Automated detection of off-label drug use.
Ontology highlight
ABSTRACT:
SUBMITTER: Jung K
PROVIDER: S-EPMC3929699 | biostudies-literature | 2014
REPOSITORIES: biostudies-literature
Publications
Automated detection of off-label drug use.
Jung Kenneth K LePendu Paea P Chen William S WS Iyer Srinivasan V SV Readhead Ben B Dudley Joel T JT Shah Nigam H NH
PloS one 20140219 2
Off-label drug use, defined as use of a drug in a manner that deviates from its approved use defined by the drug's FDA label, is problematic because such uses have not been evaluated for safety and efficacy. Studies estimate that 21% of prescriptions are off-label, and only 27% of those have evidence of safety and efficacy. We describe a data-mining approach for systematically identifying off-label usages using features derived from free text clinical notes and features extracted from two databa ...[more]
Similar Datasets
Project description:Real-world clinical practice commonly veers from formal drug approvals in off-label use, accounting for 21% of prescriptions for common drugs. Due to its ad hoc nature, off-label use typically goes undocumented, evading the safety and efficacy scrutiny of clinical trials. A systematic and automated approach to detection of these uses in the electronic health record (EHR) would enable improved safety monitoring, provide insight into prescribing patterns, and support real-world evidence appraisal. Domain knowledge provided by medication-indication knowledge bases has been shown to improve the accuracy of EHR-based automated detection of off-label use, but remains limited due to diverse concept representations and granularities across data sources. We present a method to leverage hierarchical concept knowledge to align medication-indication knowledge with EHR data for automated detection of off-label drug use in clinical practice. We demonstrate an over two-fold increase in detected off-label diagnoses when leveraging hierarchical knowledge relative to direct concept matching alone.
| S-EPMC10283126 | biostudies-literature
Project description:Plant leaf stomata are the gatekeepers of the atmosphere-plant interface and are essential building blocks of land surface models as they control transpiration and photosynthesis. Although more stomatal trait data are needed to significantly reduce the error in these model predictions, recording these traits is time-consuming, and no standardized protocol is currently available. Some attempts were made to automate stomatal detection from photomicrographs; however, these approaches have the disadvantage of using classic image processing or targeting a narrow taxonomic entity which makes these technologies less robust and generalizable to other plant species. We propose an easy-to-use and adaptable workflow from leaf to label. A methodology for automatic stomata detection was developed using deep neural networks according to the state of the art and its applicability demonstrated across the phylogeny of the angiosperms.We used a patch-based approach for training/tuning three different deep learning architectures. For training, we used 431 micrographs taken from leaf prints made according to the nail polish method from herbarium specimens of 19 species. The best-performing architecture was tested on 595 images of 16 additional species spread across the angiosperm phylogeny.The nail polish method was successfully applied in 78% of the species sampled here. The VGG19 architecture slightly outperformed the basic shallow and deep architectures, with a confidence threshold equal to 0.7 resulting in an optimal trade-off between precision and recall. Applying this threshold, the VGG19 architecture obtained an average F-score of 0.87, 0.89, and 0.67 on the training, validation, and unseen test set, respectively. The average accuracy was very high (94%) for computed stomatal counts on unseen images of species used for training.The leaf-to-label pipeline is an easy-to-use workflow for researchers of different areas of expertise interested in detecting stomata more efficiently. The described methodology was based on multiple species and well-established methods so that it can serve as a reference for future work.
| S-EPMC7487252 | biostudies-literature
Project description:Objective: To systematically analyze the supporting evidence, drug information, and the type of off-label drug use in recommendations on off-label drug use in pediatric guidelines. Methods: A cross-sectional study was performed by systematic search through MEDLINE (via PubMed) and Embase databases to identify literature published from 1 January 2018, to 31 December 2020. Only pediatric clinical practice guidelines that included recommendations on off-label use of drugs were included. We present descriptive information on the sources of the included guidelines, country, publication year, evidence grading system used, details on the types of off-label drug use, and the types of studies used as references to support the recommendations. Results: A total of 66 pediatric guidelines with 605 recommendations were included. Eighty-seven (14.4%) recommendations did not cite any references; and the remaining 518 recommendations were supported by 2,240 references (mean 4.3 references/recommendation). The most common types of studies cited were pediatric RCTs (n = 314, 14.0%), pediatric case series studies (n = 260, 11.6%), and reviews (n = 255, 11.4%). Twenty-one percent (n = 470) of the references were studies on adults. One hundred and forty (23.1%) recommendations were graded using the Grading of Recommendations, Assessments, Development, and Evaluations (GRADE) system, of which 37 (26.4%) were graded as strong but supported with only C or D level of evidence. The most commonly reported type of information in the recommendations was indication (n = 499, 82.5%). The most commonly addressed type of off-label drug use in the 523 positive recommendations was unapproved population (n = 255, 48.8%). Sixty-nine (11.4%) recommendations explicitly reported the drug use as off-label. Conclusion: Children may be exposed to medical risks due to gaps in reporting and evidence of off-label drug use recommendations in pediatric guidelines.
