Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 6 appraisals
PloS one
Enhancing the sustainability of cultural identity in science curricula through artificial intelligence as an innovative educational approach
This study investigates the role of AI as an innovative educational approach in promoting the sustainability of cultural identity within science curricula. To achieve this aim, a qualitative research design was employed, using semi-structured interviews and classroom observations as primary data collection methods. The study sample consisted of forty-seven science teachers from the Al-Qassim and Al-Ahsa regions in Saudi Arabia, selected purposively based on their experience with technology-enhanced instruction. The collected data were analyzed using systematic qualitative procedures involving open and axial coding, which facilitated the emergence of key categories related to AI integration, cultural relevance, and instructional practices. Accordingly, the analytical process is best understood as an inductive thematic qualitative approach rather than a full grounded theory methodology, as it does not include theoretical sampling, data saturation, or iterative cycles of data collection. This clarification ensures methodological transparency and alignment between design and analysis. Findings derived from both teachers' perceptions and classroom observations indicate that AI is not perceived merely as a tool for knowledge transmission, but rather as a supportive instructional mechanism that enables the contextualization of scientific concepts within local cultural settings. In addition, AI-supported instructional practices appear to foster more interactive and collaborative classroom environments, while also contributing to more responsive approaches to assessment. However, it is important to emphasize that these findings are based on teachers' reported perceptions and observed instructional practices rather than direct measurement of student learning outcomes. Despite this limitation, classroom observations suggest that integrating culturally relevant examples, analyzing cultural texts, and designing collaborative, culturally responsive learning activities may enhance student engagement and strengthen their connection to cultural identity within science learning contexts. Overall, the study highlights the potential of AI to bridge scientific knowledge and local cultural heritage in science education. It contributes to the growing body of literature on AI in education by emphasizing its role in fostering culturally responsive pedagogy and enhancing meaningful learning experiences. The study further underscores the importance of providing teachers with targeted professional development in AI integration, developing culturally responsive instructional materials, and designing assessment practices that address both conceptual understanding and cultural dimensions of learning.
24 July 2026
Read appraisal →JMIR mHealth and uHealth
Safer Sleep Guidance in Standalone UK Smartphone Apps for Parents of Newborns and Infants: Systematic mHealth App Review
BACKGROUND: Parents are increasingly using smartphone apps to help guide decision-making around aspects of caregiving, including safer sleep. However, it remains unclear whether the guidance provided aligns with best practice guidelines. OBJECTIVE: This review aimed to identify smartphone apps that provide safer sleep guidance and to assess their content, technical features, functionality, and quality. A secondary aim was to make recommendations on the improvements that existing and future apps can make to provide safer sleep guidance. METHODS: In January 2026, we searched the UK Google Play Store and Apple App Store for safer sleep guidance apps for parents and caregivers of newborns and infants (babies aged ≤12 months). Information regarding content, technical features, and functional features was extracted and summarized descriptively. Each app was evaluated by 2 raters using the Mobile App Rating Scale (MARS) and the IMS Institute for Healthcare Informatics functionality score. Adherence to best practice sleep guidance was assessed using bespoke criteria. Interrater reliability was calculated. RESULTS: A total of 1345 apps were identified, with 12 meeting the inclusion criteria. All were free to download, with 42% (5/12) containing in-app purchases. Only 17% (2/12) were affiliated with a professional health, medical, clinical, or government body. All 12 apps contained some information regarding safer sleep practices, although this was addressed and named differently (eg, calling it cot death, sudden infant death syndrome, or safe sleep). Of the 9 criteria relating to safer sleep best practice guidance, the apps mentioned an average of 6.75 (SD 1.71; range 3-9) criteria. Some promoted unsafe sleep practices (ie, around swaddling and cosleeping). The apps included an average of 5.4 of the 11 IMS functionality criteria (SD 2.5; range 2-9). The mean MARS score was 3.7 out of 5 (SD 0.43; range 2.92-4.13), and most (11/12, 92%) had a minimum acceptability score of 3 or above. Two had been formally trialed or tested. Although the language used was considered accessible (easy to understand), none of the apps allowed the language to be changed (beyond English), and very few (3/12, 25%) enabled users to change to night/dark mode, which would be considered essential when using these apps at night. CONCLUSIONS: Several apps about safer baby sleep exist. Few discuss safer sleep guidance comprehensively or raise awareness about sudden infant death syndrome and its links with unsafe sleep. All the reviewed apps met the minimum acceptability criteria for quality; however, not many were formally trialed or tested. More long-term research should be conducted on the effectiveness of safer sleep apps on adherence to safer sleep guidance. We recommend that apps relating to infant and baby safety in general include information on realistic safer sleep practices that are informed by best practice guidelines and take a risk minimization approach.
