-
Australasian Psychiatry : Bulletin of... Jun 2024To determine whether a brief educational intervention for Junior Medical Officers (JMOs), using teaching methods aimed at achieving higher outcomes on Bloom's Taxonomy,...
OBJECTIVE
To determine whether a brief educational intervention for Junior Medical Officers (JMOs), using teaching methods aimed at achieving higher outcomes on Bloom's Taxonomy, significantly improved participant confidence and knowledge in decision making about restrictive care.
METHOD
JMOs received a teaching session on restrictive medical and mental health care. Groups were randomly assigned to either sessions including a component of modern pedagogical interventions (Think-Pair-Share and SNAPPS), or sessions including a control period focusing on reviewing a condensed summary of relevant information. Pre- and post-intervention measures were recorded for subjective self-ratings of confidence and scores on standardized clinically relevant extended matching questions (EMQs).
RESULTS
There was no difference in subjective confidence improvement between groups; however, the group receiving the modern pedagogical intervention demonstrated significantly greater objective performance on knowledge-based EMQs.
CONCLUSIONS
A brief modern pedagogical intervention using interactive teaching methods shows promise for improving knowledge of restrictive care and the Mental Health and Guardianship Acts. In the control group, similarly increased confidence in knowledge did not equate to increased competence on a knowledge assessment. Refurbishing educational interventions presents opportunities for improving clinical outcomes and engaging junior doctors in psychiatry.
PubMed: 38876497
DOI: 10.1177/10398562241260170 -
International Journal of Surgery Case... Jun 2024The most frequent location of thrombosis development in acute mesenteric venous thrombosis is the superior mesenteric vein. It is an uncommon but potentially fatal...
INTRODUCTION
The most frequent location of thrombosis development in acute mesenteric venous thrombosis is the superior mesenteric vein. It is an uncommon but potentially fatal condition. Patients with underlying medical conditions that interfere with the Virchow Triad hypercoagulability, stasis, and endothelial injury are more likely to experience it.
PRESENTATION
A 37-year-old female reported to our emergency department with a 5-day history of severe abdominal discomfort, vomiting, and constipation, as well as two episodes of bleeding per rectum. The patient had a clean medical history, no HTN, no diabetes, no chronic medication, no history of contraceptive pill use or non-steroid anti-inflammatory drug use, no history of chronic disease or operation. Patient was directly transferred to the intensive care unit for additional evaluation and preoperative stabilization.
DISCUSSION
A patient with acute mesenteric venous thrombosis and possible intestinal damage is the case we've presented. Upon presentation patient was unstable, we assessed her condition and transferred to the intensive care unit for stabilization and pre-operative preparation. She didn't respond to conservative management and we had to operate, we highly emphasize how crucial it is for early intervention in these type of conditions. Acute mesenteric venous thrombosis is a complicated case due to its nonspecific symptoms, it requires a multidisciplinary team approach between internal medicine and surgical team to plan for the most appropriate treatment strategy suitable for each patient as all options are associated with significant risks. Multiple options are available for the management of mesenteric venous thrombosis. In patients with peritoneal signs to suggestive bowel infarction or perforation or those who failed to progress with conservative management, operative intervention may be necessary. Other options include anticoagulation therapy, local or systemic thrombolysis, interventional or surgical thrombectomy.
CONCLUSION
Acute mesenteric venous thrombosis is a complex situation that calls for a multidisciplinary team approach between the surgical and internal medicine departments to determine the best course of action for each patient, as there are major risks involved with each alternative. If peritonism is present, it is preferable to assess and resuscitate as soon as possible and to proceed with surgery.
PubMed: 38875832
DOI: 10.1016/j.ijscr.2024.109872 -
JMIR AI May 2024Large language models (LLMs) have the potential to support promising new applications in health informatics. However, practical data on sample size considerations for...
BACKGROUND
Large language models (LLMs) have the potential to support promising new applications in health informatics. However, practical data on sample size considerations for fine-tuning LLMs to perform specific tasks in biomedical and health policy contexts are lacking.
OBJECTIVE
This study aims to evaluate sample size and sample selection techniques for fine-tuning LLMs to support improved named entity recognition (NER) for a custom data set of conflicts of interest disclosure statements.
