2026-04-01
Jianchen Luo, Jing Ma, Danni Zhan, Yuhong Zhou, Jiayu Li, Lan Zhang, Wentao Wang
Background: Mental health problems among university students are a growing global concern, yet limited counseling resources and inadequate understanding of counseling procedures often delay timely help-seeking. Informed consent forms (ICFs) are essential for safeguarding autonomy and clarifying counseling procedures, but many universities’ counseling ICFs are incomplete, ambiguous, or overly technical. Large language models (LLMs) may offer scalable assistance for improving clarity and accessibility. Objective: This study aimed to evaluate whether LLM-based rewriting could improve the structure, readability, content quality, and comprehensibility of university counseling ICFs, and compared 2 advanced models (ChatGPT [GPT-5] and Grok-4). Methods: We conducted a comparative evaluation of counseling ICFs collected from 33 Chinese universities (original texts) and generated 2 rewritten versions for each ICF using ChatGPT (GPT-5) and Grok-4. A multidimensional framework assessed (1) textual structure and readability, (2) expert-rated content quality from a counselor perspective, and (3) volunteer-rated reading comprehension from a client perspective. Comparisons between original and rewritten texts were performed using Wilcoxon signed rank tests, with linear mixed-effects models used to validate results while accounting for rater variability. Results: Compared with the originals, both LLM-rewritten ICFs showed significant improvements across all evaluated dimensions. The mean Lee-Yang Readability Index decreased from 28.68 (SD 5.69) to 22.39 (SD 2.13) with ChatGPT (GPT-5) and 24.37 (SD 2.32) with Grok-4 (both <.001), and mean tone friendliness increased from 2.57 (SD 0.29) to 2.67 (SD 0.12) and 2.67 (SD 0.13), respectively. The mean expert-rated content quality improved from 45.33 (SD 8.74) to 52.54 (SD 7.92) and 55.49 (SD 7.81) (<.001), driven mainly by higher completeness and specificity of key information. The mean volunteer-rated reading comprehension scores increased from 19.02 (SD 1.32) to 22.33 (SD 0.81) and 22.05 (SD 0.90) (<.001), indicating improved clarity, readability, and acceptability. Across structural features, Grok-4 tended to produce longer rewritten forms than the originals, highlighting a potential trade-off between added informational content and document length. Conclusions: In this comparative evaluation of 33 Chinese university counseling ICFs, LLM-based rewriting was associated with improved readability, expert-rated content quality, and volunteer-rated comprehension relative to original forms. These findings suggest that LLMs can support the optimization of counseling documentation; however, implementation should consider practical constraints (eg, document length) and retain human oversight.
2026-04-01
Lu Ji, Xinke Zhou, Lan Yao
Background: Improving screening coverage is a central goal of the global strategy to eliminate cervical cancer. In resource-constrained settings, insufficient service accessibility remains a key barrier to expanding coverage. Supported by artificial intelligence (AI)–assisted diagnostic technology, Hubei province has pioneered China’s first provincial-level population-wide cervical cancer screening program, serving 12.67 million eligible women. This initiative provides an innovative practice for addressing such challenges. Objective: This study systematically examines major factors influencing the achievement of universal screening coverage targets through interviews with core managers and implementers of Hubei province’s screening program. It aims to provide empirical evidence and strategic recommendations for applying AI technologies in cervical cancer screening and enhancing screening coverage rates. Methods: The interview guide was developed under the guidance of the macro model of health system. A combination of purposive sampling and multistage stratified sampling was used to capture provincial-level overviews and understand regional implementation variations, respectively. Guided by the macro model of health system, interview outlines were developed. Semistructured interviews were conducted between January and August 2024 with key project personnel (one per institution) from 14 relevant institutions. Interview data were analyzed using thematic analysis, with systematic coding and management facilitated by the NVivo software. Results: Key informants reported that comprehensive screening has been largely achieved. The analysis identified government stewardship, AI-assisted screening technology, screening funding, and health literacy as the major factors for achieving universal screening coverage. Among these, government leadership and the application of AI-assisted diagnostic technologies provide significant driving factors. Additional factors encompassed structural dimensions, including multisectoral coordination, trained screening technicians, and information systems; process dimensions, such as institutional service delivery capacity, quality control measures, and community mobilization; along with outcome dimensions comprising population coverage, cytology positivity rate, follow-up, and treatment rate. Conclusions: Achieving large-scale cervical cancer screening requires coordinated efforts across four dimensions: government stewardship, screening technology, screening funding, and health literacy. Government stewardship served as the core driver in advancing population-wide screening coverage. Its mechanisms included coordinated procurement of AI-assisted screening services, secured financial investment, formulation of targeted policies, promotion of multi-sectoral collaboration, and optimization of service delivery models. These efforts systematically improved the accessibility and utilization of screening services, ultimately encouraging and facilitating active participation among residents.
