Potential Bias Research Articles

The meta-governance of quality registries for public health

Abstract Issue/problem Healthcare-related planning and policy requires evidence, which in turn requires data. Generating sensitive, high-quality data is expensive. Given the constraints of routine or study-related data for quality monitoring, quality registries are commonly used. But when is it adequate, from a public health meta-governance point of view, to establish dedicated quality registries? Description of the problem Many countries have established public patient registries over time. The lack of a meta-governance framework complicates decision-making for establishing and funding dedicated quality registries. Simultaneously, legal advancements like the European Health Data Space may enhance the use of routine data in healthcare policy, potentially transforming both registry and meta-governance practices. Results We identified key meta-governance challenges, such as setting criteria for the creation, modification, or closure of registries, deciding on public funding and choosing between routine and specially collected data. We argue against using absolute thresholds like disease prevalence, advocating for a scorecard approach that evaluates epidemiologic reasoning, policy relevance, healthcare impact and factors like data availability and registry synergies. Lessons The EHDS will enable enhanced secondary data use, necessitating decisions on when to use routine versus specifically collected data. Decision-makers must consider sustainability, ethics, practicability and data linkability, while recognizing risks such as data quality limitations and potential bias from original data collection purposes. Key messages • The necessity for quality registries varies by case, serving specific policy questions. Optimizing funds and avoiding arbitrary decisions require a strategy based on a meta-governance framework. • With the introduction of EHDS, the focus may shift more toward data curation and quality control, making it critical to align registry efforts with the broader health information system.

Prehospital point of care ultrasound in Helicopter Emergency Medical Services: a 5-year experience study in Belgium

IntroductionThe utilization of Clinical Ultrasound (CUS) in Helicopter Emergency Medical Services (HEMS) has become increasingly prevalent, reflecting its significant role in emergency care.MethodsThis descriptive, monocentric study, conducted at the Helicopter Medical Center (CMH) of Bra-sur-Lienne, Belgium, aimed to describe the use of CUS within HEMS and evaluate its potential impact on diagnostic accuracy, therapeutic interventions, and hospital destination decisions.ResultsOver a 5-year period, from January 2018 to December 2022, our retrospective analysis included 6,126 HEMS interventions, with CUS performed in 29.55% of cases. The study demonstrated that CUS led to therapeutic changes in 30.88% of instances and affected the hospital destination in 9% of cases, despite a potential bias toward directing patients to the University Hospital Center of Liège (CHU of Liège) due to operational constraints. The agreement between prehospital CUS findings and hospital imaging was substantial, with an 80.39% concordance rate, underscoring CUS's utility in enhancing prehospital diagnostic accuracy. We found no significant increase in on-scene time due to CUS usage, with most exams completed in <5 min.DiscussionOur findings highlight the interesting role of CUS in HEMS, facilitating more informed and confident medical decision-making in prehospital emergency care. However, variability in CUS application emphasize the need for standardized guidelines to optimize its use. The study's limitations include its retrospective, monocentric design and the focus on on-scene ultrasound evaluations. Future research should aim at a more extensive, prospective analysis to further validate the benefits of CUS in emergency medical services.

Can ChatGPT affect health literacy?

Abstract Background ChatGPT is a large language model based on artificial intelligence, which nowadays finds application in various fields, including medicine. Objectives To investigate the usefulness of ChatGPT related to health literacy and to highlight its potential limitations. Methods We searched PubMed/MEDLINE for published research and articles that reviewed ChatGPT in the context of health literacy, written in English over the past five years with free full text access. All original articles were screened. Results A total of 53 publications were found. Twenty-five publications met the inclusion criteria and were included in the review. Most of the works presented a positive opinion towards the use of ChatGPT. The authors of six articles believe that ChatGPT generates more readable papers, while four of them believe the opposite. In four articles authors found out possibility to simplify text. The negative sides relate to the possibly insufficient education of users, the accuracy of the information, potential bias, dependence on the outdated data, and the ethical problem of the real authors not being cited, and therefore not recognized for their work. Conclusions According to the conducted research, ChatGPT represents an opportunity to improve health literacy by generating accessible and comprehensible information about health. It has also been shown to improve the readability and accessibility of public education materials on selected health topics. Furthermore, while AI models like ChatGPT can serve as a reliable source of health information, they should be used as an addition to professional medical consultation, and not as a standalone advice service. Key messages • ChatGPT represents an opportunity to improve health literacy. • AI models like ChatGPT can serve as a reliable source of health information, but they should be used as an addition to professional medical consultation, and not as a standalone advice service.

