Category: Nevin Manimala

Eye (Lond). 2024 Jul 13. doi: 10.1038/s41433-024-03231-w. Online ahead of print.

ABSTRACT

PURPOSE: This study aimed to evaluate the accuracy of information that patients can obtain from large language models (LLMs) when seeking answers to common questions about choroidal melanoma.

METHODS: Comparative study comparing frequently asked questions from choroidal melanoma patients and queried three major LLMs-ChatGPT 3.5, Bing AI, and DocsGPT. Answers were reviewed by three ocular oncology experts and scored as accurate, partially accurate, or inaccurate. Statistical analysis compared the quality of responses across models.

RESULTS: For medical advice questions, ChatGPT gave 92% accurate responses compared to 58% for Bing AI and DocsGPT. For pre/post-op questions, ChatGPT and Bing AI were 86% accurate while DocsGPT was 73% accurate. There were no statistically significant differences between models. ChatGPT responses were the longest while Bing AI responses were the shortest, but length did not affect accuracy. All LLMs appropriately directed patients to seek medical advice from professionals.

CONCLUSION: LLMs show promising capability to address common choroidal melanoma patient questions at generally acceptable accuracy levels. However, inconsistent, and inaccurate responses do occur, highlighting the need for improved fine-tuning and oversight before integration into clinical practice.

PMID:39003430 | DOI:10.1038/s41433-024-03231-w

Mol Psychiatry. 2024 Jul 14. doi: 10.1038/s41380-024-02665-8. Online ahead of print.

ABSTRACT

Genetics research has potential to alleviate the burden of mental disorders in low- and middle-income-countries through identification of new mechanistic pathways which can lead to efficacious drugs or new drug targets. However, there is currently limited genetics data from Africa. The Uganda Genome Resource provides opportunity for psychiatric genetics research among underrepresented people from Africa. We aimed at determining the prevalence and correlates of major depressive disorder (MDD), suicidality, post-traumatic stress disorder (PTSD), alcohol abuse, generalised anxiety disorder (GAD) and probable attention-deficit hyperactivity disorder (ADHD) among participants of the Uganda Genome Resource. Standardised tools assessed for each mental disorder. Prevalence of each disorder was calculated with 95% confidence intervals. Multivariate logistic regression models evaluated the association between each mental disorder and associated demographic and clinical factors. Among 985 participants, prevalence of the disorders were: current MDD 19.3%, life-time MDD 23.3%, suicidality 10.6%, PTSD 3.1%, alcohol abuse 5.7%, GAD 12.9% and probable ADHD 9.2%. This is the first study to determine the prevalence of probable ADHD among adult Ugandans from a general population. We found significant association between sex and alcohol abuse (adjusted odds ratio [AOR] = 0.26 [0.14,0.45], p < 0.001) and GAD (AOR = 1.78 [1.09,2.49], p = 0.019) respectively. We also found significant association between body mass index and suicidality (AOR = 0.85 [0.73,0.99], p = 0.041), alcohol abuse (AOR = 0.86 [0.78,0.94], p = 0.003) and GAD (AOR = 0.93 [0.87,0.98], p = 0.008) respectively. We also found a significant association between high blood pressure and life-time MDD (AOR = 2.87 [1.08,7.66], p = 0.035) and probable ADHD (AOR = 1.99 [1.00,3.97], p = 0.050) respectively. We also found a statistically significant association between tobacco smoking and alcohol abuse (AOR = 3.2 [1.56,6.67], p = 0.002). We also found ever been married to be a risk factor for probable ADHD (AOR = 2.12 [0.88,5.14], p = 0.049). The Uganda Genome Resource presents opportunity for psychiatric genetics research among underrepresented people from Africa.

PMID:39003415 | DOI:10.1038/s41380-024-02665-8

Pharmaceut Med. 2024 Jul 13. doi: 10.1007/s40290-024-00530-1. Online ahead of print.

ABSTRACT

INTRODUCTION: Several quantitative methods have been established, in pharmacovigilance, to detect signals of disproportionate reporting (SDRs) from databases containing reports of adverse drug reactions (ADRs). The signal detection algorithms (SDAs) and the source of the reporting per product vary, but it is unclear whether any algorithm can provide satisfactory performance using data with such large variance factors.

