The findings come amid documented inequities in routine vaccination among Native Hawaiian and Pacific Islander communities in the United States, a group that also faces a higher risk of death from HPV-related cervical cancer than other populations. The researchers said broad category labels can obscure cultural, social and linguistic differences within those communities, including the experiences of Tongans.

The researchers conducted three focus group discussions in February and March 2024 with Tongan adults aged 18 and older who lived in Utah. Groups were conducted in English and Tongan. The study used the Socioecological Model and Social Cognitive Theory to examine behavioral, social and cultural influences on general vaccine attitudes and HPV literacy, experience and access.

Participants showed a generational pattern in where they preferred to obtain vaccine information. Those younger than 55 relied mainly on social media, whereas participants aged 56 to 65 and older expressed greater confidence in healthcare providers. The older group emphasized the importance of trust between patients and physicians, identifying that relationship as central to confidence in vaccination decisions.

Within families and households, participants described conformity with traditional gender expectations as a major influence. Tongan women were viewed as the primary caregivers and health decision-makers, with much of the responsibility for household health decisions, including vaccination, assigned to them. The study therefore identified gender and family roles as important parts of the environment in which vaccine acceptance and refusal were considered.

Vaccine hesitancy within close social circles was strongly influenced by misinformation, according to the participants. Their accounts also highlighted the strong reliance on traditional medicine and religious leaders in Tongan communities. At the same time, participants said discussions of sexual health and sexually transmitted diseases are constrained by social stigma. The findings indicate that misinformation and culturally influential sources operate alongside limited discussion of subjects directly relevant to HPV vaccination.

At the societal level, participants identified several structural barriers: inadequate insurance coverage for preventive and therapeutic services, limited access to health information in culturally and linguistically appropriate forms, and under-representation of Tongans among healthcare providers and staff. These gaps were described not only as problems of individual knowledge but also as conditions affecting whether useful information and services could be reached and trusted.

As a qualitative pilot study, the research was designed to explore experiences and perceptions rather than estimate vaccination rates or test an intervention. Its sample was small and geographically limited to Tongan adults in Utah, and the findings arose from three facilitated discussions conducted in two languages. The authors said the results underscore the need for culturally specific and culturally humble vaccine messaging and strategies at multiple social-ecological levels, particularly to address misinformation and barriers affecting HPV vaccine awareness and acceptance.

Subject of Research: Tongan adults’ vaccine attitudes, HPV awareness and access barriers in Utah

Article Title: Navigating vaccine beliefs: a qualitative pilot study of general vaccine attitudes and HPV awareness among Tongan adults

Article References: Obray, A. M., Skyles, T. J., Poole, B. D., Bashford, T. M., Hoelzer, H., Davis, S. C., Petersen, L. G., Fox, D., Sloan, D. S., Vergara, M. A. C., & Novilla, M. L. B. (2026). Navigating vaccine beliefs: a qualitative pilot study of general vaccine attitudes and HPV awareness among Tongan adults. BMC Public Health. https://doi.org/10.1186/s12889-026-29627-9

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29627-9

Keywords: Tongan adults, HPV vaccine, vaccine hesitancy, cervical cancer, health disparities, Utah, cultural competence, health misinformation, focus groups, public health, Navigating, vaccine

Researchers conducted semi-structured, in-depth interviews with 26 ophthalmic nurses selected through maximum-variation sampling. Participants worked in tertiary, secondary and primary eye-care settings across eastern, central and western China, providing coverage of different levels and locations of care. The interviews used the Critical Incident Technique, while the transcripts were examined with Braun and Clarke’s reflexive thematic analysis. Reporting followed the Consolidated Criteria for Reporting Qualitative Research, or COREQ.

Three themes emerged from the analysis. The first, described by the researchers as a “tentative approach,” captured a progression from curiosity to a conditional appraisal of generative AI’s competence. Nurses did not treat the technology as uniformly reliable or unreliable; instead, their trust was calibrated in a graded, task-specific manner. The researchers tentatively called this pattern “gradient calibration,” indicating that acceptance varied with the task and context rather than representing a single position toward generative AI.

The second theme, “guarding professional boundaries,” concerned the anchoring of trust in ethical obligation and the nurse–patient relationship. This focus is particularly relevant to ophthalmic nursing, where visual outcomes can be irreversible, ophthalmic medication requires precision, and visually impaired patients may depend on nurse-mediated support. The findings indicate that the possible reconfiguration of professional roles and relationships by generative AI forms part of nurses’ trust negotiations. They also show that trust is not only a judgment about technical performance: ethical tension and responsibility toward patients influence how nurses engage with the tools.

