Frequent Classes Research Articles

P0365 The evolution of Geldof fistula class in patients with Crohn’s disease perianal fistula

Abstract Background Progressing the evaluation and management of Perianal fistulizing Crohn’s disease ( CD-pAF ) , an aggressive phenotype of CD , is one of the key unmet needs in Inflammatory Bowel Disease (IBD). There is sub-optimal understanding of the behavior of CD-pAF with some patients having rapid velocity of disease progression needing radical surgery such as proctectomy or diverting stoma. More recently, a practical patient-centric classification has been developed recently by Geldof et al has been used to classify patients underpinned by the behavior of CD-pAF(1). We aimed to evaluate the fistula class trajectory using Geldof classification in a large cohort of CD-pAF patients. Methods Data for 997 patients with Crohn’s disease perianal fistula (CD-pAF) included in a cohort study of patients with existing diagnosis of CD-pAF. Data was entered to predefined time points starting at the date of diagnosis of Cd-pAF and onward until entry into the cohort. We analysed the change in fistula classification using Geldof classification over 2 years. Results Data at entry to the cohort of 997 ( Males 52% , Females 48%) patients were included in the analysis The median age at fistula diagnosis was 31 years (23, 42). Among the 997 patients 686 had their Geldof class identified at baseline and 665 at 6-12 months. Class 2a followed by class 1 were the most frequent classes reported (Class 2a at baseline 376/686(55%) and Class 1 167/686(24%), at 6-12 months class 2a were 286/665(43%) and class 1 were 252/665(38%). At 12-24 months from fistula diagnosis 599 patients’ classes were reported, the most frequent being class 1 279/599(47%). There were minimal variations in Class 2b and class 2ci over the 3 time periods. (Class 2b [112/686(16%], [92/665(14%)]and [82/549(15%)] respectively, class 2ci [21/686(3%)], [20/665(3%)] [14/549(3%)] respectively). 15/686(2.1%) had proctectomy in the first 2 years. Conclusion While there are dynamic changes in fistula classification over 2 year period from fistula diagnosis there seems to be clustering among some classes. This indicates potential opportunity for identifying distinct classes which may have refractory course and tailor therapies accordingly. References (1)Geldof J, Iqbal N, LeBlanc JF, et al. Classifying perianal fistulising Crohn’s disease: an expert consensus to guide decision-making in daily practice and clinical trials. Lancet Gastroenterol Hepatol. 2022;7(6):576-584. doi:10.1016/S2468-1253(22)00007-3

Proto-DS: A Self-Supervised Learning-Based Nondestructive Testing Approach for Food Adulteration with Imbalanced Hyperspectral Data.

Conventional food fraud detection using hyperspectral imaging (HSI) relies on the discriminative power of machine learning. However, these approaches often assume a balanced class distribution in an ideal laboratory environment, which is impractical in real-world scenarios with diverse label distributions. This results in suboptimal performance when less frequent classes are overshadowed by the majority class during training. Thus, the critical research challenge emerges of how to develop an effective classifier on a small-scale imbalanced dataset without significant bias from the dominant class. In this paper, we propose a novel nondestructive detection approach, which we call the Dice Loss Improved Self-Supervised Learning-Based Prototypical Network (Proto-DS), designed to address this imbalanced learning challenge. The proposed amalgamation mitigates the label bias on the most frequent class, further improving robustness. We validate our proposed method on three collected hyperspectral food image datasets with varying degrees of data imbalance: Citri Reticulatae Pericarpium (Chenpi), Chinese herbs, and coffee beans. Comparisons with state-of-the-art imbalanced learning techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and class-importance reweighting, reveal our method's superiority. Notably, our experiments demonstrate that Proto-DS consistently outperforms conventional approaches, achieving the best average balanced accuracy of 88.18% across various training sample sizes, whereas the Logistic Model Tree (LMT), Multi-Layer Perceptron (MLP), and Convolutional Neural Network (CNN) approaches attain only 59.42%, 60.38%, and 66.34%, respectively. Overall, self-supervised learning is key to improving imbalanced learning performance and outperforms related approaches, while both prototypical networks and the Dice loss can further enhance classification performance. Intriguingly, self-supervised learning can provide complementary information to existing imbalanced learning approaches. Combining these approaches may serve as a potential solution for building effective models with limited training data.