11 July 2026
Read appraisal →BMJ open
The Cyber Paranoia and Fear Scale-Updated (CPFS-U): development and implications for digital health engagement
OBJECTIVES: To update and revalidate the Cyber Paranoia and Fear Scale to reflect current technological contexts and examine its relevance to digital health readiness and engagement. METHODS: Using an online community sample (n=433), exploratory factor analysis was conducted to examine the factor structure of the revised item pool. Items were refined through consultation with Patient and Public Involvement and Engagement groups to ensure contemporary relevance and clarity. RESULTS: Analysis supported a four-factor structure representing artificial intelligence (AI) and digital dependence, technological risk awareness, perceived data vulnerability and surveillance-related mistrust. The updated Cyber Paranoia and Fear Scale-Updated (CPFS-U) demonstrated good internal consistency and supported construct validity. Cyber-paranoia and fear were conceptually and empirically distinct from general paranoia and anxiety, highlighting the specific cognitive and emotional responses elicited by digital technologies. CONCLUSIONS: The CPFS-U offers a psychometrically robust, modernised measure for understanding individuals' responses to digital and AI-based technologies. Its application in digital health research and practice can inform risk communication, user engagement strategies and the design of trustworthy digital interventions. By identifying individuals who may disengage due to online mistrust, the CPFS-U has the potential to inform more inclusive and psychologically informed digital health systems.
8 July 2026
Read appraisal →JMIR formative research
Exploring Informal Caregivers' Perception of the Olera Digital Caregiving Assistance Platform for Dementia Care: Mixed Methods Evaluation Study
BACKGROUND: Informal caregivers of people living with dementia often experience high rates of caregiver burnout while providing care. Although there are many websites and mobile apps available to help caregivers, many do not use digital tools. The Olera platform was developed to be an easily adoptable web-based support tool, connecting caregivers with long-term services and supports, financial assistance, and educational resources. The platform was developed based on the Build-Measure-Learn framework with input from caregiver needs assessments and usability studies. OBJECTIVE: This study aims to evaluate the quantitative and qualitative feedback of informal caregivers of people living with dementia on the second iteration of the Olera platform. The primary objective was to assess caregivers' acceptance of this caregiving platform. The secondary objective was to use qualitative methods to explore (1) the study cohort's challenges in daily caregiving to determine and compare them with prior literature, (2) their experience when using the Olera platform, and (3) their attitudes toward integrating artificial intelligence in caregiver services for future studies and platform development. METHODS: Caregivers were recruited through various sources and screened for eligibility through an initial survey. Participants used the platform for 4 weeks and completed a survey with an adapted Technology Acceptance Survey (TAS) and qualitative open-ended questions at the end of the testing period. TAS responses were summarized with descriptive statistics, while ANOVAs, t tests, and linear regressions were used to compare the differences in the overall TAS scores by caregiver characteristics. Qualitative feedback data on the platform's usefulness were analyzed via a thematic analysis framework approach. RESULTS: A total of 65 caregivers in the United States completed the study, with a mean age of 59.9 (SD 9.8) years. The majority were female (61/65, 95.3%), non-Hispanic or Latino White (45/65, 69.2%), and the adult child of their care recipient (42/65, 64.6%). Evaluation of the Olera platform showed a high acceptance rate, with each TAS item scoring above 5.0 and an overall TAS score of 5.83 (SD 0.85) out of 7. Higher platform use frequency was associated with higher TAS ratings in technology acceptance (F3,61=7.88, P<.001). Thematic analyses elicited the caregiving challenges, evaluation of the Olera platform, and feedback on artificial intelligence-assisted support. CONCLUSIONS: The Olera platform is an example of a beneficial web-based tool, though key features were requested to be included in the next iteration. Additionally, data supported prior findings regarding informal caregiver challenges and the insufficiency of conventional support mechanisms, indicating a need for more innovative digital solutions. Future research and development efforts using the Build-Measure-Learn approach are necessary to further iterate the platform's key features, enhance the tool, involve more informal caregivers in its improvements, and serve as a model for customizable, person-centered online care support.