2026-04-01
Nicky Lewis, Jennifer Reynolds, Diane Krause, Philip M Reeves, Jamie Luster, Michael D Stein, Sharon L Walsh, Amy Farmer, Michelle R Lofwall, Monica F Roberts, Hilary L Surratt, Brooke N Crockett, Kara Stephens, Kelli Bursey, Kristin Mattson, Michael D Slater
Background: Research on the effectiveness of digital health campaign strategies is lacking. Understanding performance outcomes is essential for the successful implementation of campaigns. Two studies examined platforms, tactics, and content of digital health campaigns using paid media performance data. Objective: This analysis compared 2 digital advertising methods (social media and banner or display) using click-through rate (CTR) and cost-per-click (CPC) as performance measures. Performance differences by state, community type, message approach, format, and image type were assessed. CTR and CPC served as measures in determining performance differences between social media and banner or display. Methods: This cross-sectional secondary analysis examined campaign performance for the HEALing (Helping to End Addiction Long-Term) Communities Study, which served 85,875,105 impressions. Data were collected from media buy reports, entered into templates that included method (display or banner and social media) and key performance indicators (impressions, clicks, and media spend), and CTR and CPC were calculated. Study 1 assessed differences in CTR and CPC for social media and banner or display by state (KY, NY, MA, and OH) and community type (urban and rural). Study 2 assessed differences in CTR for social media and banner or display by state (KY, NY, MA, and OH), community type (urban and rural), message approach (testimonial and information-based), format (motion graphic or graphics interchange format, video, and static image), and image type (local and stock). Separate analyses were conducted for each advertising method. Results: Study 1 found significant differences between advertising methods, where social media had higher CTR compared to banner or display. Social media had a significant main effect for state, where OH had the highest CTR. There was a statistically significant difference in CPC based on advertising method, where social media had a lower CPC compared to banner or display. Social media had a significant main effect for state, where OH had the lowest CPC. Banner or display had a significant main effect for state and community type, where OH and urban communities had the highest CPC. Study 2 found significant differences between advertising methods, where social media had higher CTR than banner or display. For social media, urban communities, static format, and local spokespersons had the highest CTR. There were significant differences between all pairs of states, where OH had the highest CTR. For display or banner, static format and local spokespersons had the highest CTR. Conclusions: This analysis provides guidance for digital health campaigns. It examined the performance of opioid use disorder campaigns using CTR and CPC measures, demonstrating utility in future campaign evaluations. Social media was more related to stimulating responses to campaign messages compared to banner or display. State-to-state variations emphasized the importance of message pilot testing. Using local spokespersons versus stock spokespersons is recommended. Trial Registration: ClinicalTrials.gov NCT04111939; https://clinicaltrials.gov/study/NCT04111939
2026-04-01
Jihye Lee, Juyoung Park, Yuna Kim, Duk-Jo Kong
Background: Preventive campaigns for older adults must decide how to allocate limited resources across media channels. However, these channel allocation and budget decisions rarely use explicit criteria for distributional equity or structured strategic planning tools. Consequently, health systems may optimize average uptake while leaving large gaps across socioeconomic groups and media use profiles. Objective: This study aimed to develop and apply a data-driven agent-based model as a strategic planning tool for preventive campaigns targeting older adults, comparing channel allocation, personalization, and loss framing options under explicit budget and equity guardrails. Methods: We built an agent-based model calibrated to national survey data from South Korea on influenza vaccination and routine health screening among older adults (vaccination, N=2405; screening, N=2400). Fifteen prespecified campaign scenarios varied channel allocation across television, digital, and print media; budget intensity; 2 equity-focused personalization strategies; and graded loss framing. Primary outcomes were final adoption and time to adoption. Equity outcomes included the minimum class-level adoption and 90‐10 gap across latent classes. Each scenario was simulated over 12 monthly steps with 100 Monte Carlo replications. We conducted sensitivity analyses varying link functions and key social reinforcement parameters. Results: Personalization improved uptake and equity relative to the integrated baseline. In the vaccination model (N=2405), adoption increased from 91.2% (n=2193) to 93.3% (n=2244) and 94.6% (n=2275). Minimum class-level adoption increased from 86.8% to 90.3% and 90.9%. The 90‐10 gap narrowed from 5.7 to 4.5 and 4.7 percentage points. In the screening model (N=2400), adoption increased from 83.8% (n=2011) to 88.2% (n=2117) and 89.5% (n=2148). Minimum class-level adoption increased from 77.6% to 83.2% and 85.3%. The 90‐10 gap narrowed from 9.2 to 7.4 and 6.2 percentage points. Television-only strategies achieved high adoption but had less favorable equity profiles than personalization. High-budget strategies achieved high adoption but required higher total exposure. Stronger loss framing produced small, monotonic gains in adoption and shortened the time to adoption without worsening equity in the tested range. Scenario rankings were stable in sensitivity analyses. Conclusions: This agent-based modeling study illustrates how ex ante planning can improve preventive campaign design by comparing channel allocation and personalization options under explicit equity and budget criteria. For campaigns targeting older adults, equity-focused reweighting and class-tailored television-digital portfolios improved or preserved mean adoption while strengthening distributional equity under fixed budgets. In contrast, undifferentiated channel diversification without personalization offered a less favorable efficiency-equity trade-off. These findings support integrating explicit equity guardrails into early-stage channel allocation and prioritizing targeted personalization over simple channel diversification. Future work should validate these patterns in other populations and health systems and link simulated diffusion trajectories with observed exposure and engagement in real-world campaigns. It should also extend guardrail-based planning tools to organizational settings and multiyear decision contexts.