Is the Foot a Prosodic Domain in European Portuguese?

It is widely accepted that languages organize speech material into prosodic domains, which are hierarchically arranged. However, it is still a matter of debate whether this prosodic hierarchy is composed of a small number of universal categories, or whether these prosodic categories are language-particular and emergent. In this article, we concentrate on one of these categories, the foot, and we investigate its role in European Portuguese (EP). Whereas research on EP has shown that other prosodic domains commonly found crosslinguistically are active in the language, it seems that EP may lack this prosodic constituent. Therefore, the goal of this article is to systematically investigate the role of the foot in a number of areas within EP grammar. In our study, we defend some new approaches to several long-standing issues in EP phonology and we conclude that many facts of the language can be, in fact, better understood resorting to the foot domain. Namely, the distribution of stress in the most frequent morphological classes (thematic non-verbs and present tense verbs), -inh-/-zinh- diminutive formation, the domain for regular vowel reduction, obligatory glide formation, the stress window, and poetic rhyme seem to benefit from a metrically-governed account. By contrast, other facts, such as vowel lowering, word clipping, minimal words, secondary stress, and schwa deletion are shown not to be conditioned by the foot in EP. Importantly, the evidence we found for metrical structure only cues the prominent foot of the word, suggesting that pretonic material may not be footed. All in all, in addition to shedding light on facts previously poorly understood, and exposing some noteworthy specificities of EP in the realm of Romance languages and other varieties of Portuguese, we draw implications for the universal characterization of the prosodic hierarchy and its acquisition.

Investigating the resistome of haemolytic bacteria in Arctic soils.

Microorganisms inhabiting hostile Arctic environments express a variety of functional phenotypes, some of clinical interest, such as haemolytic ability and antimicrobial resistance. We studied haemolytic bacterial isolates from Arctic habitats, assessing their minimum inhibitory concentration (MIC) against antimicrobials. We then performed whole genome sequencing and analysed them for features conferring antimicrobial resistance. MIC data showed that Micromonospora spp. belong to 33% non-wild type (NWT) for erythromycin and penicillin and 22% NWT for tetracycline. Both Pseudomonas spp. belong to 43% NWT for nalidixic acid and streptomycin and 29% NWT for colistin. Finally, the Pedobacter isolate was in 80% NWT for antimicrobials tested. Whole-genome sequencing analyses revealed that fluoroquinolones, tetracyclines, macrolides and penams were the most frequent drug classes against which genotypic resistance was found. Additionally, resistance genes to heavy metals and disinfectants were identified. Our research demonstrates the presence of antimicrobial resistance in bacteria from Arctic habitats and highlights the importance of conservation efforts in these environments, where anthropogenic influence is becoming more evident. Furthermore, our data suggest the possible presence of novel resistance mechanisms, which could pose a threat if the responsible genes are transferable between species or become widespread due to environmental stress and alterations brought about by climate change.

The medication usage pattern and prevalence of polypharmacy among patients with sickle cell disease: a population-based study in southern Iran