6 July 2026
Read appraisal →JMIR cancer
Large Language Models for Breast and Cervical Cancers Communication: Mixed Methods Evaluation Study Assessing Linguistic Quality, Safety, and Accessibility
BACKGROUND: Effective communication about breast and cervical cancers remains a public health challenge, with widespread misinformation and barriers to cancer-related language understanding. Large language models (LLMs) offer potential for scalable health communication, yet trade-offs between quality, safety, and accessibility of general-purpose and medical-domain LLMs remain underexplored. OBJECTIVE: This study aimed to propose a comprehensive evaluation framework and systematically assess the performance of LLMs in generating breast and cervical cancer information, with a focus on linguistic quality, safety and trustworthiness, and communication accessibility and affectiveness. METHODS: This mixed methods evaluation study assessed outputs from 5 general-purpose and 3 medical LLMs using real-world breast and cervical cancer-related questions curated from publicly available medical datasets. LLM-generated responses were evaluated in a controlled offline setting. Primary outcomes included linguistic quality (fluency, coherence, and accuracy), safety and trustworthiness (toxicity, bias, and harm potential), and communication accessibility and affectiveness (readability, empathy, and clarity). Qualitative ratings were performed by domain experts, while quantitative metrics were compared across models. Statistical analyses included Welch ANOVA to detect differences in metric scores, Games-Howell tests for pairwise comparisons, and Hedges g to assess effect sizes. RESULTS: General-purpose LLMs, particularly Llama 3 and Gemma, demonstrated superior linguistic quality and affectiveness but often produced complex outputs that may limit accessibility. In contrast, medical LLMs (eg, MedAlpaca and BioMistral) generated simpler content suitable for broader audiences but scored lower in safety and empathy due to higher levels of hallucination, bias, and toxicity. CONCLUSIONS: While LLMs show promise for improving digital cancer communication, our findings reveal a trade-off between domain specialization and overall communication quality and safety. Future development of health-focused LLMs should prioritize hybrid modeling strategies to enhance trust, clarity, and clinical relevance in patient-facing tools.
29 June 2026
Read appraisal →eLife
Disinformation elicits learning biases.
In open societies, disinformation is often considered a threat to the very fabric of democracy. However, we know little about how disinformation exerts its impact, especially its influence on individual learning processes. Guided by the notion that disinformation exerts its pernicious effects by capitalizing on learning biases, we ask which aspects of learning from potential disinformation align with ideal 'Bayesian' principles, and which exhibit biases deviating from these standards. To this end, we harnessed a reinforcement learning framework, offering computationally tractable models capable of estimating latent aspects of a learning process as well as identifying biases in learning. In two experiments, participants completed a two-armed bandit task, where they repeatedly chose between two lotteries and received outcome-feedback from sources of varying credibility, who occasionally disseminated disinformation by lying about true choice outcome (e.g., reporting non-reward when a reward was truly earned or vice versa). Computational modelling indicated that learning increased in tandem with source credibility, consistent with ideal-Bayesian principles. However, we also observed striking biases reflecting divergence from idealized Bayesian learning patterns. Notably, in one experiment individuals learned from sources that should have been ignored, as these were known to be fully unreliable. Additionally, the presence of disinformation elicited exaggerated learning from trustworthy information (akin to jumping to conclusions) and exacerbated a normalized measure of 'positivity bias' whereby individuals self-servingly boost their learning from positive, relative to negative, choice feedback. Thus, in the face of disinformation we identify specific cognitive mechanisms underlying learning biases, with potential implications for societal strategies aimed at mitigating its harmful impacts.
28 June 2026
Read appraisal →