The third theme placed trust negotiation within the wider organizational ecology of ophthalmic care. Institutional conditions, professional expectations and the workplace context shaped how nurses approached and used generative AI. Taken together, the themes show trust emerging through negotiation rather than as a fixed attitude of acceptance or refusal. The study does not measure whether generative AI improves diagnostic accuracy, patient outcomes or nursing performance, nor does it establish how common particular responses are among ophthalmic nurses. Because the evidence comes from interviews with 26 participants, its conclusions describe the experiences and reasoning of this sample. Transferability beyond the sampled Chinese eye-care settings requires further study.

Subject of Research: Trust negotiation in generative AI use among ophthalmic nurses in China

Article Title: Navigating the unknown: a qualitative descriptive study of ophthalmic nurses’ trust negotiation with generative artificial intelligence in China

Article References: Zhu, M., Zhang, X., Zhu, Y., Zhang, L., Li, R., Li, A., & Qu, L. (2026). Navigating the unknown: a qualitative descriptive study of ophthalmic nurses’ trust negotiation with generative artificial intelligence in China. BMC Nursing. https://doi.org/10.1186/s12912-026-05437-7

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05437-7

Keywords: ophthalmic nursing, generative artificial intelligence, trust calibration, qualitative interviews, nurse–patient relationship, clinical ethics, organizational context, eye care, thematic analysis, Navigating, unknown, qualitative

It also addresses a practical problem for immunologists: designing reliable experiments and making sense of the resulting data can be technically challenging and conceptually opaque. The authors present the review as a guide for researchers beginning to incorporate metabolomics into immunology. It outlines key methodological approaches and considers common challenges alongside potential solutions, with the aim of lowering the barrier to entry for laboratories that may otherwise regard the field as inaccessible.

The guidance is intended to help immunologists design initial experiments, interpret metabolomic data produced by others and confidently incorporate metabolic measurements into their research. A central part of the article is a step-by-step roadmap for interpreting nutrient-tracing data. The review explains how readers can work through complex isotopologue datasets and draw meaningful conclusions from them.

The review’s broader message is that the expanding recognition of metabolism in immune biology is creating demand for analytical tools capable of capturing metabolic states across different conditions. Metabolomics offers one such tool because it profiles the metabolites that reflect cells’ energy use and biosynthetic activity.

Subject of Research: Practical metabolomics methods for studying immune-cell metabolism and interpreting nutrient-tracing data.

Article Title: Navigating metabolomics in the era of immunometabolism

Article References: Slattery, K., Avizonis, D. Z., O’Connor, G. M., & Gardiner, C. M. (2026). Navigating metabolomics in the era of immunometabolism. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08998-y

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08998-y

Keywords: metabolomics, immunometabolism, immune cells, metabolic pathways, bioenergetics, nutrient tracing, isotopologues, immunology, translational medicine, Navigating, scientific research

The review, led by Alixe Ménard and colleagues at the University of Ottawa’s Interdisciplinary School of Health Sciences and LIFE Research Institute, with additional contribution from Kanika Sarwal at the University of Waterloo, set out to answer a deceptively simple question: what actually happened to stroke care when COVID-19 swept through hospitals worldwide? The answer, drawn from 107 studies that met rigorous inclusion criteria out of 1,405 initially screened, reveals a story of systemic disruption that touched every link in the stroke care chain—from the moment a patient notices symptoms to long-term rehabilitation and quality of life.

Using the Joanna Briggs Institute Methodology for Scoping Reviews, the team searched four major electronic databases—APA PsycInfo, Embase, Medline, and CINAHL—for English-language primary research articles published between December 2019 and September 2024. They focused specifically on individuals who experienced a stroke during the pandemic, extracting study characteristics and significant findings to enable a qualitative and narrative synthesis. The review followed PRISMA-ScR reporting guidelines, the preferred reporting standard for scoping reviews, ensuring transparency in how studies were identified and included.

From the 107 included articles, seven sub-themes emerged, painting a picture of stroke care under siege. The first and perhaps most striking theme concerns patient behavior and awareness. Fear of contracting COVID-19 in hospitals became a powerful deterrent, causing many people to ignore or downplay stroke symptoms rather than seek emergency care. This reluctance to present to emergency departments was not a marginal effect—it was a fundamental shift in how patients weighed the risk of staying home against the risk of going to hospital during a respiratory pandemic.