ABSTRACT Background Due to the numerous complications associated with sickle cell disease (SCD), patients often receive a variety of medications alongside their SCD treatment. However, a notable gap exists in the current literature regarding medication use patterns among them. This study aimed to investigate medication usage patterns in patients with SCD. Research design and methods This cross-sectional study, conducted in Bushehr Province, employed a stratified random sampling method to select eligible participants with SCD. A thorough interview gathered various information, including details about the medications. The Anatomical Therapeutic Chemical classification system was utilized for drug classification. Polypharmacy was defined as the concurrent use of at least five medications. Results A total of 300 individuals with SCD were included in this study. Polypharmacy was observed in 26.3% (95% CI: 20.8%-32.8%) of the study population. The analyses revealed positive associations between the use of more concurrent medication use and higher age groups and having multimorbidity. Antianemic preparations (86.7%), antineoplastic agents (58.3%), and vitamins (41.0%) were the most frequent medication classes used by the study participants. Conclusions Our study revealed notable underutilization of hydroxyurea and a high rate of polypharmacy, associated with age and multimorbidity, among patients with SCD in southern Iran.

Real-time path planning for autonomous vehicle off-road driving.

Autonomous driving is a growing research area that brings benefits in science, economy, and society. Although there are several studies in this area, currently there is no a fully autonomous vehicle, particularly, for off-road navigation. Autonomous vehicle (AV) navigation is a complex process based on application of multiple technologies and algorithms for data acquisition, management and understanding. Particularly, a self-driving assistance system supports key functionalities such as sensing and terrain perception, real time vehicle mapping and localization, path prediction and actuation, communication and safety measures, among others. In this work, an original approach for vehicle autonomous driving in off-road environments that combines semantic segmentation of video frames and subsequent real-time route planning is proposed. To check the relevance of the proposal, a modular framework for assistive driving in off-road scenarios oriented to resource-constrained devices has been designed. In the scene perception module, a deep neural network is used to segment Red-Green-Blue (RGB) images obtained from camera. The second traversability module fuses Light Detection And Ranging (LiDAR) point clouds with the results of segmentation to create a binary occupancy grid map to provide scene understanding during autonomous navigation. Finally, the last module, based on the Rapidly-exploring Random Tree (RRT) algorithm, predicts a path. The Freiburg Forest Dataset (FFD) and RELLIS-3D dataset were used to assess the performance of the proposed approach. The theoretical contributions of this article consist of the original approach for image semantic segmentation fitted to off-road driving scenarios, as well as adapting the shortest route searching A* and RRT algorithms to AV path planning. The reported results are very promising and show several advantages compared to previously reported solutions. The segmentation precision achieves 85.9% for FFD and 79.5% for RELLIS-3D including the most frequent semantic classes. While compared to other approaches, the proposed approach is faster regarding computational time for path planning.

Pharmacist initiated interventions using RxChange message communication with prescribers for electronic prescriptions: A retrospective descriptive study

BackgroundRxChange messages improve patient medication management by enhancing pharmacist-prescriber communication, but their usage patterns in the United States are not well-documented. ObjectiveTo determine intervention characteristics by pharmacists and prescribers using RxChange messages. MethodsA retrospective analysis of electronic prescription and RxChange messages from 2022 to 2023, using data from Surescripts, LLC, was conducted. This included NewRx messages and RxChange Responses, categorized by 7 RxChange use cases and Anatomical Therapeutic Chemical level 4 medication classes. Descriptive statistics and nonparametric tests were used for statistical analysis. ResultsThe study analyzed 1,361,528 RxChange messages. Therapeutic interchange was the predominant use case (76.14%). Direct approvals accounted for 10.44% of requests, approvals with changes for 42.55%, and denials for 47.01%. Script clarification had the highest approval rate (64.21%), while prior authorization faced the most frequent denials (73.38%). The top denial reason was "Request addressed through alternate methods such as phone or fax (41.50%).” The most frequent drug classes observed in the data were selective beta-2 adrenoreceptor agonists, extended-spectrum penicillins, selective serotonin reuptake inhibitors, and glucagon-like peptide 1 analogues. Time from new e-prescription issuance to RxChange request submission was longer than from request to response, with a significant statistical difference (median 1.57 vs 0.27 days, P-value < 0.05). ConclusionThis study highlights interventions pharmacists make using RxChange with electronic prescriptions to improve patient care and medication safety. It underlined the need for improved RxChange message content and data on the effectiveness of RxChange messages in improving medication use.