METHODS
A random sample of 200 disclosure statements was prepared for annotation. All "PERSON" and "ORG" entities were identified by each of the 2 raters, and once appropriate agreement was established, the annotators independently annotated an additional 290 disclosure statements. From the 490 annotated documents, 2500 stratified random samples in different size ranges were drawn. The 2500 training set subsamples were used to fine-tune a selection of language models across 2 model architectures (Bidirectional Encoder Representations from Transformers [BERT] and Generative Pre-trained Transformer [GPT]) for improved NER, and multiple regression was used to assess the relationship between sample size (sentences), entity density (entities per sentence [EPS]), and trained model performance (F-score). Additionally, single-predictor threshold regression models were used to evaluate the possibility of diminishing marginal returns from increased sample size or entity density.
RESULTS
Fine-tuned models ranged in topline NER performance from F-score=0.79 to F-score=0.96 across architectures. Two-predictor multiple linear regression models were statistically significant with multiple R ranging from 0.6057 to 0.7896 (all P<.001). EPS and the number of sentences were significant predictors of F-scores in all cases ( P<.001), except for the GPT-2_large model, where EPS was not a significant predictor (P=.184). Model thresholds indicate points of diminishing marginal return from increased training data set sample size measured by the number of sentences, with point estimates ranging from 439 sentences for RoBERTa_large to 527 sentences for GPT-2_large. Likewise, the threshold regression models indicate a diminishing marginal return for EPS with point estimates between 1.36 and 1.38.
CONCLUSIONS
Relatively modest sample sizes can be used to fine-tune LLMs for NER tasks applied to biomedical text, and training data entity density should representatively approximate entity density in production data. Training data quality and a model architecture's intended use (text generation vs text processing or classification) may be as, or more, important as training data volume and model parameter size.
PubMed: 38875593
DOI: 10.2196/52095 -
JMIR AI Jun 2023With the growing volume and complexity of laboratory repositories, it has become tedious to parse unstructured data into structured and tabulated formats for secondary...
BACKGROUND
With the growing volume and complexity of laboratory repositories, it has become tedious to parse unstructured data into structured and tabulated formats for secondary uses such as decision support, quality assurance, and outcome analysis. However, advances in natural language processing (NLP) approaches have enabled efficient and automated extraction of clinically meaningful medical concepts from unstructured reports.
OBJECTIVE
In this study, we aimed to determine the feasibility of using the NLP model for information extraction as an alternative approach to a time-consuming and operationally resource-intensive handcrafted rule-based tool. Therefore, we sought to develop and evaluate a deep learning-based NLP model to derive knowledge and extract information from text-based laboratory reports sourced from a provincial laboratory repository system.
METHODS
The NLP model, a hierarchical multilabel classifier, was trained on a corpus of laboratory reports covering testing for 14 different respiratory viruses and viral subtypes. The corpus includes 87,500 unique laboratory reports annotated by 8 subject matter experts (SMEs). The classification task involved assigning the laboratory reports to labels at 2 levels: 24 fine-grained labels in level 1 and 6 coarse-grained labels in level 2. A "label" also refers to the status of a specific virus or strain being tested or detected (eg, influenza A is detected). The model's performance stability and variation were analyzed across all labels in the classification task. Additionally, the model's generalizability was evaluated internally and externally on various test sets.
RESULTS
Overall, the NLP model performed well on internal, out-of-time (pre-COVID-19), and external (different laboratories) test sets with microaveraged F-scores >94% across all classes. Higher precision and recall scores with less variability were observed for the internal and pre-COVID-19 test sets. As expected, the model's performance varied across categories and virus types due to the imbalanced nature of the corpus and sample sizes per class. There were intrinsically fewer classes of viruses being detected than those tested; therefore, the model's performance (lowest F-score of 57%) was noticeably lower in the detected cases.
CONCLUSIONS
We demonstrated that deep learning-based NLP models are promising solutions for information extraction from text-based laboratory reports. These approaches enable scalable, timely, and practical access to high-quality and encoded laboratory data if integrated into laboratory information system repositories.
PubMed: 38875570
DOI: 10.2196/44835 -
PloS One 2024Anti-vascular endothelial growth factor (VEGF) monoclonal antibodies (mAbs) are widely used for tumor treatment, including metastatic colorectal cancer (mCRC). So far,...