Field Visits as a Pedagogical Tool to Understand Health Determinants

Abstract Background Medical education often focuses on curative medicine, neglecting health determinants. Field visits, common in other educational contexts, enhance understanding of these factors. We aim to offer insights into pedagogical methods applicable to similar educational settings. Objectives Medical students completing an internship at the Gaia-Espinho Public Health Unit participated in field visits to enhance their understanding of health determinants. Guided through two economically distinct areas within a walkable distance of 2km, students documented health determinants through notes and photos. Discussions were held on-site and after visits integrating the real-world experience with theoretical frameworks. An inductive thematic analysis of students’ satisfaction and pedagogical effectiveness of the activity was performed on the internship reports from the 2023/2024 academic year. Results Out of 18 internship reports, 10 described the activity as positive and enriching, of which six highlighted it as a crucial part of the internship. Thirteen reports noted that the activity led to new perceptions about health determinants, and nine highlighted its practical value. Nine students included photos in their reports, bike paths being the most featured. Seventeen provided a clear description and listed the determinants they were able to detect. Discussions highlighted sensory experiences that are uniquely discernible through physical site visits including unpleasant odours, loud noises, and the cooling effect of blue areas on hot days. Three did not provide any evaluative comments, but correctly listed the determinants identified. Conclusions Field visits may enhance medical students’ understanding of health determinants, being recognized as an engaging pedagogical tool. Potential biases in assessments that rely on student satisfaction reports must be considered. Findings support the continuation and expansion of visits in other Public Health internships. Key messages • Field visits, used as a pedagogical tool, may enhance medical students’ ability to understand health determinants. • User Field visits provide a new perspective on previously acquired theoretical knowledge regarding prevention.

Cultural Considerations in Intelligence Test Adaptations: a Critical Review of the WAIS-IV India and Its U.K. and U.S. Counterparts.

This study critically examined the adaptation and normative processes of the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV)India. WAIS-IV U.K. Edition Administration and Scoring Manual, WAIS-IVIndia Administration and Scoring Manual, and WAIS-IV Administration and Scoring Manual and Technical and Interpretive Manuals were used to extract test development and adaptation process data. The analyses were descriptive and qualitative. A review of the manuals showed that the normative sample were fluent English speakers with high education levels, which is not representative of India's multilingual and educational diversity. The study identified minimal cultural and linguistic adaptations in subtests, which raised concern about potential biases in the test adaptation process across cultures. The lack of clarity in item selection, demographic data, sample selection, and psychometric properties raised questions about the test's reliability and validity. Issues related to education, age, gender, and regional distribution are discussed, emphasizing the test's limitations in capturing the diverse characteristics of the heterogeneous Indian population. WAIS-IV India may not adequately represent India's population. More careful consideration of socioecological and psychometric factors when adapting intelligence (IQ) tests and use of advanced data harmonization methods can provide robust cross-national statistical harmonization in the future along with more ecologically valid local test development efforts. This can allow consideration of social determinants of health and other factors that can inform a more nuanced approach to IQ test development in non-western contexts.

9.P. Scientific session: National perspective on knowledge translation to bridge the gap between research and policy

Abstract National and international public health institutions play a crucial role in guiding public health policies. However, a key challenge in this process is to bridge the gap between the data collected and the insights required to guide decision-makers, commonly referred to as knowledge translation (KT). Various models have been proposed to facilitate the KT process in public health, and one of the widely recognized models is the Data-Information-Knowledge-Wisdom (DIKW) model, also known as the knowledge pyramid or information hierarchy. In this workshop, we provide insight into the Belgian perspective on the implementation of the DIKW pyramid, using the Belgian health status website as an example https://www.healthybelgium.be. The workshop will tackle the four distinct layers of KT, including data, information, knowledge, and wisdom. The first one involves the data collection process, and the importance of considering potential biases that may arise during this process. Information revolves around the data analysis and the translation of data into meaningful information. Knowledge builds on data-driven information by contextualizing public health analysis results. Finally, wisdom reflects the end-result of the DIKW model, highlighting the need to utilize knowledge for action and achieving results based on the analysis performed. The DIKW model provides a valuable platform by delineating the different layers required to move from raw data to knowledge and wisdom. This process is crucial in achieving an impact on public health, where the data analyzed by public health researchers informs the end-users, including policymakers and the public. Nonetheless, the DIKW model and its applicability to knowledge translation are the subjects of many debates. Within this workshop, the national perspective of the national public health institute in Belgium, Sciensano, on each level of the DIKW pyramid will offer participants concrete examples that might spark an initial step towards their nationwide implementation of the DIKW pyramid for KT. The workshop will offer a step-by-step presentation on the different KT layers, whereby the attendees will gain unique insights into the system implemented in Belgium. This presentation will foster a discussion for generating ideas and stimulating other countries to implement the DIKW pyramid. Key messages • The DIKW pyramid offers a valuable framework to guide knowledge translation in public health, which is a key task for public health institutes. • Attendees will receive an overview and a practical example of how the DIKW-model could be applied in a national context, including a discussion of the strengths and weaknesses of the pyramid.