Frequency and preventability of adverse drug events in the outpatient setting

BackgroundLimited data exist regarding adverse drug events (ADEs) in the outpatient setting. The objective of this study was to determine the incidence, severity, and preventability of ADEs in the outpatient...

Cancer fusion transcripts with human non-coding RNAs.

Cancer chimeric, or fusion, transcripts are thought to most frequently appear due to chromosomal aberrations that combine moieties of unrelated normal genes. When being expressed, this results in chimeric RNAs having upstream and downstream parts relatively to the breakpoint position for the 5'- and 3'-fusion components, respectively. As many other types of cancer mutations, fusion genes can be of either driver or passenger type. The driver fusions may have pivotal roles in malignisation by regulating survival, growth, and proliferation of tumor cells, whereas the passenger fusions most likely have no specific function in cancer. The majority of research on fusion gene formation events is concentrated on identifying fusion proteins through chimeric transcripts. However, contemporary studies evidence that fusion events involving non-coding RNA (ncRNA) genes may also have strong oncogenic potential. In this review we highlight most frequent classes of ncRNAs fusions and summarize current understanding of their functional roles. In many cases, cancer ncRNA fusion can result in altered concentration of the non-coding RNA itself, or it can promote protein expression from the protein-coding fusion moiety. Differential splicing, in turn, can enrich the repertoire of cancer chimeric transcripts, e.g. as observed for the fusions of circular RNAs and long non-coding RNAs. These and other ncRNA fusions are being increasingly recognized as cancer biomarkers and even potential therapeutic targets. Finally, we discuss the use of ncRNA fusion genes in the context of cancer detection and therapy.

Optimizing Emotional Insight through Unimodal and Multimodal Long Short-term Memory Models

The field of multimodal emotion recognition is increasingly gaining popularity as a research area. It involves analyzing human emotions across multiple modalities, such as acoustic, visual, and language. Emotion recognition is more effective as a multimodal learning task than relying on a single modality. In this paper, we present an unimodal and multimodal long short-term memory model with a class weight parameter technique for emotion recognition on the CMU-Multimodal Opinion Sentiment and Emotion Intensity dataset. In addition, a critical challenge lies in selecting the most effective fusion method for integrating multiple modalities. To address this, we applied four different fusion techniques: Early fusion, late fusion, deep fusion, and tensor fusion. These fusion methods improved the performance of multimodal emotion recognition compared to unimodal approaches. With the highly imbalanced number of samples per emotion class in the MOSEI dataset, adding a class weight parameter technique leads our model to outperform the state of the art on all three modalities — acoustic, visual, and language — as well as on all the fusion models. The challenges of class imbalance, which can lead to biased model performance, and using an effective fusion method for integrating multiple modalities often result in decreased accuracy in recognizing less frequent emotion classes. Our proposed model shows 2–3% performance improvement in the unimodal and 2% in the multimodal over the state-of-the-art achieved results.

Secretory structures in Baccharis platypoda DC. inflorescences (Asteraceae) and characterization of the chemical composition of its secretion.

Rocky outcrop environments at high altitudes have nutrient-poor soil, where species are exposed to water scarcity and high solar radiation. Baccharis platypoda DC. occurs in such an environment and has a rigid and transparent secretion that covers the entire inflorescence. We analysed and compared the secretory structures and their chemical composition in female and male inflorescences of B. platypoda, a dioecious species, to explore chemodiversity within this species and assess potential differences between individuals. Our investigation also aims to understand the occurrence of these substances in the genus Baccharis L. Chemical compounds and secretory structures were similar in female and male inflorescences. There are glandular trichomes on the epidermis of the abaxial surface of bracts, and secretory ducts in the axis of the inflorescence, as well as in sepals, petals, and bracts. Histochemical tests were positive for phenolic compounds, flavonoids, proteins, pectin, and lipids, but not for mucilage. Flavonoid content varied between 6.24% and 9.81%, being higher in female inflorescences. Chromatography revealed the presence of several phenolic compounds, some terpenes, and other less frequent classes in both female and male inflorescences. We highlight that trichomes found on these surfaces produce abundant phenolic compounds. These act as natural defence agents, absorbing UV radiation and minimizing oxidative stress to plant cells. The chemical composition of the secretion covering the inflorescences may reflect adaptation and survival mechanisms of these organisms under extreme sun exposure.