The second theme involves the rapid and often disruptive pivot to telemedicine and remote care. In-person consultations dropped sharply as healthcare systems tried to limit viral transmission, and while telemedicine offered a lifeline for continuity of care, it also introduced new barriers. Stroke assessment relies heavily on physical examination—neurological deficits, facial asymmetry, motor weakness, speech patterns—and remote evaluation cannot fully replicate that hands-on diagnostic precision. For a condition where minutes determine outcomes, the shift to virtual care was a double-edged sword.

Delays in treatment formed the third theme, with the review documenting reduced utilization of critical interventions. Time-sensitive treatments for stroke—particularly thrombolysis with tissue plasminogen activator and mechanical thrombectomy for emergent large vessel occlusion, or ELVO—depend on rapid presentation, rapid imaging such as CT, and rapid intervention. When patients delayed seeking help, or when hospital processes were reconfigured around infection control, the window for these interventions narrowed. The review found that this translated into measurable declines in the use of these treatments, meaning fewer patients received the therapies that could reverse their deficits.

The fourth theme addresses the strain on healthcare resources, which led to fragmented care. Hospitals repurposed beds, redeployed staff, and reorganized workflows around COVID-19 surges. Stroke units—specialized wards with dedicated nursing, neuroimaging access, and rehabilitation expertise—were among the resources diverted. When a stroke unit becomes a COVID ward, or when stroke neurologists are pulled to pandemic duties, the coordinated pathway from emergency department to imaging to intervention to rehabilitation breaks apart. The review describes this fragmentation as a direct consequence of resource strain, not a failure of individual clinicians.

Quality of care metrics declined as a fifth theme, with documented drops in care benchmarks and patient outcomes. Stroke care is measured through specific quality indicators—door-to-needle time for thrombolysis, proportion of patients receiving reperfusion therapy, rates of appropriate imaging, and functional outcomes at discharge. The review found that these metrics deteriorated during the pandemic, reflecting the cumulative effect of delayed presentation, reduced interventions, and fragmented systems. Greater stroke severity at presentation formed the sixth theme: when patients finally did arrive at hospital, they were often sicker, having waited longer at home while their symptoms progressed. This created a vicious cycle—fewer admissions but more severe strokes among those admitted.

The seventh and final theme is perhaps the most alarming: a reduction in stroke admissions overall, driven by both patient reluctance and changes in healthcare practices. Stroke did not become less common during the pandemic. People continued to have ischemic strokes, hemorrhagic strokes, and transient ischemic attacks. But fewer of them reached hospital. This suggests a hidden burden of untreated stroke in communities—people who survived at home with deficits that could have been minimized, or who died without ever receiving a diagnosis. The review underscores that this reduction was not a reflection of reduced disease incidence but of reduced access and willingness to seek care.

The implications of this review extend far beyond historical documentation. The authors conclude that the pandemic significantly impacted stroke care globally, leading to increased patient reluctance to seek medical attention, delays in treatment, and a rise in stroke severity. The shift to telemedicine and the strain on healthcare resources further disrupted stroke care, resulting in declines in quality metrics and reduced stroke admissions. But the review is not simply a catalogue of failures—it is a call to action. The authors emphasize the need for adaptable stroke care systems, improved telemedicine access, and enhanced patient education to address pandemic-related challenges and inform future emergency preparedness.

For clinicians and policymakers, the message is clear: stroke care cannot be paused during a crisis. The systems that deliver thrombolysis and thrombectomy must be designed with surge capacity in mind, with contingency plans that preserve time-critical pathways even when infection control demands reorganization. Telemedicine must be developed not as a stopgap but as a robust complement to in-person care, with protocols that account for the limitations of remote neurological assessment. And public health messaging must evolve to address the dual fear that pandemic conditions create—the fear of the virus and the fear of the hospital—so that patients understand that stroke symptoms warrant immediate emergency attention regardless of circulating pathogens.

As the world moves past the acute phase of COVID-19, the lessons embedded in this review serve as a blueprint for resilience. Stroke will remain a leading cause of death and disability worldwide, and the next pandemic or large-scale emergency will test healthcare systems again. The evidence gathered by Ménard and colleagues shows that when systems are not prepared, the consequences ripple through every stage of stroke care—from prevention and awareness through acute intervention to rehabilitation and quality of life. Building adaptable, resilient stroke systems is not optional; it is the difference between recovery and permanent disability for millions of people who will experience stroke in the years to come.