OBJECTIVE: Determine the appropriate SDA for Biogen’s internal Global Safety Database (GSD) given the characteristics of the database including frequencies of events, data skewness, outliers, and missing information. Compare performance of standard approaches (EBGM, EB05, PRR, and ROR), well accepted by industry, to a Biogen-developed Machine Learning (ML) Regression Decision Tree (RDT) model, across several Biogen products, to determine a champion SDA.

METHODS: All data associated with seven marketed Biogen products were chosen and a historical subset of reported ADRs were considered. Six SDAs (five common industry disproportionality methods) and RDT were evaluated. The SDRs were calculated on training and test data composed of quarterly reporting intervals from 2004-2019. The performance measures used were sensitivity, precision, time to detect new events, and frequency of detected cases for each algorithm for each product. Outcomes in the test data are known a priori and easily compared to predicted outcomes. Validation was performed via rates of misclassification. This work solely represents Biogen’s internal information, intentionally chosen to serve the performance review of its signal detection systems, and results will not necessarily be generalizable to other external sources.

RESULTS: Several algorithms performed differently among products, but no one method dominated any other. Performance was dependent on the thresholds used to define a signal according to different criteria. However, those different statistics subtly influenced the achievable performance. The relative performance of RDT and Medicines and Healthcare products Regulatory Agency (MHRA) algorithms were superior and paired across products. A reduction in precision for all methods spanning the products was present. Hence, companies evaluating signal detection approaches, search for innovative methods to minimize this effect.

CONCLUSIONS: In designing signal detection systems, careful consideration should be given to the criteria that are used to define SDRs. The choice of disproportionality statistics does not affect the achievable range of signal detection performance. These choices should consider mainly ease of implementation and interpretation. The implementation of a method is specific to its accuracy. The RDT attempted to take advantage of known methods and compare results on a per-product basis. Many factors influencing ADRs may improve RDT in future efforts. In this experiment, RDT demonstrated superiority in terms of quickest time to detect and capturing of the highest number of ADRs. Next steps include expansion of data for products representing other indications and testing models in external databases to investigate generalizability of estimates when comparing SDAs.

PMID:39003400 | DOI:10.1007/s40290-024-00530-1

Ann Surg Oncol. 2024 Jul 13. doi: 10.1245/s10434-024-15786-9. Online ahead of print.

ABSTRACT

BACKGROUND: The prognostic impact of genetic mutations for patients who undergo cytoreductive surgery (CRS) with hyperthermic intraperitoneal chemotherapy (HIPEC) of colorectal origin (CRC) is not well defined.

OBJECTIVE: We aimed to describe the genetic classifications in an unsupervised fashion, and the outcomes of this patient population.

METHODS: A retrospective, bi-institutional study was performed on patients who underwent CRS-HIPEC with targeted mutation data with a median follow-up time of 61 months. Functional link analysis was performed using STRING v11.5. Genes with similar functional significance were clustered using unsupervised k-means clustering. Chi-square, Kaplan-Meier, and the log-rank test were used for comparative statistics.

RESULTS: Sixty-four patients with peritoneal carcinomatosis from CRC origin underwent CRS-HIPEC between 2007 and 2022 and genetic mutation data were extracted. We identified 19 unique altered genes, with KRAS (56%), TP53 (33%), and APC (22%) being the most commonly altered; 12.5% had co-altered KRAS/TP53. After creating an interactome map, k-means clustering revealed three functional clusters. Reactome Pathway analysis on three clusters showed unique pathways (1): Ras/FGFR3 signaling; (2) p53 signaling; and (3): NOTCH signaling. Seventy-one percent of patients in cluster 1 had KRAS mutations and a median overall survival of 52.3 months (p < 0.05).

CONCLUSIONS: Patients with peritoneal carcinomatosis (PC) of CRC origin who underwent CRS-HIPEC and with tumors that harbored mutations in cluster 1 (Ras/FGFR3 signaling) had worse outcomes. Pathway disruption and a cluster-centric perspective may affect prognosis more than individual genetic alterations in patients with PC of CRC origin.

PMID:39003380 | DOI:10.1245/s10434-024-15786-9

Aging Clin Exp Res. 2024 Jul 13;36(1):142. doi: 10.1007/s40520-024-02794-8.

ABSTRACT

BACKGROUND: Hip fracture (HF) in older adults is strongly associated with a greater decline in their activities of daily living (ADLs) and health-related (HRQoL). The present study aimed to evaluate the effects of a family-based care transition program (FBCTP) on ADLs, HRQoL and social support in this age group after HF surgery.