#1602 Lupus nephritis multidisciplinary consultation: a new strategy for increasingly complex patients

Abstract Background and Aims Lupus nephritis (LN) is a major course of morbidity and mortality in patients with systemic lupus erythematosus (SLE), best managed by a multidisciplinary group. Recent management advances require greater collaboration between specialists and individualized treatments in severe cases. Method We conducted a retrospective, single-center study of a cohort of incident patients which were diagnosed with LN between 2015 and 2022 and began follow-up in the multidisciplinary rheumatology and nephrology consultation of our institution. Clinical and analytical characteristics at diagnosis and during follow-up were studied. The primary endpoint was complete remission rate at follow-up end, according to GLOSEN (Grupo de Estudio de Enfermedades Glomerulares de la SEN) [1] criteria. Changes in immunosuppression schemes over time were studied. Results 26 patients (92.3% women) were included. Median age at SLE diagnosis was 30.5 years [Interquartile range (IQR) 23.7 – 44.2]. Most patients were Hispanic (14 patients, 53.8%). Median time between SLE and LN diagnosis was 21 months (IQR 0-129 months). In 9 patients (34.6%), SLE and LN onset happened simultaneously. Most common renal manifestation was hematuria, which appeared in 19 patients (73.1%). Median proteinuria was 1.83 g/d (IQR 0.88-2.92) and serum creatinine level, 0.69 (IQR 0.51-1.18). Hypocomplementemia (88.5%) and positivity to anti-DNA (69.2%) were common. Most frequent classes found in renal biopsies were IV (30.8%) and V (23.1%). At LN debut, SLEDAI score, which assesses SLE manifestations, had a median value of 16 (IQR 12-20). Mycophenolate mofetil (96.2%) was the main induction agent. Intravenous steroids were administered in 14 patients (53.8%) and the rest received steroids at a dose of 0.5 mg/kg/day. Addition of another therapy, mainly tacrolimus (6 patients, 23.1%) and belimumab (4 patients, 15.4%), was needed to achieve remission in 12 patients (46.2%). In 8 of them (66.6%), treatments were added within the first 6 months. These 12 patients had less chronicity in the biopsy (median chronicity index: 0 vs. 2; p 0.004) and had higher proteinuria (median 1.41 g/day vs. 0.4 g/day; p 0.001). Likewise, SLEDAI score at 3 months was significantly higher in these 12 patients than in the standard of care group (median SLEDAI score 13 vs. 8; p 0.016). No differences were found in 3 month serum creatinine (p 0.181) or 3 month hematuria (p 0.496) After a median follow-up of 47 months, 15 (57.7%) and 9 patients (34.6%) achieved complete and partial remission, respectively. No differences in percentages of remission were observed between the 2 groups. Steroids were withdrawn in 12 patients (46.2%). Median steroid dose at the end of follow-up was 5 mg (IQR 3-5). Conclusion In our LN cohort, higher SLEDAI and proteinuria indicated the need for an individualized approach. Collaboration between Rheumatology and Nephrology specialists is essential to identify patients with these needs.