Subject of Research: Stroke care during the COVID-19 pandemic: a scoping review of healthcare disruption, telemedicine, and emergency preparedness

Article Title: Navigating stroke care during the COVID-19 pandemic: a scoping review

Article References: Ménard, A., McLean, M., Oostlander, S., Sarwal, K., Manji, I., O’sullivan, T., & Fraser, S. (2026). Navigating stroke care during the COVID-19 pandemic: a scoping review. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15676-8

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15676-8

Keywords: Navigating, stroke, care, during, COVID-19, pandemic, scoping, review, scientific research

The study, published in the journal Metabarcoding and Metagenomics as part of the Horizon Europe project eDNAqua-Plan, took an unusually comprehensive look at how eDNA data are generated, archived and shared across the continent. Rather than relying on a single method, the team combined four complementary approaches: a qualitative evaluation of the major databases that store eDNA records, a quantitative audit of metadata completeness, a questionnaire survey of 69 researchers from 29 countries, and a large-scale analysis of more than 1,600 scientific publications using a fine-tuned large language model. The verdict is stark. While some repositories demonstrate strong curation and accessibility, others lack essential metadata descriptors, and the gaps are heavily concentrated in exactly the areas that determine whether data can ever be reused.

The technical core of the problem lies in metadata, the structured information that describes how each sample was collected, processed and sequenced. Without knowing the filter pore size used to capture DNA particles, the volume of water extracted, the primers deployed during PCR amplification, or the bioinformatic pipeline applied to the raw reads, a downstream user cannot meaningfully compare one dataset with another. The international standards designed to solve this exist. The Genomic Standards Consortium’s MIxS framework defines mandatory and recommended metadata terms for sequence submissions, including a Water extension with fields such as size_frac_low and samp_vol_we_dna_ext. The Darwin Core standard underpins biodiversity platforms like GBIF and OBIS. Yet the audit of 5,280 European aquatic eDNA sample accessions in the European Nucleotide Archive revealed that these standards are being applied only patchily.

The numbers are remarkable. Although every retrieved sample carried the mandatory environmental context fields, Environment Ontology terms appeared in only about 18 percent of entries. The sample collection method was provided in fewer than 20 percent of records, and the size fraction used for pre-filtration in just 13.5 percent. The volume or mass of sample used for DNA extraction was reported in a mere 1.5 percent of cases, and the nucleic acid extraction method in 2 percent. PCR primers, a critical determinant of taxonomic coverage and amplification bias, were given for 23 percent of records, while the target gene fragment appeared in only 2 percent. Even the MIxS term nucl_acid_ext, designed specifically to hold extraction protocol information, did not appear in a single record. In practice, much of the metadata that does exist was submitted as inconsistent free text rather than standardised values, forcing anyone hoping to harmonise the data into laborious manual curation.

The questionnaire survey exposed a second, more troubling layer to the problem: a significant gap between what researchers believe they are doing and what the repositories actually record. Respondents reported higher compliance with metadata submission practices than the automated audit revealed. Sampling methods were reported by 59 percent of survey participants and DNA extraction protocols by 51 percent, yet the actual BioSample records showed far lower volumes of reported metadata. The authors attribute this mismatch to a combination of survey biases, including self-selection of more engaged respondents and social desirability bias, alongside structural factors such as misperceptions of the full scope of metadata requirements, lack of training, the complexity of submission systems, time constraints and limited incentives or enforcement.

The large language model analysis added a third dimension by revealing how these weaknesses recur across the literature itself. The team assembled 117 manually curated publications as a training set, then used a fine-tuned Llama 2 model to automatically extract methodological information from an expanded corpus of 1,607 eDNA papers, achieving validation accuracy of 97 to 99 percent. The synthesis showed a strong concentration of studies around a limited set of genetic markers, with the mitochondrial COI marker most frequently reported at 19 percent, followed by 16S rRNA at 17 percent and ITS at 14.5 percent. GenBank dominated as the reference database for taxonomic assignment, cited in 19.7 percent of papers where the model could infer one, with SILVA at 18.3 percent and BOLD at 9.9 percent. But the most striking finding concerned protocol transparency: of 1,440 publications for which methodological information could be inferred, only six, or 0.4 percent, cited experimental protocols deposited in protocols.io, and none of those protocol DOIs appeared in the formal reference sections of the papers.