Are answers affected by data collection on paper or online? The case of Belgian adolescents

Abstract Background In the early 2000s, the prevalence estimations were compared in relation with the mode of data collection, i.e. paper or online. Following the widely spread routine use of connected devices, especially in adolescents, such a monitoring is of interest. Within the 2022 “Health Behaviour in school-aged children” survey in French-speaking primary and secondary schools of Belgium, those participating could choose between paper and online collection. Our objective was to compare a large set of indicators between the two modes after adjusting for the students’ characteristics. Methods After a two-stage random selection of schools and classes, 11,751 10-to-20-year-old adolescents completed the questionnaire either on paper or online. Descriptive analysis and comparisons were conducted using chi2 and multiple logistic regressions adjusted for the students’ sociodemographic characteristics that may vary according to the questionnaire mode. Results Overall, 54.9% of the questionnaires were completed online. The choice of the collection mode provided to the schools’ headteachers was associated with the adolescents’ characteristics. Those who completed the questionnaire online were more likely to be in secondary schools, in Wallonia, and of a more favourable socioeconomic profile. After adjustments, no difference in the indicators regarding health, well-being, substance use, or sleep duration was observed. Conversely, a difference was observed for screen time (internet ≥2h/weekday: 70.4% on paper vs. 72.3% online; adjusted P-value: 0.02), some food groups (daily vegetables: 51.2% vs. 58.0%; P-value<0.001) and body weight status (overweight including obesity: 16.7% vs. 14.0%; P-value<0.001). Conclusions In some circumstances such as data collection in the school setting, mixed modes are still necessary due to operational issues. While most indicators did not show a difference between online and paper collection, potential implications for others need to be further understood. Key messages • Collecting data through online or paper questionnaires does not lead to major differences in prevalence estimates. • With the rise of online surveys, further assessment is required to understand the potential declaration biases.

Prognostic impact of SGLT2-inhibitor therapy on survival in patients with transthyretin amyloid cardiomyopathy

Abstract Background Sodium-glucose co-transporter 2 inhibitors (SGLT2i) significantly improve cardiovascular outcomes in patients with heart failure across the whole spectrum of ejection fraction. Recently published data indicates the safety and tolerability of SGLT2i therapy in patients with transthyretin amyloid cardiomyopathy (ATTR-CM). However, there is a lack of evidence regarding the impact of SGLT2i therapy on clinical outcomes in those patients. Purpose This study aims to investigate the association between SGLT2i therapy and clinical outcomes in patients with ATTR-CM. Methods This is an analysis of a prospective registry conducted at an expert center for hypertrophic cardiomyopathies. We included all participants with a confirmed diagnosis of ATTR-CM, who were enrolled until December 2022. Prior and concomitant SGLT2i treatment was systemically assessed according to the patient interview and medical records. Patients’ outcomes were retrieved from local medical and health insurance records. Results We enrolled 116 patients, of whom 7 patients (6%) were treated with SGLT2i at inclusion and 44 patients (38%) were started on SGLT2i therapy during the observational period while 65 patients (56%) were SGLT2i-naïve. There were no significant differences in baseline characteristics between groups, as shown in the table. Median (interquartile range) age was 80 (76-82) years with 15% females. Median N-terminal pro-brain natriuretic peptide (NT-proBNP) was 2844 (1519-5033) pg/ml, median estimated glomerular filtration rate (eGFR) 58 (46-69) ml/min/1,73m2, and median troponin T 56 (34-83) pg/ml. During a median follow-up of 2.6 (1.8-3.7) years, 28 patients (24%) died, of whom 6 patients (5%) were in the treatment group and 22 patients (19%) in the control group. In univariate Cox-regression analysis, SGLT2i treatment was significantly associated with a lower mortality (HR 0.325, 95%CI 0.132-0.802, p=0.015). This association persisted after adjusting for age and sex (HR 0.347, 95%CI 0.139-0.867, p=0.023). However, when treatment initiation time was considered, the benefit did not remain statistically significant. No significant associations between SGLT2i treatment and worsening heart failure hospitalizations or cardiovascular death were observed. Conclusion In a prospective registry-based cohort of patients with ATTR-CM, SGLT2i therapy was associated with significantly better survival, with a potential treatment initiation time bias. These results suggest that SGLT2i may exert beneficial effects in patients with ATTR-CM and warrant a randomized controlled trial.