Deep contrastive learning based hybrid network for Typhoon intensity classification

The accurate classification of Typhoon Intensity (TI) based on cloud patterns in satellite images is crucial for effective disaster warning and management. However, this task is particularly challenging due to the striking visual similarities among different sub-classes of typhoons, coupled with a long-tailed distribution of class occurrences. Existing deep learning methods often struggle with biased classification stemming from imbalanced datasets and face difficulty discerning subtle differences between categories. This paper proposes a novel solution to these challenges through a hybrid framework that integrates contrastive learning and classifier learning. Contrastive learning is employed to increase the separation of similar sub-classes within the feature space, while classifier learning aims to train a discriminative and unbiased classifier. The proposed approach is evaluated extensively on the publicly available DeepTI dataset, demonstrating enhanced performance for both prominent and less frequent classes. The model achieves the 70.88% accuracy for seven categories typhoon intensity classification. The code implementation is at https://github.com/chen-huanxin/contrastive-hybrid-network.

The remodelling of u-stem adjectives in West Germanic

In Proto-Germanic, strong adjectives are declined according to stem type. The minor u­-stem class of adjectives is only preserved in Gothic, and even there it is under pressure from more frequent classes. In West Germanic, members of this class are redistributed into the common a-/ō-stem and ja-/jō-stem classes. The present study analysed the reflexes of all Proto-Germanic u-stem adjectives in Kroonen (2013) in the different West Germanic languages. It was found that a plurality are transferred into the ja-/jō-stem class in all languages. However, a considerable number of them have a-/ō-stem reflexes in some languages but ja-/jō-stem ones in others, while certain adjectives can show reflexes of both types within one language. Anglo-Frisian has the largest number of a-/ō-stem outcomes, and High German the most ja-/jō-stem ones, with Dutch and Low German falling somewhere in between. Therefore, the remodelling of u-stem adjectives necessarily postdates a period of West Germanic unity.

Analysis of the hyperparameter optimisation of four machine learning satellite imagery classification methods

The classification of land use and land cover (LULC) from remotely sensed imagery in semi-arid Mediterranean areas is a challenging task due to the fragmentation of the landscape and the diversity of spatial patterns. Recently, the use of deep learning (DL) for image analysis has increased compared to commonly used machine learning (ML) methods. This paper compares the performance of four algorithms, Random Forest (RF), Support Vector Machine (SVM), Multilayer Perceptron (MLP) and Convolutional Network (CNN), using multi-source data, applying an exhaustive optimisation process of the hyperparameters. The usual approach in the optimisation process of a LULC classification model is to keep the best model in terms of accuracy without analysing the rest of the results. In this study, we have analysed such results, discovering noteworthy patterns in a space defined by the mean and standard deviation of the validation accuracy estimated in a 10-fold cross validation (CV). The point distributions in such a space do not appear to be completely random, but show clusters of points that facilitate the discovery of hyperparameter values that tend to increase the mean accuracy and decrease its standard deviation. RF is not the most accurate model, but it is the less sensitive to changes in hyperparameters. Neural Networks, tend to increase commission and omission errors of the less represented classes because their optimisation lead the model to learn better the most frequent classes. On the other hand, RF and MLP prediction layers are the most accurate from a general qualitative point of view.

Prescribed psychotropic medication patterns among treated Foster Care enrollees: a single institution study.

