Researchers tested whether generative AI could handle complex medical datasets as well as human experts. In some cases, the AI matched or outperformed teams that had spent months building prediction models. By generating usable analytical code from precise prompts, the systems dramatically reduced the time needed to process health data. The findings hint at a future where AI helps scientists move faster from data to discovery.
Tag: nevin manimala
Am J Clin Pathol. 2026 Jan 5;165(2):aqaf153. doi: 10.1093/ajcp/aqaf153.
ABSTRACT
OBJECTIVE: To evaluate the ability of 4 artificial intelligence large language models (LLMs) to create items that align with the item writing standards of the American Board of Pathology (ABPath) for continuing certification.
METHODS: An informatics item writing application was developed and used with prompts based on the ABPath item writing standards. Uniform prompts were used for the LLMs evaluated, with the content of the items generated tailored to the expertise of the reviewing subject matter experts (SMEs). The SMEs were blinded to the identity of the LLM that generated each item. The 14 SMEs graded 4 written items and 4 practical items, with 1 item from each set of 4 generated from each of the LLMs. The 19 questions used for grading concentrated on item anatomy (ie, item structure), accuracy, relevance, and level of item difficulty.
RESULTS: The overall scores for the 4 LLMs for the written items were as follows: Claude, 229 of 266 (86.1%); ChatGPT, 212 of 266 (79.7%); Llama, 175 of 266 (65.8%); and Titan, 162 of 266 (60.9%). The overall scores for the 4 LLMs for the practical items were as follows: Claude, 247 of 266 (92.9%); ChatGPT, 216 of 266 (81.2%); Llama, 175 of 266 (65.8%); and Titan, 151 of 266 (56.8%). Statistically significant differences existed between the LLMs.
CONCLUSIONS: We observed significant differences in the ability of the 4 LLMs evaluated to draft items consistent with the ABPath guidelines based on SME scoring. It is important to assess the various LLMs available to determine which model best meets the needs of the user for the proposed task and not to assume equivalence.
PMID:41722024 | DOI:10.1093/ajcp/aqaf153
Oncol Ther. 2026 Feb 21. doi: 10.1007/s40487-026-00418-x. Online ahead of print.
ABSTRACT
INTRODUCTION: Venetoclax represents a significant advancement in target anticancer therapy in the management of relapsed/refractory chronic lymphocytic leukemia (RR-CLL). This study aims to compare the efficacy and safety of the venetoclax + rituximab (VenR) regimen with other therapies approved in Brazil.
METHODS: A systematic review and network meta-analysis (NMA) was conducted to evaluate the efficacy and safety of treatments approved in Brazil for RR-CLL. Comprehensive literature searches were performed to identify randomized controlled trials. Risk of bias and certainty of evidence for each outcome were assessed across the included studies. The NMA was conducted using a frequentist framework. The primary efficacy outcomes were progression-free survival, overall survival, overall response rate, and time to next therapy. Safety was assessed on the basis of the incidence of serious adverse events.
RESULTS: A total of 24 publications related to 12 trials were identified and included. VenR was associated with better survival outcomes when compared with standard regimens such as rituximab (HR 0.16; 95% CI 0.03-0.74) and physician choice (HR 0.17; 95% CI 0.04-0.81). In terms of progression-free survival, VenR achieved HR < 0.20, supported by narrow confidence intervals, when compared to treatments such as bendamustine plus rituximab, ofatumumab and physician choice, in addition to significantly favorable results compared to ibrutinib and acalabrutinib. Approximately half of the studies presented a low risk of bias, and the certainty of evidence assessed using the GRADE-NMA approach resulted in very low certainty of evidence, mainly due to risk of bias, intransitivity, and imprecision.
CONCLUSION: This NMA provides valuable evidence to support rational therapeutic choices for RR-CLL in Brazil, highlighting VenR and Bruton tyrosine kinase inhibitors as leading treatment options across diverse clinical scenarios.
PMID:41722016 | DOI:10.1007/s40487-026-00418-x
J Patient Rep Outcomes. 2026 Feb 21. doi: 10.1186/s41687-026-01021-4. Online ahead of print.
ABSTRACT
BACKGROUND: Prosthetic sockets are fundamental in stabilizing the residual limb and providing a secure attachment for the prosthesis, therefore aiding in achieving a more natural and efficient walking motion. Patients who are fit with a prosthesis report increased mobility which correlates to increased quality of life and satisfaction. While timely receipt of an initial prosthesis following amputation improves patient outcomes, the impact of the socket or prosthesis being replaced is less documented. The purpose of this analysis was to assess the change in outcomes resulting from socket and prosthesis replacements.