METHODS: A quasi-experimental design was conducted on 100 older adults who had undergone HFS and were selected by convenience sampling and allocated to the IG (n = 50) and the CG (n = 50). Data were collected utilizing the Barthel Index, the 12-item Short Form Health Survey (SF-12), and the Multidimensional Scale of Perceived Social Support. The FBCTP was delivered in-hospital education sessions, home visit, and a follow-up and telephone counselling session. The data were collected at three stages, including the baseline, four weeks after discharge, and eight weeks later. The level of statistical significance was set at 0.05.

RESULTS: The results of the study indicated that the effects of time and group on the increase in ADLs were 15.2 and 36.69 (p < 0.000), respectively, following the completion of the FBCTP. Furthermore, time and group were found to have a positive effect on HRQoL, with an increase of 2.82 and 5.60 units, respectively (p < 0.000). In this context, time and group also interacted in the IG compared to the CG, with scores increasing by 1.86 units over time (p < 0.000). Although the study results indicated that social support improved by 1.98 units over time (p < 0.000), the effects of group alone and the time × group interaction were not statistically significant. This indicates that the program was not effective in accelerating social support.

CONCLUSION: Consequently, nurses, policymakers, and planners engaged in geriatric healthcare may utilize these results to enhance the health status of this age group following HFS.

PMID:39002096 | DOI:10.1007/s40520-024-02794-8

J Nephrol. 2024 Jul 13. doi: 10.1007/s40620-024-02028-0. Online ahead of print.

NO ABSTRACT

PMID:39002093 | DOI:10.1007/s40620-024-02028-0

Environ Geochem Health. 2024 Jul 13;46(9):304. doi: 10.1007/s10653-024-02071-7.

ABSTRACT

Accumulating animal studies have demonstrated associations between ambient air pollution (AP) and metabolic dysfunction-associated fatty liver disease (MAFLD), but relevant epidemiological evidence is limited. We evaluated the association of long-term exposure to AP with the risk of incident MAFLD in Northwest China. The average AP concentration between baseline and follow-up was used to assess individual exposure levels. Cox proportional hazard models and restricted cubic spline functions (RCS) were used to estimate the association of PM_2.5 and its constituents with the risk of MAFLD and the dose-response relationship. Quantile g-computation was used to assess the joint effects of mixed exposure to air pollutants on MAFLD and the weights of the various pollutants. We observed 1516 cases of new-onset MAFLD, with an incidence of 10.89%. Increased exposure to pollutants was significantly associated with increased odds of MAFLD, with hazard ratios (HRs) of 2.93 (95% CI: 1.22, 7.00), 2.86 (1.44, 5.66), 7.55 (3.39, 16.84), 4.83 (1.89, 12.38), 3.35 (1.35, 8.34), 1.89 (1.02, 1.62) for each interquartile range increase in PM_2.5, SO₄^2-, NO₃^–, NH₄⁺, OM, and BC, respectively. Stratified analyses suggested that females, frequent exercisers and never-drinkers were more susceptible to MAFLD associated with ambient PM_2.5 and its constituents. Mixed exposure to SO₄^2-, NO₃^–, NH₄⁺, OM and BC was associated with an increased risk of MAFLD, and the weight of BC had the strongest effect on MAFLD. Exposure to ambient PM_2.5 and its constituents increased the risk of MAFLD.

PMID:39002087 | DOI:10.1007/s10653-024-02071-7

Environ Sci Pollut Res Int. 2024 Jul 13. doi: 10.1007/s11356-024-33992-6. Online ahead of print.