While several state-based studies have shown that children in foster care are more likely to be prescribed psychotropic medications and experience concomitant medication use both within and among medication class, these patterns have not been explored in the state of Nevada, which lacks state mandated oversight of psychotropic prescribing for foster care enrolled youth. Data from an electronic medical record system from a single institution were analyzed to examine the prevalence of psychotropic prescribing and concomitant medication use in children ages 2 to 19 who were enrolled and received psychotropic prescriptions between July 2019 to June 2022. Out of 569 distinct psychotropic medication treatment episodes within this cohort, the most frequent psychotropic classes prescribed were non-stimulant ADHD medications (alpha-agonists and atomoxetine, 31.5%), atypical antipsychotics (22.1%), antidepressants (20.6%), and stimulants (16.0%). The use of stimulants and non-stimulant ADHD medications decreased in older age groups while the use of antidepressants and antipsychotics increased in older age groups. During the three-year period studied, 24.0% of psychotropic medications prescriptions increased in dosage. Treatments were prescribed for only one month in 43.8% of youth. In children prescribed psychotropic medications, concomitant medication use for at least 60 days occurred in 28.0% of children who had any psychotropic medication prescribed. Within the cohort of 273 foster care enrolled subjects aged 2 to 19 years old who received psychotropic medication prescriptions, non-stimulant ADHD medications (both alpha-agonists and atomoxetine) and atypical antipsychotics were more commonly co-prescribed additional psychotropic medication compared to other co-prescribed medication categories. This study illustrates prescribing patterns in a community mental health clinic focused on judicious prescribing of psychotropic medications in foster care enrolled youth. Despite this, 41% of the youth treated in this clinic received at least one prescription for psychotropic medication, and of those, 27.8% were prescribed more than one psychotropic medication at the same time. More studies are necessary to understand the underlying causes of high prevalence of concomitant medication use and prescribing practices of psychotropic medications use in foster care involved pediatric populations.

Assessment of clinical pharmacist interventions using a web-based application in a Saudi Arabian Tertiary Hospital.

Medication-related problems are a top concern of clinical pharmacists. Medication-related problems can cause patient harm and increase the number of visits, hospital admissions, and length of hospital stay. The objective was to assess clinical pharmacy medication-related problem-related interventions in a tertiary care setting. A retrospective cohort study was conducted at King Fahad Armed Forces Hospital in Jeddah (Saudi Arabia) between June 2021 and June 2022. The data were extracted monthly from a new web-based Microsoft Excel application documenting medication-related problems during any stage of the medication use process. A total of 5310 medication-related problem-related interventions in 1494 patients were performed. The departments associated with the highest frequency of medication-related problem-related interventions were the critical care unit (26.9%), intensive care unit (23.8%), anticoagulation clinic (17.1%), medical ward (11.3%), and nephrology unit (6.8%). The most common type of medication-related problem-related interventions included inappropriate dosage regimens (25.6%), monitoring drug effect or therapeutic drug monitoring (24.4%), requirement of additional drug therapy (21.9%), and inappropriate drug selection (14.1%). The proposed interventions were accepted by physicians in 97% of the incidents. The most frequent medication classes associated with medication-related problem-related interventions were cardiovascular agents (47.6%), antimicrobial agents (27.2%), and nutrition and blood substitute agents (11.4%). The most frequent medication groups associated with medication-related problem-related interventions were anticoagulants (25.6%) and antibiotics (25.2%). The current findings characterize the medication-related problem-related interventions addressed in clinical pharmacy at a tertiary care setting. The high rate of physician acceptance emphasizes the integral patient safety role of clinical pharmacy services.

Drug Therapy Problems Identified by Clinical Pharmacists at a General Surgery Ward of an Academic Referral Hospital in Jordan.

Drug therapy problems (DTPs) continuously occur in hospitalized patients. This study aims to emphasize the role of clinical pharmacists in evaluating the DTP's frequencies, causes, severity ratings, and contributing factors at a general surgery ward in Jordan. This prospective observational study was conducted at one of the major teaching and referral hospitals in Jordan. Data were collected through clinical pharmacist reviews of paper and electronic medical records as well as patient interviews. DTPs were identified using Cipolle's classification system and rated for severity on a scale of 10. Multiple linear regression was performed to identify factors contributing to DTPs. Drug classesprimarily associated with DTPs were specified. During enrollment, a total of 80 patients were recruited in this study. The mean age of the enrolled patients was 52.35 ± 14.82 years, and 49 (61.25%) of them were males. Within the study period, 192 DTPs were identified by clinical pharmacists in 87.5%of the total recruited patients. The mean number of DTPs per patient was 2.40 ± 1.83. The most common categories of DTPs were "needs additional therapy" 46 (23.96%), "unnecessary drug therapy" 45 (23.44%), and "dosage too low" 39 (20.31%). Of the total DTPs, 127 (66.15%) were rated as severe. Multiple linear regression revealed that patients' length of hospital stay and the number of current medications had a statistically significant effect on the number of DTPs identified during hospitalization. Endocrine and metabolic drugs 51 (26.56%) and cardiovascular drugs 36 (18.75%) were the most frequent classes of drugs contributing to DTPs. DTPs are common in the general surgery ward. Clinical pharmacists can provide medication reviews for surgical patients to identify DTPs and rate their severities. Detecting risk factors for DTPs and the most common drug classes associated with them can assist in decision-making relevant to reducing DTPs in the surgical ward.