BACKGROUND
Anti-vascular endothelial growth factor (VEGF) monoclonal antibodies (mAbs) are widely used for tumor treatment, including metastatic colorectal cancer (mCRC). So far, there are no biomarkers that reliably predict resistance to anti-VEGF mAbs like bevacizumab. A biomarker-guided strategy for early and accurate assessment of resistance could avoid the use of non-effective treatment and improve patient outcomes. We hypothesized that repeated analysis of multiple cytokines and angiogenic growth factors (CAFs) before and during treatment using machine learning could provide an accurate and earlier, i.e., 100 days before conventional radiologic staging, prediction of resistance to first-line mCRC treatment with FOLFOX plus bevacizumab.
PATIENTS AND METHODS
15 German and Austrian centers prospectively recruited 50 mCRC patients receiving FOLFOX plus bevacizumab as first-line treatment. Plasma samples were collected every two weeks until radiologic progression (RECIST 1.1) as determined by CT scans performed every 2 months. 102 pre-selected CAFs were centrally analyzed using a cytokine multiplex assay (Luminex, Myriad RBM).
RESULTS
Using random forests, we developed a predictive machine learning model that discriminated between the situations of "no progress within 100 days before radiological progress" and "progress within 100 days before radiological progress". We could further identify a combination of ten out of the 102 CAF markers, which fulfilled this task with 78.2% accuracy, 71.8% sensitivity, and 82.5% specificity.
CONCLUSIONS
We identified a CAF marker combination that indicates treatment resistance to FOLFOX plus bevacizumab in patients with mCRC within 100 days prior to radiologic progress.
Topics: Humans; Colorectal Neoplasms; Bevacizumab; Leucovorin; Antineoplastic Combined Chemotherapy Protocols; Female; Organoplatinum Compounds; Male; Fluorouracil; Middle Aged; Aged; Drug Resistance, Neoplasm; Prospective Studies; Adult; Neoplasm Metastasis; Biomarkers, Tumor
PubMed: 38875244
DOI: 10.1371/journal.pone.0304324 -
Qualitative and quantitative analysis of lipid droplets in mature 3T3-L1 adipocytes using oil red O.STAR Protocols Jun 2024By differentiating into mature adipocytes, 3T3-L1 cells can be utilized as a model cell line to investigate (pre)adipocyte function in vitro. Here, we present a...
By differentiating into mature adipocytes, 3T3-L1 cells can be utilized as a model cell line to investigate (pre)adipocyte function in vitro. Here, we present a protocol for combining qualitative and quantitative analysis of lipid droplets in mature 3T3-L1 adipocytes using oil red O. We describe steps to differentiate 3T3-L1 preadipocytes to adipocytes and give detailed procedures to determine total lipid amount as well as lipid droplet size and number using microscopic devices and an ImageJ macro. For complete details on the use and execution of this protocol, please refer to Kaczmarek et al..
PubMed: 38875117
DOI: 10.1016/j.xpro.2024.102977 -
JMIR Research Protocols Jun 2024The Veteran-Directed Care (VDC) program serves to assist veterans at risk of long-term institutional care to remain at home by providing funding to hire veteran-selected...
BACKGROUND
The Veteran-Directed Care (VDC) program serves to assist veterans at risk of long-term institutional care to remain at home by providing funding to hire veteran-selected caregivers. VDC is operated through partnerships between Department of Veterans Affairs (VA) Medical Centers (VAMCs) and third-party Aging and Disability Network Agency providers.
OBJECTIVE
We aim to identify facilitators, barriers, and adaptations in VDC implementation across 7 VAMCs in 1 region: Veterans Integrated Service Network (VISN) 8, which covers Florida, South Georgia, Puerto Rico, and the US Virgin Islands. We also attempted to understand leadership and stakeholder perspectives on VDC programs' reach and implementation and identify veterans served by VISN 8's VDC programs and describe their home- and community-based service use. Finally, we want to compare veterans served by VDC programs in VISN 8 to the veterans served in VDC programs across the VA. This information is intended to be used to identify strategies and propose recommendations to guide VDC program expansion in VISN 8.