Geographic disparities compound the technical gaps. Analysis of European aquatic eDNA sequence records in the International Nucleotide Sequence Database Collaboration, which comprises ENA, NCBI and the DNA Data Bank of Japan, revealed that records are heavily concentrated in Northern and Western European countries, while Eastern and Southeastern Europe contribute substantially fewer. The imbalance reflects not just differences in sequencing volume but disparities in project-level participation, suggesting that in some regions DNA research activity exists but projects are smaller in scale or involve lower sequencing intensity. Meanwhile, the mobilisation of processed eDNA-derived biodiversity data into secondary platforms remains limited: only 25 datasets in OBIS contained the DNA-derived data extension from European marine waters, and just 15 percent of questionnaire respondents reported submitting such data to biodiversity portals.

The way forward, the authors argue, lies in a structured framework combining three complementary sets of principles. The FAIR principles, ensuring data are Findable, Accessible, Interoperable and Reusable, form the technical foundation, drawing on Darwin Core and MIxS standards alongside GBIF guidelines for publishing DNA-derived data. The CARE principles, emphasising Collective benefit, Authority to control, Responsibility and Ethics, address Indigenous data governance and responsible sample stewardship. The TRUST principles, covering Transparency, Responsibility, User focus, Sustainability and Technology, provide criteria for repository reliability and long-term digital preservation. Recent work by Takahashi and colleagues on Making eDNA FAIR has already produced a comprehensive metadata checklist integrating these standards with new terms tailored specifically to eDNA workflows, and core terms of MIxS and Darwin Core have been mapped to improve interoperability between the sequence and biodiversity data infrastructures.

Importantly, the authors stress that standardisation does not imply a single universal eDNA protocol. Rather, it entails a portfolio of context-dependent, transparent and well-documented methodological approaches that are comparable because they adhere to shared reporting principles. PCR conditions and primer sequences should be consistently reported, as they are a critical determinant of taxonomic coverage and amplification bias. The cost of poor metadata standardisation, in terms of wasted research investment, duplicated efforts and reduced funding efficiency, was not quantified in this study, but the authors describe it as a substantial hidden burden on the research ecosystem.

As Europe moves toward coordinated aquatic biodiversity monitoring at scale, the message from this assessment is unambiguous. eDNA can become a cornerstone of conservation strategy, environmental policy and rigorous evaluation of restoration interventions, but only if the data infrastructure catches up with the science. Coordinated efforts between field and laboratory method experts, data custodians and policymakers are essential to embed minimum metadata requirements into harmonised reporting frameworks, ensuring that high-quality metadata becomes a natural outcome of methodological compliance rather than an afterthought. The technology is ready. The question now is whether the community can align its incentives, training and repositories to make every eDNA dataset as reusable as the science it enables.

Subject of Research: Navigating the European aquatic eDNA landscape: Opportunities for metadata standardisation and data mobilisation

Article Title: Navigating the European aquatic eDNA landscape: Opportunities for metadata standardisation and data mobilisation

Article References: Kavakiotis, I., Woollard, P., Babo, C., Egge, E., Krawczyk, D., Laamanen, T., Paupério, J., Bacela-Spychalska, K., Pade, N., Costa, F. O., Grabowski, M., Norros, V., Meissner, K., & Papakostas, S. (2026). Navigating the European aquatic eDNA landscape: Opportunities for metadata standardisation and data mobilisation. Metabarcoding and Metagenomics, 10, Article e173612. https://doi.org/10.3897/mbmg.10.173612

Image Credits: AI Generated

DOI: 10.3897/mbmg.10.173612

Keywords: Navigating, European, aquatic, eDNA, landscape, Opportunities, metadata, standardisation, data, mobilisation, scientific research

The study addresses a specific vulnerability in the Personal Health Train (PHT) paradigm, a manifesto for implementing federated learning with real-world private health data. In this framework, data remain in local repositories called Stations, while algorithms called Trains visit each station to conduct analysis locally, ensuring that individual-level data never leave the originating institution. A central Track orchestrates the interactions and an Aggregator combines the local model gradients from various stations into a single global model. However, recent research has pointed out that certain federated learning methods may become vulnerable to specific attacks that could result in patient data leakage, even when the raw data themselves are never transmitted.

