ABSTRACT
RATIONALE AND OBJECTIVES: Given the high volume of chest radiographs, radiologists frequently encounter heavy workloads. In outpatient imaging, a substantial portion of chest radiographs show no actionable findings. Automatically identifying these cases could improve efficiency by facilitating shorter reading workflows. PURPOSE: A large-scale study to assess the performance of AI on identifying chest radiographs with no actionable disease (NAD) in an outpatient imaging population using comprehensive, objective, and reproducible criteria for NAD. MATERIALS AND METHODS: The independent validation study includes 15000 patients with chest radiographs in posterior-anterior (PA) and lateral projections from an outpatient imaging center in the United States. Ground truth was established by reviewing CXR reports and classifying cases as NAD or actionable disease (AD). The NAD definition includes completely normal chest radiographs and radiographs with well-defined non-actionable findings. The AI NAD Analyzer1 (trained with 100 million multimodal images and fine-tuned on 1.3 million radiographs) utilizes a tandem system with image-level rule in and compartment-level rule out to provide case level output as NAD or potential actionable disease (PAD). RESULTS: A total of 14057 cases met our eligibility criteria (age 56 ± 16.1 years, 55% women and 45% men). The prevalence of NAD cases in the study population was 70.7%. The AI NAD Analyzer correctly classified NAD cases with a sensitivity of 29.1% and a yield of 20.6%. The specificity was 98.9% which corresponds to a miss rate of 0.3% of cases. Significant findings were missed in 0.06% of cases, while no cases with critical findings were missed by AI. CONCLUSION: In an outpatient population, AI can identify 20% of chest radiographs as NAD with a very low rate of missed findings. These cases could potentially be read using a streamlined protocol, thus improving efficiency and consequently reducing daily workload for radiologists.
ABSTRACT
BACKGROUND: Time is a measurable and critical resource that affects the quality of services provided in clinical practice. There is limited insight into the effects of time restrictions on clinicians' cognitive processes with the electronic health record (EHR) in providing ambulatory care. OBJECTIVE: To understand the impact of time constraints on clinicians' synthesis of text-based EHR clinical notes. METHODS: We used an established clinician cognitive framework based on a think-aloud protocol. We studied interns' thought processes as they accomplished a set of four preformed ambulatory care clinical scenarios with and without time restrictions in a controlled setting. RESULTS: Interns most often synthesized details relevant to patients' problems and treatment, regardless of whether or not the time available for task performance was restricted. In contrast to previous findings, subsequent information commonly synthesized by clinicians related most commonly to the chronology of clinical events for the unrestricted time observations and to investigative procedures for the time-restricted sessions. There was no significant difference in the mean number of omission errors and incorrect deductions when interns synthesized the EHR clinical notes with and without time restrictions (3.5±0.5 vs. 2.3±0.5, p=0.14). CONCLUSION: Our results suggest that the incidence of errors during clinicians' synthesis of EHR clinical notes is not increased with modest time restrictions, possibly due to effective adjustments of information processing strategies learned from the usual time-constrained nature of patient visits. Further research is required to investigate the effects of similar or more extreme time variations on cognitive processes employed with different levels of expertise, specialty, and with different care settings.
Subject(s)
Electronic Health Records , Practice Patterns, Physicians' , User-Computer InterfaceABSTRACT
Evidence-based clinical guidelines are being developed to bridge the gap between research and practice with the goals of improving health care quality and population health. However, disseminating, implementing, and ensuring ongoing use of clinical guidelines in practice settings is challenging. The purpose of this study was to demonstrate the feasibility of encoding evidence-based clinical guidelines using the Omaha System. Clinical documentation with Omaha System-encoded guidelines generates individualized, meaningful data suitable for program evaluation and health care quality research. The use of encoded guidelines within the electronic health record has potential to reinforce use of guidelines, and thus improve health care quality and population health. Research using Omaha System data generated by clinicians has potential to discover new knowledge related to guideline use and effectiveness.