Examining the local Universe isotropy with galaxy cluster velocity dispersion scaling relations

In standard cosmology, the Universe is assumed to be statistically homogeneous and isotropic. This assumption suggests that the expansion rate of the Universe, as measured by the Hubble parameter, should be the same in all directions. However, our recent study based on galaxy clusters finds an apparent angular variation of approximately $9<!PCT!>$ in the Hubble constant, $H_0$, across the sky. In the study, the authors utilised galaxy cluster scaling relations between various cosmology-dependent cluster properties and a cosmology-independent property, i.e. the temperature of the intracluster gas ($T$). A position-dependent systematic bias of $T$ measurements can, in principle, result in an overestimation of apparent $H_0$ variations. Therefore, it is crucial to confirm or exclude this possibility. In this work, we search for directional $T$ measurement biases by examining the relationship between the member galaxy velocity dispersion and gas temperature ($ v -T$) of galaxy clusters. Both measurements are independent of any cosmological assumptions and do not suffer from the same potential systematic biases. Additionally, we search for apparent $H_0$ angular variations independently of $T$ by analysing the relations between the X-ray luminosity and Sunyaev-Zeldovich signal with the velocity dispersion, $L_ X v $ and $Y_ SZ v To study the angular variation of scaling relation parameters, we determined the latter for different sky patches across the extra-galactic sky. We constrained the possible directional $T$ bias using the $ v -T$ relation, as well as the apparent $H_0$ variations using the $L_ X v $ and $Y_ SZ v $ relations. We utilised Monte Carlo simulations of isotropic cluster samples to quantify the statistical significance of any observed anisotropies. We calculated and rigorously took into account a correlation of $L_ X $ and $Y_ SZ $ residuals. No significant directional $T$ measurement biases are found from the $ v -T$ anisotropy study. The probability that the previously observed $H_0$ anisotropy is caused by a directional $T$ bias is only $0.002<!PCT!>$. On the other hand, from the joint analysis of the $L_ X v $ and $Y_ SZ v $ relations, the maximum variation of $H_0$ is found in the direction of $(295^ with a statistical significance of $3.64 fully consistent with our previous results. Our findings, based on the analysis of new scaling relations utilising a completely independent cluster property, $ v $, strongly corroborate the previously detected anisotropy of galaxy cluster scaling relations. The underlying cause, for example, $H_0$ angular variation or large-scale bulk flows of matter, remains to be identified.

Cost impact of the use of a universal cardiac implantable electronic devices remote monitoring solution: results of the Evidence RM study