The threat landscape is grouped by the type of adversary involved. Curious participants are legitimate members of the collaboration network who attempt to infer sensitive information about their partners by intercepting partial model updates, potentially carrying out membership inference attacks or even reconstructing training data through model inversion. Colluding participants coordinate their actions jointly, sharing their partial models with each other to mount a coordinated attack against other consortium partners. Malicious attackers from outside might exploit security vulnerabilities to penetrate the federated infrastructure, intercepting communications or injecting poisoned updates to manipulate the global model. The study notes that curious or colluding participants might also employ data poisoning and model poisoning in an attempt to forestall the global model from converging, buying themselves more time to carry out inference or inversion attacks.

The researchers focused on Cox Proportional Hazards models, a cornerstone of clinical survival analysis used for risk-based decision-making in oncology and beyond. Previous work had established that a federated Cox fit reproduces centralized estimates exactly, because the Cox partial log-likelihood decomposes additively over the risk sets shared at each event time, allowing the federated Newton-Raphson procedure to converge to the same maximum-likelihood estimates that a pooled fit on combined data would produce. However, later work drew attention to the residual exposure of such models: the event times and iterative gradient updates that must be shared can leak information about individual patients. What had not been characterized is how far the standard responses to that exposure move the survival estimates away from the exact federated baseline.

To answer this question, the team implemented two Privacy-Enhancing Techniques on top of a Personal Health Train infrastructure using Vantage6, an open-source federated learning framework. The first technique, called Discretized Cox analysis, groups event times into finite bins before they are sent to the central server for aggregation. Because several distinct event times fall within each bin, every value that leaves a station is shared by multiple events, so an individual event time can no longer be singled out from the others in its bin. This provides a k-anonymity-style effect. The researchers explored two binning strategies: fixed binning, where events are divided into equal-sized bins across all data stations, and quantile binning, where events are categorized based on an optimal number of quantiles so that events are evenly distributed across each bin. The optimal number of bins was determined using Sturges’ rule, resulting in 12 bins for the datasets used.

The second technique applies Local Differential Privacy by introducing calibrated noise to the local model gradients before they are transmitted. Each station adds Laplacian noise to its local gradients, so the noise is introduced at the data source and the guarantee does not depend on trusting the server or the aggregator. Formally, this implements epsilon-local differential privacy through the Laplace mechanism, where the noise scale is calibrated to the L1-sensitivity of the gradient and a privacy-loss budget epsilon. A smaller epsilon means more noise and stronger privacy. The researchers conducted a series of Cox models to determine the optimal epsilon, finding that the Harrell’s concordance index remains stable when the noise scale is low but decreases beyond a threshold of around 4, indicating a decline in model accuracy due to excessive noise. They selected a noise scale of 3.3, corresponding to a sensitivity of 1 and epsilon of 0.3, to balance added noise and model accuracy. They also tested a variant in which Laplacian noise was additionally applied to the input predictors for 25 percent of patients at each station before the gradients were computed.

The experiments used three public datasets from The Cancer Imaging Archive: HN1 with 137 patients, HEAD-NECK with 298 patients, and OPC with 606 patients, comprising 1,041 patients in total with differing event compositions. A private Maastro dataset, HN3 with 165 patients, provided an independent check that the chosen noise scale generalized. The results showed that both fixed and quantile binning preserved the estimated hazard ratios, with the Wald test statistics for the predictors in each privacy-enhanced model being relatively low and p-values considerably above the 0.05 significance level, suggesting no statistically significant difference between the coefficients of the privacy-enhanced and standard models. Adding Laplacian noise to the local gradients likewise maintained model fidelity, with the hazard ratios remaining consistent with those of the standard Cox model.

For the cumulative baseline hazard curves, computed using the Breslow estimator, the picture was more nuanced. Fixed binning showed a steep initial rise, suggesting a slight compromise in the hazard estimation even as privacy improved. Quantile binning closely aligned with the standard Cox model during the initial follow-up periods but showed some divergence later, which may indicate increased sensitivity to data sparsity over time. When additional noise was applied to a 25 percent subset of the predictors along with the noise on local gradients, the results revealed a more pronounced divergence, suggesting that while the combination further enhances privacy, it may introduce a degree of overestimation in the baseline hazard values, especially over longer follow-up periods. The researchers note that the number of individuals still at risk falls sharply toward the end of follow-up, and the largest divergences fall in this sparse region where estimates are inherently unstable, so they restrict their interpretation to the period over which the risk set remains adequate.