Subject(s)
Depression/diagnosis , Depression/therapy , Electronic Health Records , Mental Health Services/standards , Practice Guidelines as Topic , Vocabulary, Controlled , Clinical Coding , Feasibility Studies , Humans , Meaningful Use , Medical Informatics , Quality of Health CareABSTRACT
OBJECTIVES: Benchmark client outcomes across public health nursing (PHN) agencies using Omaha System knowledge, behavior, and status ratings as benchmarking metrics. DESIGN AND SAMPLE: A descriptive, comparative study of benchmark attainment for a retrospective cohort of PHN clients (low-income, high-risk parents, primarily mothers) from 6 counties. MEASURES: Omaha System Problem Rating Scale for Outcomes data for selected problems. Benchmark measures were defined as a rating of 4 on a scale from 1 (lowest) to 5 (highest). INTERVENTION: Family home visiting services to low-income, high-risk parents. RESULTS: The highest percentage of benchmark attainment was for the Postpartum problem (knowledge, 76.2%; behavior, 94.0%; status, 96.6%), and the lowest was for the Interpersonal relationship problem (knowledge, 21.7%; behavior, 69.0%; status, 40.7%). All counties showed significant increases in client knowledge benchmark attainment, and 4 of 6 counties showed significant increases from baseline in behavior and status benchmark attainment. Significant differences were found between counties in client characteristics and benchmark attainment for knowledge, behavior, and status outcomes. CONCLUSIONS: There were consistent patterns in benchmark attainment and outcome improvement across counties and family home visiting studies. Benchmarking appears to be useful for comparison of population health status and home visiting program outcomes.
Subject(s)
Benchmarking/methods , Child Welfare/statistics & numerical data , Clinical Competence/standards , Home Nursing/standards , Maternal Welfare/statistics & numerical data , Public Health Nursing/standards , Adult , Child , Child, Preschool , Female , Health Knowledge, Attitudes, Practice , Home Nursing/methods , Humans , Infant , Infant, Newborn , Minnesota , Pregnancy , United States , Young AdultABSTRACT
Disease named entity recognition (NER) is a critical task for most biomedical natural language processing (NLP) applications. For example, extracting diseases from clinical trial text can be helpful for patient profiling and other downstream applications such as matching clinical trials to eligible patients. Similarly, disease annotation in biomedical articles can help information search engines to accurately index them such that clinicians can easily find relevant articles to enhance their knowledge. In this paper, we propose a domain knowledge-enhanced long short-term memory network-conditional random field (LSTM-CRF) model for disease named entity recognition, which also augments a character-level convolutional neural network (CNN) and a character-level LSTM network for input embedding. Experimental results on a scientific article dataset show the effectiveness of our proposed models compared to state-of-the-art methods in disease recognition.
ABSTRACT
This paper explores cutting-edge deep learning methods for information extraction from medical imaging free text reports at a multi-institutional scale and compares them to the state-of-the-art domain-specific rule-based system - PEFinder and traditional machine learning methods - SVM and Adaboost. We proposed two distinct deep learning models - (i) CNN Word - Glove, and (ii) Domain phrase attention-based hierarchical recurrent neural network (DPA-HNN), for synthesizing information on pulmonary emboli (PE) from over 7370 clinical thoracic computed tomography (CT) free-text radiology reports collected from four major healthcare centers. Our proposed DPA-HNN model encodes domain-dependent phrases into an attention mechanism and represents a radiology report through a hierarchical RNN structure composed of word-level, sentence-level and document-level representations. Experimental results suggest that the performance of the deep learning models that are trained on a single institutional dataset, are better than rule-based PEFinder on our multi-institutional test sets. The best F1 score for the presence of PE in an adult patient population was 0.99 (DPA-HNN) and for a pediatrics population was 0.99 (HNN) which shows that the deep learning models being trained on adult data, demonstrated generalizability to pediatrics population with comparable accuracy. Our work suggests feasibility of broader usage of neural network models in automated classification of multi-institutional imaging text reports for a variety of applications including evaluation of imaging utilization, imaging yield, clinical decision support tools, and as part of automated classification of large corpus for medical imaging deep learning work.
