Luo Z, et al. World J Mens Health. 2025 Sep;43:e70. https://doi.org/10.5534/wjmh.250144

Large language models (LLMs) are increasingly used to draft public-facing HIV materials, but the balance between readability and information quality remains unclear. This study provides a methodological, content-level evaluation of unedited LLM outputs. Fifteen high-frequency, HIV-relevant lay queries were drawn from Google Trends (using “AIDS” only as the entry term for salience) and mapped to MeSH. From March 13–20, 2025, each query was posed to ChatGPT-4, DeepSeek-R1, and Grok-3 across three conversational rounds (15 × 3 × 3 = 135 texts). Readability of raw outputs was computed using ARI, FRE, GFI, FKGL, and TSI. Information quality was rated with DISCERN by two non-specialist clinicians for all texts and by two infectious-disease specialists for a randomized subset (45 texts). Inter-rater reliability used two-way random-effects ICC; group differences used one-way ANOVA with Tukey’s post hoc or Kruskal–Wallis with Dunn’s post hoc, as appropriate. ARI and GFI showed no between-model differences, whereas FRE, FKGL, and TSI indicated that Grok-3 required higher reading levels than ChatGPT-4 and DeepSeek-R1. In non-specialist ratings, Grok-3 scored higher on DISCERN (56.5 ± 7.7) than ChatGPT-4 (36.3 ± 5.2) and DeepSeek-R1 (36.3 ± 4.9) (both p < 0.0001). Among specialists, Grok-3 (71 [69–73]) exceeded ChatGPT-4 (43.5 [38–49], p < 0.0001) and DeepSeek-R1 (54 [49–59], p = 0.0006); ChatGPT-4 vs. DeepSeek-R1 was not significant (p = 0.2055). ICC among non-specialists was excellent for DeepSeek-R1 (0.95) and Grok-3 (0.98) but moderate for ChatGPT-4 (0.58). Among specialists, ICCs were moderate (≈ 0.59–0.72) with wide confidence intervals. A consistent readability–quality tension emerged: outputs rated higher for information quality tended to be harder to read, whereas more readable texts were less complete. These findings suggest that LLMs may be best suited for upstream use as drafting aids, paired with expert review for accuracy, stigma sensitivity, and audience fit, then edited to target literacy bands and localized, rather than as direct-to-patient systems. This methodological study does not assess reach, behavior change, or clinical outcomes; future work should test comprehension and actionability in target populations and explore low-bandwidth, localized deployments to determine whether content improvements translate into meaningful public-health benefits.

Lee GH, et al. World J Mens Health. 2023 Oct;41:e79. https://doi.org/10.5534/wjmh.230122

Background: The association between the triglyceride-glucose (TyG) level, its obesity-related derivatives, and the occurrence of endometriosis (EMS) remains ambiguous, particularly in individuals with higher levels of education. This study sought to explore the relationship between TyG, its obesity-related derivatives, and EMS across various educational backgrounds. Methods: This study utilized a substantial dataset obtained from four cycles of the National Health and Nutrition Examination Survey (NHANES) conducted between 1999 and 2006. To explore the relationship between TyG, its obesity-related derivatives, and EMS, we employed a variety of analytical methods, including multivariable logistic regression models, smooth curve fitting, threshold effect analysis, and subgroup analysis, which were applied to participants with varying educational levels. Results: Among the 2347 participants, 203 (8.65%) were diagnosed with EMS. In the overall population, only the TyG, TyG-waist-to-height ratio (TyG-WHtR), and TyG-waist circumference (TyG-WC) variables demonstrated a positive association with EMS. However, within the group with high educational attainment, TyG, TyG-WHtR, TyG-WC, and TyG-body mass index (TyG-BMI) all exhibited positive correlations with EMS. These associations remained robust after adjustment for multiple potential confounding variables. The subgroup analysis demonstrated that these associations were consistent across different subgroups (p > 0.05). Furthermore, both linear and nonlinear relationships were observed between TyG and its obesity-related derivatives and EMS, as evidenced by the smooth curve fittings and threshold effect analyses. In contrast, no significant associations were identified in the group with lower levels of education. Conclusions: Our study suggests that there is variation in the association between TyG and its obesity-related derivatives and EMS across different educational levels, warranting further investigation. In individuals with higher education, elevated levels of TyG and its obesity-related derivatives were associated with a higher prevalence of EMS. Conversely, this correlation was not observed among those with lower educational levels.

Lee GH, et al. World J Mens Health. 2024 Oct;42:e94. https://doi.org/10.5534/wjmh.240071

Integrating large language models (LLMs) with virtual patient platforms offers a novel approach to teaching clinical reasoning. This study evaluated the performance and educational value of combining Body Interact with two AI models, ChatGPT-4 and DeepSeek-R1, across acute care scenarios. Three standardized cases (coma, stroke, trauma) were simulated by two medical researchers. Structured case summaries were input into both models using identical prompts. Outputs were assessed for diagnostic and treatment consistency, alignment with clinical reasoning stages, and educational quality using expert scoring, AI self-assessment, text readability indices, and Grammarly analysis. ChatGPT-4 performed best in stroke scenarios but was less consistent in coma and trauma cases. DeepSeek-R1 showed more stable diagnostic and therapeutic output across all cases. While both models received high expert and self-assessment scores, ChatGPT-4 produced more readable outputs, and DeepSeek-R1 demonstrated greater grammatical precision. Our findings suggest that ChatGPT-4 and DeepSeek-R1 each offer unique strengths for AI-assisted instruction. ChatGPT-4’s accessible language may better support early learners, whereas DeepSeek-R1 may be more aligned with formal clinical reasoning. Selecting models based on specific teaching goals can enhance the effectiveness of AI-driven medical education.