Abstract Background Remote monitoring (RM) is considered the standard of care for patients with cardiac implantable electronic devices (CIED). In 2023 the HRS/EHRA/APHPRS/LAHRS expert consensus highlighted the potential interest of alert-based monitoring and the use of a third-party platform for RM management. By lightening the RM workload for clinical staff, valuable time and resources can be redirected towards patient care. Purpose This study aims to assess the impact on healthcare expenditures of the adoption of a Universal, vendor-neutral, and alert-focused remote monitoring (RM) platform for CIED in France, as opposed to the Conventional RM conducted via device specific manufacturers' platforms. Methods This study utilizes the French National Health Database (SNDS) to evaluate the effectiveness of RM solutions among patients with implantable cardioverter defibrillators (ICD), including cardiac resynchronization therapy defibrillators (CRT-D). All patients under RM were categorized based on the type of monitoring, either the Universal RM or the Conventional RM. The analysis was conducted on year 2019 and included only patients maintaining consistent RM solutions and device types throughout the entirety of the study period. To mitigate potential biases, costs were adjusted according to age, gender, device type, year of first implantation, year of RM initiation, medical center experience with RM, and Elixhauser score for comorbidities. Results Study cohort consisted of 36,401 patients (age 67.3 ± 13.0 years / male 78.4% / CRT-D 40.1%), 1,482 patients followed using the Universal RM platform and 34,419 patients monitored with the Conventional RM solutions. The study findings revealed a 4% decrease in corrected total costs and a notable 17.8% reduction in hospital costs among patients utilizing the Universal RM. Analysis further identified that this decrease in hospital expenses was primarily influenced by a reduction of the costs associated with cardiovascular diseases. Conversely, the group utilizing the Universal RM experienced a 7.9% increase in total outpatient costs compared to Conventional RM, while ambulatory visit costs remained unchanged. As costs incurred by patients were not included, total costs may however be underestimated. Patients under the Universal RM solution may benefit of more proactive preventive measures delivered through outpatient care. By addressing issues preemptively, critical conditions may be averted, enhancing overall patient management, and diminishing hospital costs. The adoption of a third-party Universal platform may thus yield to cost savings, exemplified by a negative Incremental Cost-Effectiveness Ratio (ICER) of -103€ per Day Alive and Out of Hospital (DAOH) as observed in this study. Conclusions The use of a third-party Universal RM platform showed a positive impact in terms of costs reduction for the French healthcare system on this ICD population.

Contrast-enhanced ultrasound combined with percutaneous ultrasound-guide core needle biopsies in the diagnosis of pancreatic lesions: a propensity score-matched study

ObjectiveTo compare the diagnostic performance of pancreatic lesions using percutaneous ultrasound (US)-guided core needle biopsy (CNB) with and without contrast-enhanced ultrasound (CEUS).MethodThe patients were divided into two groups, US and CEUS group, based on whether CEUS was performed prior to biopsy. Before and after propensity score matching (PSM), the CNB-relevant characteristics of the two groups, including the first puncture success rate, the number of sampling, complication rate, type of complications, and degree of abdominal pain, were compared. The accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) of percutaneous US-guided CNB were compared between the groups.ResultsThis study included 277 patients (173 men and 104 women) with pancreatic lesions who underwent percutaneous CNB before PSM; 190 patients in the CEUS group, and 87 in the US group prior to CNB. After controlling for potential biases using PSM, no significant differences were observed in the complication rate, type of complications, or degree of abdominal pain between the CEUS and US groups (P > 0.05). However, significant differences were observed in the first puncture success rate and the number of sampling (P < 0.05). Importantly, before and after PSM, the CEUS group achieved a higher first-puncture success rate while obtaining a lower number of sampling (P < 0.05). Compared to the US group, the CEUS group demonstrated improved accuracy, sensitivity, specificity, PPV, and NPV of 13.1%, 14.9%, 13.4%, 2.5%, and 38.7%, respectively. Furthermore, the significant difference was observed in the AUC for diagnostic performance between the two groups when compared using DeLong’s test (P = 0.043).ConclusionsPerforming CEUS before percutaneous CNB for pancreatic lesions can help achieve better biopsy results, reduce the number of punctures samples, increase the success rate of biopsies, and avoid the need for repeat biopsies.

Bias analysis on socioeconomic status and lung cancer in a pooled international case-control study

Abstract Background Low socioeconomic status (SES) groups showed increased lung cancer risks even after adjustment for smoking habits and other exposures to lung carcinogens. Although several biases were often indicated and discussed, only few studies quantified the impact of potential biases. Methods We conducted a bias analysis on the association of lung cancer and SES using data from the SYNERGY project, including 12 case-control studies with 18 study centres from Europe and Canada (16,550 cases, 20,147 controls). SES in quartiles was derived from the International Socio-Economic Index of occupational status (ISEI). Odds ratios (OR) with 95% confidence intervals (CI) were estimated by logistic regression adjusting for age, study centre, and smoking. In addition, we estimated natural direct SES effects and natural indirect smoking effects by inverse odds ratio weighting. In a multiple quantitative bias analysis, we considered impacts of misclassification of smoking status, selection bias, and unmeasured mediator-outcome confounding by a protective genetic factor, and created 95% simulation intervals (SI) by bootstrap. All analyses were stratified by sex. Results Adjustment for smoking as well as natural effects estimation showed that nearly half of lung cancer risks of lower SES groups in men and up to one third in women were attributable to smoking. Consideration of all types of bias reduced lung-cancer risks in the fully adjusted logistic regression models, with the strongest impact by selection bias: For the 4th versus 1st (highest) ISEI quartile OR decreased from 1.83 (1.69-1.98 CI) to 1.50 (1.30-1.73 SI) in men, and OR 1.48 (1.27-1.72 CI) to 1.20 (0.96-1.53 SI) in women. Conclusions Smoking is the main target for prevention of lung cancer, along with occupational and environmental exposures, in particular in lower SES groups. This finding remains, though our analysis revealed reduced lung-cancer risks of lower SES groups after multiple bias adjustment. Key messages • Impact of potential biases was quantified in the association of SES and lung cancer. • Lung cancer risks were partially attributable to smoking and multiple biases.