Subject(s)
Deep Learning , Neural Networks, Computer , Pulmonary Embolism/diagnostic imaging , Radiography, Thoracic , Humans , Information Storage and RetrievalABSTRACT
The emerging penetration of Health IT in Latin America (especially in Brazil) has exacerbated the ever-increasing amount of Electronic Health Record (EHR) clinical free text documents.This imposes a workflow efficiency challenge on clinicians who need to synthesize such documents during the typically time-constrained patient care. We propose an ontology-driven semantic search framework that effectively supports clinicians' information synthesis at the point of care.
Subject(s)
Biological Ontologies/organization & administration , Decision Support Systems, Clinical/organization & administration , Electronic Health Records/organization & administration , Information Storage and Retrieval/methods , Point-of-Care Systems/organization & administration , Semantics , Brazil , Humans , Machine Learning , Natural Language Processing , Portugal , Terminology as Topic , User-Computer Interface , WorkflowABSTRACT
Clinicians utilize electronic health record (EHR) systems during time-constrained patient encounters where large amounts of clinical text must be synthesized at the point of care. Qualitative methods may be an effective approach for uncovering cognitive processes associated with the synthesis of clinical documents within EHR systems. We utilized a think-aloud protocol and content analysis with the goal of understanding cognitive processes and barriers involved as medical interns synthesized patient clinical documents in an EHR system to accomplish routine clinical tasks. Overall, interns established correlations of significance and meaning between problem, symptom and treatment concepts to inform hypotheses generation and clinical decision-making. Barriers identified with synthesizing EHR documents include difficulty searching for patient data, poor readability, redundancy, and unfamiliar specialized terms. Our study can inform recommendations for future designs of EHR clinical document user interfaces to aid clinicians in providing improved patient care.
Subject(s)
Cognition , Electronic Health Records , Internship and Residency , Humans , Physicians/psychologyABSTRACT
Ontologias terminológicas padronizadas e corretamente traduzidas são essenciais para o desenvolvimento de aplicações de processamento de linguagem natural na área da saúde. Para o desenvolvimento de uma aplicação de busca semântica em narrativas clínicas em português se fez necessária a utilização dos termos clínicos da Unified Medical Language System (UMLS). OBJETIVOS: Traduzir termos da UMLS em Português Europeu para Português Brasileiro. MÉTODOS: Foi desenvolvido um algoritmo de tradução semi-automática baseada em regras de substituição de texto. RESULTADOS: Após execução do algoritmo e avaliação por parte de especialistas, o algoritmo deixou de traduzir corretamente apenas 0.1% dos termos da base de testes. CONCLUSÃO: A utilização do método proposto se mostrou efetivo na tradução dos termos da UMLS e pode auxiliar em posteriores adaptações de listagens em Português Europeu para Português Brasileiro.
Correctly translated and standardized clinical ontologies are essential for development of Natural LanguageProcessing application for the medical domain. To develop an ontology-driven semantic search application for Portuguese clinical notes we needed to implement the Unified Medical Language System (UMLS) ontologies, specifically for Brazilian Portuguese. OBJECTIVES: To translate UMLS terms from European Portuguese to Brazilian Portuguese. METHODS: To develop a semi-automatic translation algorithm based on string replacement rules. RESULTS: Following the experiments and specialists' evaluation the algorithm mis-translated only 0.1% of terms in our test set. CONCLUSION: The proposed method proved to be effective for UMLS clinical terms translation and can be useful for posterior adaption ofa set of clinical terms from European Portuguese to Brazilian Portuguese.