understand symptoms, testing, and treatment, yet the quality of patient-facing information varies widely. As generative artificial intelligence tools are increasingly used for health questions, their comparative performance for OSA education has not been systematically evaluated using blinded expert review. In this blinded comparative study, generative AI systems, particularly a retrieval-augmented medical model, provided more accurate, clear, complete, and useful answers to common OSA questions than standard web search. These findings highlight that the choice of digital information source can meaningfully influence the quality of patient education in sleep medicine and support further evaluation of AI tools within clinical practice.
Abstract Objective To examine walking behavior during pregnancy and its association with obstetric and neonatal outcomes. Methods This retrospective cohort study included pregnant patients who delivered between 2020 and 2024 at a tertiary medical center. Daily step counts were recorded via the Health Auto Expert mobile application from three months before conception until admission to the labor ward. Pregnancy and delivery data were obtained from medical records. The primary outcome was a composite of delivery complications by step-count category: Above-Average and Below-Average, based on the cohort mean. Secondary outcomes included postpartum complications, mode of delivery, and neonatal outcomes. Results The mean daily step count during pregnancy was 2539, declining progressively with gestation. Of 176 patients, 71 (40%) were Above-Average and 105 (60%) Below-Average. The Above-Average group had higher pre-pregnancy activity (4541.7 vs 2330.3 steps/day, p < 0.001) and higher maternal age. No significant differences were found in composite delivery or postpartum complications. Vacuum extraction was more frequent in the Above-Average group (15.5% vs 3.8%, p = 0.014), though not significant after adjustment. Median neonatal birth-weight percentile was 54.5, lower in the Above-Average group (44.0 [25.0–71.0] vs 60.0 [38.0–78.0], p = 0.024), without significance in multivariable analysis. Conclusions Walking patterns declined across pregnancy. Step count was not associated with delivery complications or adverse maternal or neonatal outcomes. Walking during pregnancy appears safe and was not associated with adverse maternal or neonatal outcomes in this predominantly sedentary cohort of low-risk pregnancies.
Malignancy-associated hemophagocytic lymphohistiocytosis (mHLH), a hyperinflammatory syndrome, has poor prognosis and no standard therapy. Emapalumab, a fully human monoclonal antibody that neutralizes the proinflammatory cytokine interferon-gamma, is approved for treating primary hemophagocytic lymphohistiocytosis (pHLH). REAL-HLH, a retrospective chart review conducted across 33 US hospitals, evaluated real-world treatment patterns and outcomes in patients treated with at least 1 dose of emapalumab between November 20, 2018, and October 31, 2021. Data are presented for the subset of patients with mHLH. Overall, 51/105 (48.6%) patients did not meet the pHLH classification criteria and were categorized as presenting with secondary HLH; 17/51 patients had underlying malignancy (mHLH). At HLH diagnosis, median age (range) was 15.0 (3.0-27.0) years, 6/14 (42.9%) patients with available data had a positive Optimized HLH Inflammatory index indicating pathologic inflammation; 9/17 (52.9%) had infections, and 10/17 (58.8%) received emapalumab in an intensive care unit. Emapalumab was primarily initiated for treating refractory (10/17; 58.8%) or progressive (3/17, 17.7%) disease. Most patients received HLH-related therapies before (16/17; 94.1%) and/or concurrent with (15/17; 88.2%) emapalumab. Most key laboratory parameters improved, and some (fibrinogen [11/13; 84.6%], absolute neutrophil count [6/10; 60%], and CXCL9 [7/8; 87.5%]), normalized or stabilized per physician assessment, following treatment with emapalumab-containing regimens. Overall survival at the end of follow-up and 12-month survival probability from emapalumab initiation were 23.5% and 22.1%, respectively. In conclusion, emapalumab-containing regimens improved or normalized most laboratory parameters in patients with mHLH. Future studies are warranted to establish appropriate emapalumab dosing and utility in this high-risk population.
CONTEXT:Glucagon-like peptide 1 receptor agonists (GLP-1RAs) are increasingly prescribed to older adults with type 2 diabetes for their metabolic and cardiovascular benefits, but their effects on bone health in this high-risk population are not well established. OBJECTIVE:The aim of this study was to assess the risk of fragility fractures associated with GLP-1RA therapy compared to sodium-glucose cotransporter-2 inhibitors or dipeptidyl peptidase-4 inhibitors in older adults with type 2 diabetes. METHODS:We conducted a population-based retrospective new-user cohort study using the nationwide Clalit Health Services database. Participants included people with type 2 diabetes aged 65 years or older who initiated GLP-1RA therapy or comparator drugs during 2018 to 2022. The primary outcome was first fragility fracture. Follow-up continued until fracture occurrence, death, or March 31, 2024. An intention-to-treat approach was used. Baseline characteristics were balanced using inverse probability of treatment weighting based on propensity scores. Hazard ratios (HRs) were estimated using multivariate survival analysis accounting for the competing risk of death. RESULTS:The study included 46 177 participants: 11 257 GLP-1RA users and 34 920 comparator drug users. During a median follow-up of 34.7 months, 4086 (8.8%) individuals experienced a fragility fracture and 4595 (10.0%) died. In the propensity score-weighted model adjusted for baseline characteristics and competing risk of death, new users of GLP-1RAs had an 11% increased fracture risk compared to the comparator group (HR 1.11; 95% CI, 1.01-1.21). CONCLUSION:Initiating GLP-1RA therapy was associated with a modestly increased risk of fragility fractures in older adults with type 2 diabetes. These findings may inform clinical decisions by balancing clinical benefits against potential skeletal risks in this vulnerable population.