METHODS
The mixed methods study design encompasses electronically delivered surveys, semistructured interviews, and administrative data. It is guided by the Consolidated Framework for Implementation Research (CFIR version 2.0). Participants included the staff of VAMCs and partnering aging and disability network agencies across VISN 8, leadership at these VAMCs and VISN 8, veterans enrolled in VDC, and veterans who declined VDC enrollment and their caregivers. We interviewed selected VAMC site leaders in social work, Geriatrics and Extended Care, and the Caregiver Support Program. Each interviewee will be asked to complete a preinterview survey that includes information about their personal characteristics, experiences with the VDC program, and perceptions of program aspects according to the CFIR (version 2.0) framework. Participants will complete a semistructured interview that covers constructs relevant to the respondent and facilitators, barriers, and adaptations in VDC implementation at their site.
RESULTS
We will calculate descriptive statistics including means, SDs, and percentages for survey responses. Facilitators, barriers, number of patients enrolled, and staffing will also be presented. Interviews will be analyzed using rapid qualitative techniques guided by CFIR domains and constructs. Findings from VISN 8 will be collated to identify strategies for VDC expansion. We will use administrative data to describe veterans served by the programs in VISN 8.
CONCLUSIONS
The VA has prioritized VDC rollout nationwide and this study will inform these expansion efforts. The findings from this study will provide information about the experiences of the staff, leadership, veterans, and caregivers in the VDC program and identify program facilitators and barriers. These results may be used to improve program delivery, facilitate growth within VISN 8, and inform new program establishment at other sites nationwide as the VDC program expands.
INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)
DERR1-10.2196/57341.
Topics: Humans; United States; United States Department of Veterans Affairs; Veterans; Self Care; Program Evaluation; Caregivers
PubMed: 38875003
DOI: 10.2196/57341 -
European Journal of Sport Science Jun 2024The aim of this study was to investigate sleep-wake behavior and gain insights into perceived impairment (sleep, fatigue, and cognitive function) of athletes competing...
The aim of this study was to investigate sleep-wake behavior and gain insights into perceived impairment (sleep, fatigue, and cognitive function) of athletes competing in two international multi-day adventure races. Twenty-four athletes took part across two independent adventure races: Queensland, Australia and Alaska, USA. Individual sleep periods were determined via actigraphy, and racers self-reported their perceived sleep disturbances, sleep impairment, fatigue and cognitive function. Each of these indices was calculated for pre-, during- and post-race periods. Sleep was severely restricted during the race period compared to pre-race (Queensland, 7:46 [0:29] vs. 2:50 [1:01]; Alaska, 7:39 [0:58] vs. 2:45 [2:05]; mean [SD], hh:mm). As a result, there was a large cumulative sleep debt at race completion, which was not 'reversed' in the post-race period (up to 1 week). The deterioration in all four self-reported scales of perceived impairment during the race period was largely restored in the post-race period. This is the first study to document objective sleep-wake behaviors and subjective impairment of adventure racers, in the context of two geographically diverse, multi-day, international adventure races. Measures of sleep deprivation indicate that sleep debt was extreme and did not recover to pre-race levels within 1 week following each race. Despite this objective debt continuing, perceived impairment returned to pre-race levels quickly post-race. Therefore, further examination of actual and perceived sleep recovery is warranted. Adventure racing presents a unique scenario to examine sleep, performance and recovery.
PubMed: 38874812
DOI: 10.1002/ejsc.12143 -
Acta Orthopaedica Jun 2024There is conflicting evidence regarding treatment outcomes after minimally invasive sacroiliac joint fusion for long-lasting severe sacroiliac joint pain. The primary...
BACKGROUND AND PURPOSE
There is conflicting evidence regarding treatment outcomes after minimally invasive sacroiliac joint fusion for long-lasting severe sacroiliac joint pain. The primary aim of our cohort study was to investigate change in patient-reported outcome measures (PROMs) after minimally invasive sacroiliac joint surgery in daily practice in the Swedish Spine Registry. Secondary aims were to explore the proportion of patients reaching a patient acceptable symptom score (PASS) and the minimal clinically important difference (MCID) for pain scores, physical function, and health-related quality of life outcomes; furthermore, to evaluate self-reported satisfaction, walking distance, and changes in proportions of patients on full sick leave/disability leave and report complications and reoperations.
METHODS
Data from the Swedish Spine Registry was collected for patients with first-time sacroiliac joint fusion, aged 21 to 70 years, with PROMs available preoperatively, at 1 or 2 years after last surgery. PROMs included Oswestry Disability Index (ODI), Numeric Rating Scale (NRS) for low back pain (LBP) and leg pain, and EQ-VAS, in addition to demographic variables. We calculated mean change from pre- to postoperative and the proportion of patients achieving MCID and PASS.