ABSTRACT

In the heavy industrial city of Northeast China, there has been a significant decrease in particulate matter pollution while experiencing a sharp increase in ozone (O₃) pollution. However, the main influencing factors and source contributions to O3 remain unclear. Taking the case of Siping as an example, this study analyzed the spatiotemporal characteristics, assessed local source contributions to O₃, and revealed regional transmission effects using numeric simulation and statistical methods. Temporally, higher O₃ concentrations were observed in summer and the afternoon, with hourly peaks up to 254 µg/m³. Spatially, O₃ pollution was mainly contributed by background concentrations (34.52%), external transport (34.50%), and local emissions (30.98%) during the case study period (June 11-18, 2021). Among the local emission sources, biological emissions, the industrial sector, and the traffic sector accounted for 35.30%, 32.09%, and 23.58% of the O₃ concentration, respectively. For regional atmospheric transmission, high O₃ pollution was accompanied by wind from the southwest directions, and the trajectory of air mass transport suggests that eastern Mongolia, the Korean Peninsula, and its neighboring regions contribute to O₃ pollution. Furthermore, sensitivity analysis showed that O₃ pollution in Siping is a co-controlled region by anthropogenic volatile organic compounds (AVOCs) and NO_X, which implies control in an optimal ratio of VOCs and NO_X emissions. Thus, our results highlight the importance of joint prevention and control of O₃ pollution in the region, optimization of biogenic landscape ecology, and control of VOCs and NOx in both the industrial and transport sectors.

PMID:39002081 | DOI:10.1007/s11356-024-33992-6

Breast Cancer Res Treat. 2024 Jul 13. doi: 10.1007/s10549-024-07429-0. Online ahead of print.

ABSTRACT

PURPOSE: To establish a reliable machine learning model to predict malignancy in breast lesions identified by ultrasound (US) and optimize the negative predictive value to minimize unnecessary biopsies.

METHODS: We included clinical and ultrasonographic attributes from 1526 breast lesions classified as BI-RADS 3, 4a, 4b, 4c, 5, and 6 that underwent US-guided breast biopsy in four institutions. We selected the most informative attributes to train nine machine learning models, ensemble models and models with tuned threshold to make inferences about the diagnosis of BI-RADS 4a and 4b lesions (validation dataset). We tested the performance of the final model with 403 new suspicious lesions.

RESULTS: The most informative attributes were shape, margin, orientation and size of the lesions, the resistance index of the internal vessel, the age of the patient and the presence of a palpable lump. The highest mean negative predictive value (NPV) was achieved with the K-Nearest Neighbors algorithm (97.9%). Making ensembles did not improve the performance. Tuning the threshold did improve the performance of the models and we chose the algorithm XGBoost with the tuned threshold as the final one. The tested performance of the final model was: NPV 98.1%, false negative 1.9%, positive predictive value 77.1%, false positive 22.9%. Applying this final model, we would have missed 2 of the 231 malignant lesions of the test dataset (0.8%).

CONCLUSION: Machine learning can help physicians predict malignancy in suspicious breast lesions identified by the US. Our final model would be able to avoid 60.4% of the biopsies in benign lesions missing less than 1% of the cancer cases.

PMID:39002069 | DOI:10.1007/s10549-024-07429-0

Breast Cancer Res Treat. 2024 Jul 13. doi: 10.1007/s10549-024-07426-3. Online ahead of print.

ABSTRACT

PURPOSE: We evaluated T- and B-cell receptor (TCR and BCR) repertoire diversity and 38 serum cytokines in pre- and post-treatment peripheral blood of 66 patients with triple-negative breast cancer (TNBC) who received neoadjuvant chemotherapy plus durvalumab and assessed associations with pathologic response and immune-related adverse events (irAEs) during treatment.

METHODS: Genomic DNA was isolated from buffy coat for TCR and BCR clonotype profiling using the Immunoseq platform and diversity was quantified with Pielou’s evenness index. MILLIPLEX MAP Human Cytokine/Chemokine Magnetic Bead Panel was used to measure serum cytokine levels, which were compared between groups using moderated t-statistic with Benjamini-Hochberg correction for multiple testing.

RESULTS: TCR and BCR diversity was high (Pielou’s index > 0.75) in all samples. Baseline receptor diversities and change in diversity pre- and post-treatment were not associated with pathologic response or irAE status, except for BCR diversity that was significantly lower post-treatment in patients who developed irAE (unadjusted p = 0.0321). Five cytokines increased after treatment in patients with pathologic complete response (pCR) but decreased in patients with RD, most prominently IL-8. IFNγ, IL-7, and GM-CSF levels were higher in pre-treatment than in post-treatment samples of patients who developed irAEs but were lower in those without irAEs.

CONCLUSION: Baseline peripheral blood cytokine levels may predict irAEs in patients treated with immune checkpoint inhibitors and chemotherapy, and increased post-treatment B-cell clonal expansion might mediate irAEs.

PMID:39002068 | DOI:10.1007/s10549-024-07426-3