| S-EPMC9218488 | biostudies-literature
Project description:The chest X-ray is one of the most common radiological examination types for the diagnosis of chest diseases. Nowadays, the automatic classification technology of radiological images has been widely used in clinical diagnosis and treatment plans. However, each disease has its own different response characteristic receptive field region, which is the main challenge for chest disease classification tasks. Besides, the imbalance of sample data categories further increases the difficulty of tasks. To solve these problems, we propose a new multi-label chest disease image classification scheme based on a multi-scale attention network. In this scheme, multi-scale information is iteratively fused to focus on regions with a high probability of disease, to effectively mine more meaningful information from data. A novel loss function is also designed to improve the rationality of visual perception and multi-label image classification, which forces the consistency of attention regions before and after image transformation. A comprehensive experiment was carried out on the Chest X-Ray14 and CheXpert datasets, separately containing over 100,000 frontal-view and 200,000 front and side view X-ray images with 14 diseases. The AUROC is 0.850 and 0.815 respectively on the two data sets, which achieve the state-of-the-art results, verified the effectiveness of this method in chest X-ray image classification. This study has important practical significance for using AI algorithms to assist radiologists in improving work efficiency and diagnostic accuracy.
| S-EPMC8157016 | biostudies-literature
Project description:Medical datasets inherently contain errors from subjective or inaccurate test results, or from confounding biological complexities. It is difficult for medical experts to detect these elusive errors manually, due to lack of contextual information, limiting data privacy regulations, and the sheer scale of data to be reviewed. Current methods for training robust artificial intelligence (AI) models on data containing mislabeled examples generally fall into one of several categories-attempting to improve the robustness of the model architecture, the regularization techniques used, the loss function used during training, or selecting a subset of data that contains cleaner labels. This last category requires the ability to efficiently detect errors either prior to or during training, either relabeling them or removing them completely. More recent progress in error detection has focused on using multi-network learning to minimize deleterious effects of errors on training, however, using many neural networks to reach a consensus on which data should be removed can be computationally intensive and inefficient. In this work, a deep-learning based algorithm was used in conjunction with a label-clustering approach to automate error detection. For dataset with synthetic label flips added, these errors were identified with an accuracy of up to 85%, while requiring up to 93% less computing resources to complete compared to a previous model consensus approach developed previously. The resulting trained AI models exhibited greater training stability and up to a 45% improvement in accuracy, from 69 to over 99% compared to the consensus approach, at least 10% improvement on using noise-robust loss functions in a binary classification problem, and a 51% improvement for multi-class classification. These results indicate that practical, automated a priori detection of errors in medical data is possible, without human oversight.
| S-EPMC10638377 | biostudies-literature
Project description:Mont Blanc, the summit of Western Europe, is a popular but demanding high-altitude ascent. Drug use is thought to be widespread among climbers attempting this summit, not only to prevent altitude illnesses, but also to boost physical and/or psychological capacities. This practice may be unsafe in this remote alpine environment. However, robust data on medication during the ascent of Mont Blanc are lacking. Individual urine samples from male climbers using urinals in mountain refuges on access routes to Mont Blanc (Goûter and Cosmiques mountain huts) were blindly and anonymously collected using a hidden automatic sampler. Urine samples were screened for a wide range of drugs, including diuretics, glucocorticoids, stimulants, hypnotics and phosphodiesterase 5 (PDE-5) inhibitors. Out of 430 samples analyzed from both huts, 35.8% contained at least one drug. Diuretics (22.7%) and hypnotics (12.9%) were the most frequently detected drugs, while glucocorticoids (3.5%) and stimulants (3.1%) were less commonly detected. None of the samples contained PDE-5 inhibitors. Two substances were predominant: the diuretic acetazolamide (20.6%) and the hypnotic zolpidem (8.4%). Thirty three samples were found positive for at least two substances, the most frequent combination being acetazolamide and a hypnotic (2.1%). Based on a novel sampling technique, we demonstrate that about one third of the urine samples collected from a random sample of male climbers contained one or several drugs, suggesting frequent drug use amongst climbers ascending Mont Blanc. Our data suggest that medication primarily aims at mitigating the symptoms of altitude illnesses, rather than enhancing performance. In this hazardous environment, the relatively high prevalence of hypnotics must be highlighted, since these molecules may alter vigilance.
| S-EPMC4890938 | biostudies-literature
Project description:The term off-label drug use (OLDU) is used extensively in the medical literature, continuing medical education exercises, and the media. Yet, we propose that many health care professionals have an underappreciation of its definition, prevalence, and implications. This article introduces and answers 10 questions regarding OLDU in an effort to clarify the practice's meaning, breadth of application, acceptance, and liabilities. Off-label drug use involves prescribing medications for indications, or using a dosage or dosage form, that have not been approved by the US Food and Drug Administration. Since the Food and Drug Administration does not regulate the practice of medicine, OLDU has become common. It occurs in every specialty of medicine, but it may be more common in areas of medicine in which the patient population is less likely to be included in clinical trials (eg, pediatric, pregnant, or psychiatric patients). Pharmaceutical companies are not allowed to promote their medications for an off-label use, which has lead to several large settlements for illegal marketing. To limit liability, physicians should prescribe medications only for indications that they believe are in the best interest of the patient. In addition, health care professionals should educate themselves about OLDU to weigh the risks and benefits and provide the best possible care for their patients.