Survival penalty in patients undergoing transcatheter versus surgical aortic valve replacement: is it just a question of frailty? Insights from a large cohort with demographic matching

Abstract Background Transcatheter aortic valve replacement (TAVR) and surgical aortic valve replacement (SAVR) are the reference treatments for aortic stenosis. Conflicting data exist on late survival results of these two techniques. While randomised controlled trials (with highly selected patients) assert the non-inferiority of TAVR, real-world data from large registries (with potential uncorrected confounders) suggest the opposite. Comparing the survival rates of SAVR and TAVR with those of a matched reference population could help circumvent potential biases of direct procedural comparisons and quantify the impact of frailty on post-operative survival. Purpose Our study aims to compare the 5-year overall survival rates of SAVR, non-frail TAVR, and frail TAVR patients with the survival of their demographically matched reference population. Methods This study included all consecutive patients undergoing either SAVR or TAVR at a tertiary hospital between January 2010 and December 2021. Each patient underwent a comprehensive geriatric assessment by an experienced geriatric team, from which a Clinical Frailty Scale (CFS) was derived using a validated classification tree. Frailty was defined as a CFS ≥ 6, in accordance with ESC guidelines. The patients were stratified into 3 subgroups: SAVR, non-frail TAVR, and frail TAVR patients. For each subgroup, a corresponding reference population was generated using national actuarial tables to ensure individual-level matching based on age, gender, and year of observation. Survival of each subgroup was then compared to its corresponding reference population over a 5-year period using standardized mortality ratios (SMR) and one-sample logrank tests. Results Our cohort included 1183 SAVR, 254 non-frail TAVR, and 151 frail TAVR patients, with mean ages of 74.1, 85.4, and 85.7 years, and median EuroSCORE II of 2.0%, 4.2%, and 5.4%, respectively. Survival of SAVR patients was equivalent to the reference population (80.1 vs 80.9% survival; SMR = 1.06 [0.89 – 1.25]; p = 0.454; Figure 1A). Similarly, non-frail TAVR patients also achieved survival rates comparable to those of the reference population (50.9 vs 54.0% survival; SMR = 0.94 [0.76 – 1.18]; p = 0.549; Figure 1B). However, frail TAVR patients faced a significantly reduced survival rate (41.7% vs 54.0% survival; SMR = 1.48 [1.15 – 1.90]; p &amp;lt; 0.001; Figure 1C). The frailty of this subgroup could therefore explain a 48% increase in mortality risk compared with their reference population after 5 years. Conclusions Treating non-frail patients with aortic stenosis through TAVR or SAVR restores life expectancy to levels comparable with their reference population, challenging results from other real-world registries. In contrast, TAVR in frail patients is associated with worse survival, raising the question of treatment futility in this specific subgroup. Refining the selection of frail patients who may benefit from TAVR remains an important clinical challenge.

Do You Trust Broken Mirrors? Confronting Structural Inequalities in Student Evaluations of Teaching in Higher Education

ABSTRACT In recent years, the presence of foreign-born female academics (FBFA) in U.S. higher education has noticeably increased. Despite this progress in inclusion and diversity, they continue to face significant challenges related to race and gender biases in education. These challenges are evident when student evaluations of teaching (SETs) are the mechanism that evaluates the effectiveness of their teaching. Adopting a duo-ethnography methodology, researchers examined the effects of SETs in their professional lives, exploring the intersections of their experiences in their academic roles. Results highlight the potential biases in SETs, emphasizing the need to question the trustworthiness of these measures.