The study is careful about the limits of its claims. The authors emphasize that this work does not claim that either technique prevents membership inference or reconstruction attacks outright, as no attack was actually implemented. Their evidence concerns the effect of each technique on the model outputs rather than a measured reduction in attack success. They also note that in a healthcare consortium, membership inference is only one component of re-identification risk and is not equivalent to identifying a specific individual, and that this risk is also managed at the governance level through federated learning agreements that constrain who participates and prohibit such attacks. Not every risk therefore needs to be solved mathematically.

The findings suggest practical options for multi-institutional research while reducing the exposure that could enable re-identification. In most federated modeling, end-to-end encryption may provide sufficient privacy protection, but in more sensitive cases where additional layers are necessary, both binning and differential privacy offer valid options. Researchers could begin with binning event times and then layer differential privacy as required, adjusting the noise based on privacy requirements. Binning raises the difficulty of model-inversion attacks that attempt to reverse-engineer individual data points from shared values, while differential privacy reduces the information any single update carries about an individual. The choice and configuration of each technique should be guided by the sensitivity of the data and the requirements of the specific project, with future work focusing on optimizing these strategies for larger datasets, more institutions, and other clinical applications.

Subject of Research: Navigating the privacy-accuracy tradeoff: federated survival analysis with binning and differential privacy

Article Title: Navigating the privacy-accuracy tradeoff: federated survival analysis with binning and differential privacy

Article References: Gouthamchand, V., van Soest, J., Arcuri, G., Dekker, A., Damiani, A., & Wee, L. (2026). Navigating the privacy-accuracy tradeoff: federated survival analysis with binning and differential privacy. Frontiers in Applied Mathematics and Statistics, 12, Article 1917462. https://doi.org/10.3389/fams.2026.1917462

Image Credits: AI Generated

DOI: 10.3389/fams.2026.1917462

Keywords: Navigating, privacy-accuracy, tradeoff, federated, survival, analysis, binning, differential, privacy, scientific research

Breast cancer is the second most common cancer among women in the United States, with an estimated 310,720 new cases in 2024 alone, and it remains the second-leading cause of cancer-related mortality in women. While survivorship brings physical, emotional, and cognitive challenges for all patients, rurality adds structural, socioeconomic, and environmental barriers that urban patients rarely face. Rural survivors often must travel hours for treatment, live in areas with limited broadband access, and contend with a scarcity of mental health professionals. Studies consistently show that rural breast cancer survivors report higher levels of anxiety and depression than their urban counterparts, and evidence suggests they may experience poorer prognoses despite lower incidence rates, a persistent inequality in cancer survivorship.

Expressive writing, first pioneered by psychologist James Pennebaker and colleague Sandra Beall in 1986, asks participants to write continuously for at least 15 minutes about their deepest thoughts and feelings surrounding traumatic life events. Decades of research have linked the practice to improved emotional well-being, reduced psychological distress, and enhanced quality of life among cancer survivors, particularly those with low emotional support. Yet, as the researchers point out, nearly all prior studies analyzed the written essays themselves. None had asked rural survivors, in their own words, what the experience of expressive writing actually felt like or how it fit into their cancer journeys. The new study was designed to fill that gap.

The interviews drew from a parent randomized controlled trial conducted between October 2021 and April 2022, in which rural breast cancer survivors were assigned either to write about their deepest thoughts and feelings about cancer or to write about neutral topics. The intervention spanned six weeks, with three writing tasks scheduled two weeks apart; participants could type or write by hand, and follow-up questionnaires were collected at one, three, and six months. For the qualitative follow-up, the research team invited participants from the expressive writing arm to complete semi-structured interviews of 45 to 60 minutes each, 13 conducted via Zoom and one in person, during November and December 2023. The mean age of interviewees was 54.71 years, most identified as White, three identified as Hispanic, and cancer stages ranged from stage I to stage III.

Using an inductive thematic analysis, guided by principles of reflexivity, the team identified six interrelated themes. The first, rural-specific challenges, confirmed the structural weight these women carry. One 62-year-old survivor with stage III disease described slow internet connections that hampered virtual care, while another recounted a weekly drive of three and a half to four hours to reach her treatment center. More unexpectedly, rural community dynamics itself created barriers: in towns of a few hundred people, privacy is scarce. One 67-year-old participant from a town of 235 residents said she avoided local disclosure because everybody knows her and neighbors are, in her words, terrible gossips. The writing intervention, conducted privately at home, sidestepped that social constraint entirely.