| S-EPMC3538391 | biostudies-literature
Project description:BackgroundOff-label drug use in children is common and potentially harmful. In most previous off-label use research, authors studied hospitalized children, specific drug classes, or non-US settings. We characterized frequencies, trends, and reasons for off-label systemic drug orders for children in ambulatory US settings.MethodsUsing nationally representative surveys of office-based physicians (National Ambulatory Medical Care Surveys, 2006-2015), we studied off-label orders of systemic drugs for children age <18 based on US Food and Drug Administration-approved labeling for age, weight, and indication. We characterized the top classes and diagnoses with off-label orders and analyzed factors and trends of off-label orders using logistic regression.ResultsPhysicians ordered ≥1 off-label systemic drug at 18.5% (95% confidence interval: 17.7%-19.3%) of visits, usually (74.6%) because of unapproved conditions. Off-label ordering was most common proportionally in neonates (83%) and in absolute terms among adolescents (322 orders out of 1000 visits). Off-label ordering was associated with female sex, subspecialists, polypharmacy, and chronic conditions. Rates and reasons for off-label orders varied considerably by age. Relative and absolute rates of off-label orders rose over time. Among common classes, off-label orders for antihistamines and several psychotropics increased over time, whereas off-label orders for several classes of antibiotics were stable or declined.ConclusionsUS office-based physicians have ordered systemic drugs off label for children at increasing rates, most often for unapproved conditions, despite recent efforts to increase evidence and drug approvals for children. These findings can help inform education, research, and policies around effective, safe use of medications in children.
| S-EPMC7286122 | biostudies-literature
Project description:BackgroundMedications are frequently prescribed outside their regulatory approval (off-label) by physicians particularly where appropriate therapies are not available. However, the risk/benefit ratio of drugs in off-label use needs to be critically appraised because it may differ from approved on-label usage. Therefore, an extensive exploration of current evidence on clinical data is well-advised. The objective of this study was to develop a search strategy that facilitates detection of the off-label drug use documents in EMBASE via OvidSP.MethodsWe constructed two sets of gold standards from relevant records to off-label drug use by a sensitive search of MEDLINE and EMBASE. Search queries, including search words and strings, were conceived based on definition of off-label use of medications as well as text analysis of 500 randomly selected relevant documents. The selected terms were searched in EMBASE (from 1988 to 2011) and their retrieval performance was compared with the gold standards. We developed a sensitivity-maximizing, and a sensitivity- and precision-maximizing search strategy.ResultsFrom 4067 records relevant to off-label drug use in our full gold standard set, 3846 records were retrievable from EMBASE. "off label*.af." was the most sensitive single term (overall sensitivity 77.5%, sensitivity within EMBASE 81.9%, precision 88.1%). The highest sensitive search strategy was achieved by combining 36 search queries with overall sensitivity of 94.0% and precision of 69.5%. An optimal sensitive and precise search strategy was yielded precision 87.4% at the expense of decreasing overall sensitivity to 89.4%.ConclusionWe developed highly sensitive search strategies to enhance the retrieval of studies on off-label drug use in OvidSP EMBASE.
| S-EPMC3543848 | biostudies-literature
Project description:In the microbiology laboratory, there is an augmented need for rapid screening methods for the detection of bacteria in urine samples, since about two-thirds of these samples will not yield any bacteria or will yield insignificant growth when cultured. Thus, a reliable screening method can free up laboratory resources and can speed up the reporting of a negative urine result. In this study, we have evaluated the detection of leukocytes, bacteria, and a new sediment indicator, the "all small particles" (ASP), by an automated instrument, the iQ200 urine analyzer, to detect negative urine samples that can be excluded from culture. A coupled automated strip reader (iChem Velocity), enabling the detection of nitrite and leukocyte esterase, was tested in parallel. In total, 963 urine samples were processed through both conventional urine culture and the iQ200/iChem Velocity workstation. Using the data, a multivariate regression model was established, and the predicted specificity and the possible reduction in urine cultures were calculated for the indicators and their respective combinations (leukocytes plus bacteria plus ASP and leukocyte esterase plus nitrite). Among all options, diagnostic performance was best using the whole microscopic content of the sample (leukocytes plus bacteria plus ASP). By using a cutoff value of ? 10(4) CFU/ml for defining a positive culture, a given sensitivity of 95% resulted in a specificity of 61% and a reduction in urine cultures of 35%. By considering the indicators alone, specificity and the culture savings were both much less satisfactory. The regression model was also used to determine possible cutoff values for running the instrument as part of daily routine. By using a graphical representation of all combinations possible, we derived cutoff values for leukocyte, bacterial, and ASP count, which should enable the iQ200 microscope to screen out approximately one-third of the urine samples, significantly reducing the workload in the microbiology laboratory.
| S-EPMC4136165 | biostudies-literature