Efficiency or Equity? How Public Values Shape Bureaucrats' Willingness to Use Artificial Intelligence to Reduce Administrative Burdens

Generative and predictive Artificial Intelligence (AI) has the potential to revolutionize many features of policy implementation, but only if bureaucrats understand and decide to utilize AI. While existing research has examined the public's demand for AI in policy implementation, very little scholarship has examined the views of bureaucrats. In this article, we implement two survey experiments—one on predictive AI and another on generative AI—among government employees to reveal how information about the effects of AI on public values (equity and efficiency) affect bureaucrats' decisions on whether to utilize AI to ease administrative burdens on clients. We find that, regardless of AI type (i.e., predictive vs. generative) and randomly assigned policy contexts with different socially constructed target populations, bureaucrats are less likely to adopt AI when they are presented with information about the potential biases embedded in algorithms and the negative impacts on social equity. On the other hand, we find that highlighting the efficiency gains of utilizing AI has no impact on support for using AI. These findings suggest that concerns over equity may be a key barrier to bureaucrats' decisions on adopting AI in policy implementation.

Evaluating mammography screening in observational cohort designs: the importance of avoiding lead time bias.

To investigate the potential lead time bias of the evaluation model (extended follow-up for women diagnosed with breast cancer) used to evaluate mammography screening in a recent Danish study. This model was compared with two traditional models. We retrieved data on women diagnosed with breast cancer in each county of Norway from 1986 to 2016. In a population-based open cohort study, the change in incidence-based mortality (IBM) was estimated by relative rate ratios comparing a screening period with a historical period for each of three age groups: women eligible for screening and younger and older ineligible women. We applied the evaluation model, and for comparison two traditional IBM models from a recent Norwegian study: one without extended follow-up and no possibility of lead time bias and one with extended follow-up irrespective of diagnosis, possibly diluting any screening effect. The evaluation model estimated an extra 11% reduction in breast cancer mortality among the screening eligible relative to ineligible women. However, this result could largely be ascribed to lead time bias inflated by overdiagnosis and a decreasing mortality from other causes among eligible women. A reduction in breast cancer mortality was observed for both eligible and younger and older ineligible women across models, and relative rate ratios close to 1 were obtained using the two traditional IBM models, indicating no effect of screening on breast cancer mortality. Two models without lead time bias found no reduction in breast cancer mortality, whereas the evaluation model estimated a reduction attributable to lead time bias.

Machine learning based prediction models for cardiovascular disease risk using electronic health records data: systematic review and meta-analysis

Abstract Background Cardiovascular disease (CVD) remains a major cause of mortality in the UK, prompting the need for improved risk predictive models for primary prevention. Machine learning (ML) models utilizing electronic health records (EHR) offer potential enhancements over traditional risk scores like QRISK3 and ASCVD. Objective To systematically evaluate and compare the efficacy of ML models against conventional CVD risk prediction algorithms using EHR data for medium to long-term (5-10 years) CVD risk prediction. Methods A systematic review and random-effect meta-analysis were conducted according to PRISMA guidelines, assessing studies from 2010 to 2024. We retrieved 32 ML models and 26 conventional statistical models from 20 selected studies, focusing on performance metrics such as Area Under the Curve (AUC) and heterogeneity across models. Results ML models, particularly random forest and deep learning, demonstrated superior performance, with the highest recorded pooled AUCs of 0.865 (95% CI: 0.812-0.917) and 0.847 (95% CI: 0.766-0.927), respectively. These significantly outperformed the conventional risk score of 0.765 (95% CI: 0.734-0.796). However, significant heterogeneity (I² &amp;gt; 99%) and potential publication bias were noted across the studies. Conclusion While ML models show enhanced calibration for CVD risk, substantial variability and methodological concerns limit their current clinical applicability. Future research should address these issues by enhancing methodological transparency and standardization to improve the reliability and utility of these models in clinical settings. Implications This study highlights the advanced capabilities of ML models in CVD risk prediction and emphasizes the need for rigorous validation to facilitate their integration into clinical practice. This review was registered with PROSPERO (CRD42021250058).

The Oncology QCARD Initiative: Fostering efficient evaluation of initial real-world data proposals.