METHODOLOGY: This study included patients within a national database from a national privately owned provider of prosthetic care. Inclusion criteria consisted of adults with a unilateral, lower limb amputation that received either a replacement socket or prosthesis. Data were collected during routine clinical care with questionnaires about quality of life and satisfaction (Prosthesis Evaluation Questionnaire-Well Being), and mobility (Prosthesis Limb Users Survey of Mobility). Statistical analyses consisted of Student’s t-tests and estimated marginal means to determine significant changes within each group. Hours worn was collected as a secondary variable.
RESULTS: A total of 19,185 longitudinal outcomes were included in the final analysis and divided into four groups based on amputation level and socket or prosthesis replacement status (above knee socket replacement n = 2,515, below knee socket replacement n = 7,746, above knee replacement prosthesis n = 1,717, and below knee replacement prosthesis n = 7,207). Quality of life, satisfaction, and mobility improved significantly across all groups (p < 0.01), with hours worn increasing after socket replacement in both groups (p < 0.01) but not in the prosthesis replacement groups. Individuals with vascular disease or diabetes had reduced mobility compared to other etiologies. Older age negatively impacted mobility.
CONCLUSIONS: On average, patients with lower limb amputation receiving an above-knee or below-knee socket replacement or prosthesis replacement can expect an improvement in their mobility and well-being.
CLINICAL RELEVANCE: The findings highlight the clinical relevance of prosthetic socket and prosthesis replacements by demonstrating significant improvements in quality of life, satisfaction, and mobility for patients with lower limb amputations. This evidence supports the importance of socket and prosthesis replacements in enhancing patient outcomes and informs evidence-based clinical practices. By addressing gaps in the literature, this research provides valuable insights that can guide prosthetists and physicians in optimizing care. Overall, these results underscore the long-term benefits of replacements, advocating for prioritizing resources and support for prosthetic care to improve health outcomes and quality of life for individuals with an amputation.
PMID:41722013 | DOI:10.1186/s41687-026-01021-4
Clin Transl Oncol. 2026 Feb 21. doi: 10.1007/s12094-026-04238-3. Online ahead of print.
ABSTRACT
BACKGROUND: Cervical cancer continues to be a major worldwide health concern. Diagnostic methods are limited, and therefore, more specific, non-invasive serum biomarkers are required for greater efficacy in detecting and monitoring the disease.
OBJECTIVE: This research aims to investigate the diagnostic and prognostic utility of serum Cytokeratin 5 (CK5) and p63 as biomarkers for cervical cancer.
MATERIAL AND METHODS: A case-control study was conducted with 50 women with newly diagnosed, histologically confirmed cervical cancer and 50 age-matched healthy controls. Serum levels of CK5 and p63 were quantified using enzyme-linked immunosorbent assay (ELISA).
RESULTS: Serum levels of both CK5 (286 ± 94 vs. 77.5 ± 28.2 ng/L, p < 0.001) and p63 (35.1 ± 10.6 vs. 7.5 ± 3.2 ng/mL, p < 0.001) were significantly elevated in cervical cancer patients compared to controls. ROC analysis revealed excellent diagnostic accuracy, with an area under the curve (AUC) of 0.96 for CK5 and 0.97 for p63 both biomarkers being statistically significantly more in advanced-stage than in early-stage of the disease on p = 0.002 and p < 0.0001, respectively, and independent predictors of advanced disease in multivariate analysis.
CONCLUSION: Serum CK5 and p63 had an excellent diagnostic efficiency for cervical cancer and correlated with disease progression, positioning them as promising non-invasive biomarkers for diagnosis and staging.
PMID:41722012 | DOI:10.1007/s12094-026-04238-3
Ophthalmol Ther. 2026 Feb 21. doi: 10.1007/s40123-026-01330-3. Online ahead of print.
ABSTRACT
INTRODUCTION: To determine the incidence, clinical characteristics, and spatial associations of retinal arterial vasospasm in retinal vein occlusion (RVO), and to explore its relationship with aqueous humor endothelin-1 (ET-1) concentration.