2026-04-01
Claire Eades, Anh Nguyen-Hoang, Louise Hoyle, Dawn Cameron, Josie MM Evans
Background: Previous gestational diabetes incurs an 8-fold risk of developing type 2 diabetes, but lifestyle change can prevent or delay progression. Technology-based interventions may help overcome challenges women face in making postpartum lifestyle changes. Objective: This study aimed to assess whether technology-based diabetes prevention interventions improve outcomes related to the onset of type 2 diabetes among women with a previous diagnosis of gestational diabetes. Methods: Cochrane Central Register of Controlled Trials, CINAHL, Embase, PsycINFO, and Midwives Information and Resource Service were searched to October 2025 using subject headings and free-text terms. Titles and abstracts were independently screened by 2 authors, as were retrieved full-text articles. Studies were eligible if they examined technology-based diabetes prevention interventions delivered between gestational diabetes diagnosis and any time post partum, assessing anthropometric outcomes, glycemic control, health behavior, or psychological outcomes. Risk of bias was assessed by 1 reviewer using the National Institute for Clinical Excellence checklist, and certainty of evidence was assessed by 2 reviewers using the Grading of Recommendations Assessment, Development, and Evaluation. Data were summarized narratively, and results were pooled, where possible, using a random effects model. Results: This review identified 15 studies, including 1257 participants. Pooled analysis of 7 studies showed significantly greater weight loss among those receiving technology-based interventions (mean difference –1.01, SE 0.35, 95% CI –1.86 to –0.16 kg; P =.03). Interventions delivered using technology only showed increased weight loss (mean difference –1.13, 95% CI –3.12 to 0.86 kg) as did those with a longer follow-up (mean difference –1.58, 95% CI –3.93 to 0.76 kg) compared with combined technology and telemedicine approaches (mean difference –0.89, 95% CI –2.51 to 0.73 kg) and studies with shorter follow-up (mean difference –0.7, 95% CI –1.21 to –0.18 kg), but these differences were not significant (mode of delivery: χ 2 1 =0.08; P =.78; follow-up: χ 2 1 =1.06; P =.30). Meta-analysis showed no significant differences in BMI (mean difference –0.22, SE 0.1, 95% CI –0.4 to –0.01 kg/m 2 ; P =.27; n=2 studies), fasting glucose (mean difference –0.03, SE 0.16, 95% CI –0.49 to 0.49 mmol/L; P =.99; n=4 studies), 2-hour glucose (mean difference 0.12, SE 0.19, 95% CI –0.47 to 0.72 mmol/L; P =.56; n=4 studies), hemoglobin A 1c (mean difference –0.01%, SE 0.02%, 95% CI –0.24% to 0.23%; P =.74; n=2 studies), or homeostasis model assessment of insulin resistance (mean difference 0.07, SE 0.02, 95% CI –0.16 to 0.31; P =.16; n=2). Certainty of evidence for all pooled outcomes was very low. Conclusions: Technology-based interventions may help support women in reducing their risk of type 2 diabetes following gestational diabetes mellitus, but substantial heterogeneity, significant risk of bias, and very low certainty in the evidence mean that the findings should be interpreted cautiously. Trials with larger samples and longer follow-up are required to draw firm conclusions. Trial Registration: PROSPERO CRD42024324019; https://www.crd.york.ac.uk/PROSPERO/view/CRD42024324019
2026-03-31
Boon-How Chew