The oncology quality, characterization, and assessment of real-world data (Oncology QCARD) Initiative was formed to develop a set of minimum study design and data elements needed to evaluate the fitness of the real-world data (RWD) source(s) proposed in an initial study concept as part of early interaction with scientific reviewers. A multidisciplinary executive committee (EC) was established to guide the Oncology QCARD Initiative. The EC conducted a landscape review of published literature, guidances, and guidelines to evaluate relevant dimensions of data quality measurement. Guided by the review and informed by expert feedback, the Oncology QCARD Initial Protocol Characterization (IPC) provides a summary of minimum elements needed to adequately describe an initial clinical study concept that involves RWD and is intended to support decision-making. Fit-for-use data and fit-for-purpose design emerged as themes from the landscape analysis. Data that are fit-for-use are both relevant (sufficiently capturing exposure, outcomes, and covariates) and reliable (understanding data accrual and quality control and whether the data represent the underlying concepts they are intended to represent) to answer a specific research question. A fit-for-purpose design takes appropriate steps to ensure internal and external validity and allows for transparency in reporting. The QCARD-IPC focuses on high-level characteristics of RWD sources and study design domains including data temporality, population, medical product exposure, comparators, and covariates, endpoints, statistical analysis, and data quality assurance plans. Evaluation of studies including RWD requires understanding the data source, study design, and potential biases to preliminarily evaluate whether selected RWD are fit-for-use for the research question. The Oncology QCARD-IPC provides a structured, transparent approach to facilitate early review and enhanced communication between study sponsors and scientific reviewers of initial study proposals including RWD.

Implications of CMIP6 Models‐Based Climate Biases and Runoff Sensitivity on Runoff Projection Uncertainties Over Central India

ABSTRACTAccurate runoff projections are vital for developing climate adaptation strategies, yet significant uncertainties persist. The commonly employed approaches to constrain these uncertainties rely on the stationarity of climate biases and runoff sensitivity, which may not hold for climate‐sensitive regions (e.g., semi‐arid regions). This study investigates the validity of the stationarity assumption across 29 CMIP6 models, encompassing diverse climate biases (Dry Warm, Wet Warm, Dry Cold, and Wet Cold), utilising a semi‐arid region in central India as a testbed. The implications of this assumption on runoff projection uncertainties were comprehensively assessed across the runoff modelling chain for three time periods (the 2030s, 2060s and 2090s) based on the Soil and Water Assessment Tool (SWAT) simulations. The results highlight the non‐stationary nature of climate biases and runoff sensitivity under future scenarios, challenging the widespread applicability of common uncertainty‐constraining approaches. Moreover, the impact of non‐stationarity on runoff projection uncertainty was found to be strongly influenced by the choice of GCMs, preprocessing methods and climate change scenarios. In the 2030s, GCMs dominate runoff uncertainty, with dry models exhibiting ~10%–15% higher uncertainty compared to warm models, which is further amplified when interacting with warm biases. However, from the mid‐century onwards, the bias‐adjustment approaches and climate change scenarios significantly shape runoff projection uncertainties under non‐stationary conditions. These findings emphasise the potential of climate bias and runoff sensitivity‐based GCM selection for reducing runoff uncertainty in near‐future assessment (2030s). For mid‐term and long‐term runoff projections, addressing diverse climate biases through bias‐adjustment approaches is more viable. This study offers critical insights to prioritise the development of a non‐stationarity‐based approach for reliable runoff projections in climate‐sensitive regions.

Diversity, environments, and biogeography of Homoloida (Decapoda: Brachyura) from the Late Jurassic to the present

ABSTRACT The section Homoloida with the sole superfamily Homoloidea comprises both extinct and extant brachyuran crab taxa, with records ranging from the Tithonian (Jurassic) to the present. Examination of fossil and modern occurrence records for this clade provides insight into the biogeography, biodiversity, and environmental preference of the homoloid crabs over time. The fossil record of Homoloidea, comprised of the families Tithonohomolidae, Mithracitidae, Homolidae, Latreilliidae, and Poupiniidae, is global in distribution and peaks in diversity during the uppermost Cretaceous with the most recent fossils known from the Miocene. Modern homoloids are absent from only a handful of regions globally, with particularly high diversity and abundance in the tropical Indo-West Pacific. Comparison of interpreted depositional environments for fossil occurrences and depth records for modern occurrences sheds light on the environmental preferences and potential spatial sampling biases of this clade. By assessing both fossil and modern records of Homoloidea, patterns (e.g., absence from cold-water latitudes, Oligo-Miocene disappearance from the northeastern Pacific) emerge that may have important implications for future study of this clade, particularly in relation to dispersal/turnover events and environmental constraints.