Drug Utilization Research and Predictors of Outcomes in the Intensive Care Unit of a Tertiary Care Hospital: A Prospective Observational Study.

Multiple drugs are commonly prescribed to intensive care unit (ICU) patients owing to the disease profile, multiple organ dysfunction, prophylaxis, management of stress ulcers, nosocomial infections, etc. This study aimed to evaluate the drug utilization patterns and factors influencing mortality and duration of stay in ICU patients. Methodology: A prospective observational study was conducted in the ICUof our tertiary care hospital, Postgraduate Institute of Medical Sciences, Rohtak. Data was collected from treatment charts of patients using a structured pretested proforma. World Health Organization Anatomical Therapeutic Chemical/Defined Daily Dose (WHO ATC/DDD) methodology and core prescribing indicators were used to assess drug utilization data. The effect of different variables on mortality and duration of stay in the ICU was evaluated using regression analysis. An average of 8.78 drugs were prescribed per patient. Among the 922 prescriptions, anti-infectives, anti-inflammatory drugs, and drugs acting on the gastrointestinal tract were the most frequent medication classes prescribed. Polypharmacy and trade name prescribing were common. For most of the drugs, the prescribed daily dose corresponded to the WHO-DDD except ceftriaxone and levofloxacin. Age, presence of cardiac disorders, and Glasgow Coma Scale (GCS) score at admission directly correlated with mortality while the use of diuretics had a negative correlation with the duration of ICU stay. Conclusions: There is a need to rationalize drug therapy in the ICUwith regard to limiting polypharmacyand emphasizing generic drug name prescribing and adherence to the essential drug list. Antibiotic prescription patterns, in particular, deserve a special focus keeping in mind the multitude of factors demanding aggressive antibiotic use in critically ill intensive care patients.

A Critical Assessment of Generative Models for Synthetic Data Augmentation on Limited Pneumonia X-ray Data.

In medical imaging, deep learning models serve as invaluable tools for expediting diagnoses and aiding specialized medical professionals in making clinical decisions. However, effectively training deep learning models typically necessitates substantial quantities of high-quality data, a resource often lacking in numerous medical imaging scenarios. One way to overcome this deficiency is to artificially generate such images. Therefore, in this comparative study we train five generative models to artificially increase the amount of available data in such a scenario. This synthetic data approach is evaluated on a a downstream classification task, predicting four causes for pneumonia as well as healthy cases on 1082 chest X-ray images. Quantitative and medical assessments show that a Generative Adversarial Network (GAN)-based approach significantly outperforms more recent diffusion-based approaches on this limited dataset with better image quality and pathological plausibility. We show that better image quality surprisingly does not translate to improved classification performance by evaluating five different classification models and varying the amount of additional training data. Class-specific metrics like precision, recall, and F1-score show a substantial improvement by using synthetic images, emphasizing the data rebalancing effect of less frequent classes. However, overall performance does not improve for most models and configurations, except for a DreamBooth approach which shows a +0.52 improvement in overall accuracy. The large variance of performance impact in this study suggests a careful consideration of utilizing generative models for limited data scenarios, especially with an unexpected negative correlation between image quality and downstream classification improvement.