The second theme centered on emotion regulation and adaptive coping. Survivors described expressive writing as a cathartic outlet, a safe and private space to release frustrations they could not voice to anyone. One 32-year-old stage III survivor said the writing helped her release frustrations she could not openly talk about, while another explained that when she cannot talk to anybody, she writes it down to let it go. Beyond venting, participants used the pages as a problem-solving tool, identifying stressors, examining what made situations more stressful, and thinking through strategies to get through them. One participant noted the writing was particularly important for managing lingering memory problems from chemotherapy, suggesting the practice served cognitive as well as emotional functions.

The remaining themes mapped closely onto the post-traumatic growth framework, the well-established model describing positive psychological change after trauma across five domains: personal strength, improved relationships, new possibilities, appreciation of life, and spiritual growth. Survivors reported personal strength through meaning-making, describing how writing helped them summarize and better understand what had happened to them, reframe their diagnosis in a resilient light, and even find hope and love within a traumatic experience. The researchers emphasize that trauma alone does not produce post-traumatic growth; deliberate rumination, the kind of structured narrative reconstruction that expressive writing facilitates, is considered a strong precursor of growth.

Interpersonal benefits emerged just as clearly. Participants said writing helped them recognize that others were going through the same thing, validating their feelings and reducing the profound isolation of rural survivorship. A notable subtheme of altruism also appeared: several women framed their participation as a way to help someone else facing breast cancer, echoing research showing that survivors often develop increased empathy and altruistic behavior after cancer. Others described renewed purpose, expressing a desire to be needed and to contribute meaningfully rather than simply be taken care of, and a heightened appreciation for life, from gratitude for simply being alive and having a choice to renewed appreciation for physical abilities like walking that were once taken for granted. Spiritual growth, by contrast, was mentioned by only one participant and did not emerge as a pervasive theme, consistent with surveys identifying spiritual change as the least pronounced growth domain.

The authors caution that their findings come with limitations. The sample consisted of intervention completers who agreed to be interviewed, which may bias results toward participants who enjoyed the writing and found it beneficial; dissenting voices may be underrepresented. Social desirability concerns, the homogeneity of the largely White, digitally connected, female sample, and the retrospective, self-reported nature of the accounts all constrain generalizability. Meta-analyses of expressive writing across cancer populations have found small or null pooled effects, warning against overgeneralizing its efficacy. Still, the researchers argue that rural survivors, a subgroup facing heightened disclosure constraints and scarce psychosocial resources, may be precisely the population that benefits most from targeted expressive writing interventions emphasizing post-traumatic growth.

The practical implications are considerable. Because expressive writing is low-cost, private, and scalable, it can reach women whom geography, stigma, and infrastructure leave behind. The authors recommend that future virtual interventions account for the rural digital divide by offering low-bandwidth or offline options with phone-based follow-up, embedding writing tasks in patient portals, and integrating in-person care where feasible. Future studies should triangulate interviews with essay content and objective biomarkers to clarify the mechanisms linking writing to growth. For now, the testimony of these 14 women makes a compelling case that the page can be a powerful place: a private room of one’s own where rural survivors can release what they cannot say aloud, rewrite their trauma into meaning, and discover strength, connection, and gratitude on the other side of cancer.

Subject of Research: Expressive writing as a psychosocial intervention promoting post-traumatic growth among rural breast cancer survivors

Article Title: Navigating rural challenges and promoting posttraumatic growth via expressive writing: a qualitative study among rural breast cancer survivors

Article References: Xu, Y. A., Choi, E., Wong-Meli, C. C. Y., Borjas, M., Fei, F., Mann, S., Jin, W., Dawkins-Moultin, L., Lun, D., & Lu, Q. (2026). Navigating rural challenges and promoting posttraumatic growth via expressive writing: a qualitative study among rural breast cancer survivors. Supportive Care in Cancer, 34(10), Article 965. https://doi.org/10.1007/s00520-026-11204-w

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11204-w

Keywords: expressive writing, rural health, breast cancer, cancer survivors, post-traumatic growth, qualitative research, psychosocial intervention, cancer disparities, emotion regulation, telehealth, survivorship care, Navigating