RESULTS
68 patients had available pre- and postoperative data, with a mean age of 45 years (range 25-70) and 59 (87%) were female. At follow-up the mean reduction was 2.3 NRS points (95% confidence interval [CI] 1.6-2.9; P < 0.001) for LBP and 14.8 points (CI 10.6-18.9; P < 0.001) for ODI. EQ-VAS improved by 22 points (CI 15.4-30.3, P < 0.001) at follow-up. Approximately half of the patients achieved MCID and PASS for pain (MCID NRS LBP: 38/65 [59%] and PASS NRS LBP: 32/66 [49%]) and physical function (MCID ODI: 27/67 [40%] and PASS ODI: 24/67 [36%]). The odds for increasing the patient's walking distance to over 1 km at follow-up were 3.5 (CI 1.8-7.0; P < 0.0001), and of getting off full sick leave or full disability leave was 0.57 (CI 0.4-0.8; P = 0.001). In the first 3 months after surgery 3 complications were reported, and in the follow-up period 2 reoperations.
CONCLUSION
We found moderate treatment outcomes after minimally invasive sacroiliac joint fusion when applied in daily practice with moderate pain relief and small improvements in physical function.
Topics: Humans; Patient Reported Outcome Measures; Middle Aged; Sweden; Female; Male; Registries; Adult; Sacroiliac Joint; Minimally Invasive Surgical Procedures; Aged; Cohort Studies; Spinal Fusion; Pain Measurement; Low Back Pain; Disability Evaluation; Quality of Life; Patient Satisfaction; Young Adult; Minimal Clinically Important Difference; Treatment Outcome
PubMed: 38874434
DOI: 10.2340/17453674.2024.40817 -
Turkish Neurosurgery May 2023Background: Accurate preoperative grading and isocitrate dehydrogenase (IDH) identification is highly important for proper treatment planning and prognosis evaluation in...
AIM
Background: Accurate preoperative grading and isocitrate dehydrogenase (IDH) identification is highly important for proper treatment planning and prognosis evaluation in glioma patients. Purpose To explore the applicability of histogram features from non-invasive arterial spin labeling (ASL)-weighted MRI in differentiating isocitrate dehydrogenase mutant (IDH-mut) and wild type (IDH-wt), and separating lower-grade gliomas (LGGs) from glioblastoma multiforme (GBM).
MATERIAL AND METHODS
Methods One hundred thirty-one patients scanned with ASL-weighted and anatomic MRI were retrospectively included. Cerebral blood flow (CBF) maps were calculated, from which 10 histogram features describing the CBF distribution were extracted within the tumor region. Correlation analysis was used to determine correlations between histogram features, and tumor grades and IDH genotypes. Independent t-tests and Fisher\'s exact tests were used to determine the differences in extracted histogram features, age at diagnosis, and gender among different glioma subtypes. Binary logistic regression was used to combine multivariates, and the diagnostic performances were evaluated with receiver operating characteristic curves.
RESULTS
Results CBF histogram features were significantly correlated with tumor grades and IDH genotypes, and facilitate the efficacious differentiation of LGGs from GBM, and IDH-mut from IDH-wt gliomas. A model combining the CBF 30th percentile and age at diagnosis resulted in an area under the receiving operating characteristic curve (AUC) of 0.73 in judging LGGs from GBM. Integrating age at the time of diagnosis and CBF 10th percentile allows more comprehensive differentiation of IDH-mut and IDH-wt gliomas, and the combined model achieved an improved AUC of 0.856 (sensitivity, 84.4%; and specificity, 82.9%).
CONCLUSION
Conclusion Histogram features from non-invasive ASL-weighted MRI were significantly correlated with tumor grade and IDH genotypes, and facilitated efficacious differentiation of glioma subtypes. Combining age at the time of diagnosis and perfusion histogram features resulted in a more comprehensive identification of tumor subtypes, which indicated that ASL-weighted MRI can serve as non-invasive tool for the pre-surgical evaluation of gliomas.
PubMed: 38874235
DOI: 10.5137/1019-5149.JTN.42484-22.3