METHODS: This retrospective cohort study included 58 eyes with RVO that underwent wide-field fluorescein angiography (FA) within 1 month of presentation. In addition, aqueous humor samples were prospectively collected from a subset of 18 treatment-naïve RVO eyes. Arterial vasospasm was defined as a focal narrowing of the arterial lumen on early-phase FA. Cotton wool spots (CWS) and retinal hemorrhage were evaluated using color fundus photography and electronic medical records. Topographic concordance was assessed using quadrant-based, horizontal, and vertical hemifield classifications. Aqueous humor samples were analyzed for ET-1 concentration using enzyme-linked immunosorbent assay (ELISA).
RESULTS: Arterial vasospasm was identified in 21 eyes (36.2%), including six eyes with central RVO and 15 eyes with branch RVO. Vasospasm was most commonly observed in the superotemporal quadrant and predominantly on second-order arteriolar branches (78.6%). CWS were more frequent in eyes with vasospasm compared with those without (81.0% vs. 54.1%; p = 0.050). Significant spatial concordance between vasospasm and CWS was observed in the vertical hemifield (p = 0.049), with no concordance observed with retinal hemorrhage location or severity. Exploratory aqueous humor analysis showed higher ET-1 concentrations in eyes with vasospasm (8.37 ± 4.59 vs. 5.74 ± 1.88 pg/mL), although statistical interpretation was limited by the sample size.
CONCLUSION: Retinal arterial vasospasm is a relatively common but under-recognized feature of RVO. Its association with CWS and higher ET-1 levels suggests localized arterial endothelial dysfunction. Vasospasm may represent an arterial component of RVO pathophysiology and warrants further investigation.
PMID:41722008 | DOI:10.1007/s40123-026-01330-3
Eur J Epidemiol. 2026 Feb 21. doi: 10.1007/s10654-025-01342-6. Online ahead of print.
ABSTRACT
Ease of access to big data and automated analysis tools can facilitate the rapid generation of poorly designed epidemiological studies, which collectively pose a risk to the quality of medical literature. Member organizations of the TriNetX network have the ability to mass-produce retrospective cohort studies at speed using the federated data network’s statistical power and streamlined analytics pipeline. This exploratory meta-research study collated 13 published TriNetX-based retrospective cohort studies that claim to have used a design that is, in fact, impossible on the platform (the setting of a pseudo-index event on the TriNetX platform). Of these, 8 studies described their analysis as being conducted on the platform alone, making their description of the index event impossible. When we queried seven different generative artificial intelligence (AI) tools for advice on how to set an index event on TriNetX, six tools suggested at least one strategy that cannot be implemented on the platform. Unlike previously documented errors in TriNetX-based studies, we argue that the reporting of impossible index event designs in the identified publications likely constitute either distortion of the reported methods or the uncritical adoption of false AI-generated methodological advice. In an age of accelerating and increasingly automated medical research, editors and peer-reviewers must be informed of limitations with emerging epidemiological datasets and analytic tools.
PMID:41721987 | DOI:10.1007/s10654-025-01342-6
J Mol Neurosci. 2026 Feb 21;76(1):36. doi: 10.1007/s12031-026-02494-0.
NO ABSTRACT
PMID:41721963 | DOI:10.1007/s12031-026-02494-0
Adm Policy Ment Health. 2026 Feb 21. doi: 10.1007/s10488-026-01492-z. Online ahead of print.
NO ABSTRACT
PMID:41721949 | DOI:10.1007/s10488-026-01492-z
Bull Math Biol. 2026 Feb 21;88(3):42. doi: 10.1007/s11538-026-01593-8.
ABSTRACT
This study explores the role of mass testing in controlling the COVID-19 pandemic using an age-stratified compartmental model. The model evaluates the impact of different testing strategies on the pandemic’s reproduction number, , while also considering social distancing measures and demographic characteristics. The analysis highlights the importance of combining mass testing with isolation strategies to reduce the spread of the virus. The simulations demonstrate that in countries characterized by high levels of elderly cohabitation with younger individuals, vertical isolation is insufficient; horizontal isolation with work restrictions, alongside testing and susceptibility reduction measures, is crucial. For aged developed countries, where cohabitation of the elderly with younger individuals is less prevalent, and for least developed countries, where the population has a predominantly youthful age structure, pandemic control is more feasible with fewer tests. The study also emphasizes the critical role of identifying asymptomatic cases to achieve optimal epidemic control. Lastly, the cost-effectiveness of various testing strategies is examined, providing insights for public health policy decision-making.
PMID:41721946 | DOI:10.1007/s11538-026-01593-8