Predicting Future Disorders via Temporal Knowledge Graphs and Medical Ontologies.

Despite the vast potential for insights and value present in Electronic Health Records (EHRs), it is challenging to fully leverage all the available information, particularly that contained in the free-text data written by clinicians describing the health status of patients. The utilization of Named Entity Recognition and Linking tools allows not only for the structuring of information contained within free-text data, but also for the integration with medical ontologies, which may prove highly beneficial for the analysis of patient medical histories with the aim of forecasting future medical outcomes, such as the diagnosis of a new disorder. In this paper, we propose MedTKG, a Temporal Knowledge Graph (TKG) framework that incorporates both the dynamic information of patient clinical histories and the static information of medical ontologies. The TKG is used to model a medical history as a series of snapshots at different points in time, effectively capturing the dynamic nature of the patient's health status, while a static graph is used to model the hierarchies of concepts extracted from domain ontologies. The proposed method aims to predict future disorders by identifying missing objects in the quadruple 〈s, r, ?, t 〉, where s and r denote the patient and the disorder relation type, respectively, and t is the timestamp of the query. The method is evaluated on clinical notes extracted from MIMIC-III and demonstrates the effectiveness of the TKG framework in predicting future disorders and of medical ontologies in improving its performance.

Foresight-a generative pretrained transformer for modelling of patient timelines using electronic health records: a retrospective modelling study.

BACKGROUND: An electronic health record (EHR) holds detailed longitudinal information about a patient's health status and general clinical history, a large portion of which is stored as unstructured, free text. Existing approaches to model a patient's trajectory focus mostly on structured data and a subset of single-domain outcomes. This study aims to evaluate the effectiveness of Foresight, a generative transformer in temporal modelling of patient data, integrating both free text and structured formats, to predict a diverse array of future medical outcomes, such as disorders, substances (eg, to do with medicines, allergies, or poisonings), procedures, and findings (eg, relating to observations, judgements, or assessments). METHODS: Foresight is a novel transformer-based pipeline that uses named entity recognition and linking tools to convert EHR document text into structured, coded concepts, followed by providing probabilistic forecasts for future medical events, such as disorders, substances, procedures, and findings. The Foresight pipeline has four main components: (1) CogStack (data retrieval and preprocessing); (2) the Medical Concept Annotation Toolkit (structuring of the free-text information from EHRs); (3) Foresight Core (deep-learning model for biomedical concept modelling); and (4) the Foresight web application. We processed the entire free-text portion from three different hospital datasets (King's College Hospital [KCH], South London and Maudsley [SLaM], and the US Medical Information Mart for Intensive Care III [MIMIC-III]), resulting in information from 811 336 patients and covering both physical and mental health institutions. We measured the performance of models using custom metrics derived from precision and recall. FINDINGS: Foresight achieved a precision@10 (ie, of 10 forecasted candidates, at least one is correct) of 0·68 (SD 0·0027) for the KCH dataset, 0·76 (0·0032) for the SLaM dataset, and 0·88 (0·0018) for the MIMIC-III dataset, for forecasting the next new disorder in a patient timeline. Foresight also achieved a precision@10 value of 0·80 (0·0013) for the KCH dataset, 0·81 (0·0026) for the SLaM dataset, and 0·91 (0·0011) for the MIMIC-III dataset, for forecasting the next new biomedical concept. In addition, Foresight was validated on 34 synthetic patient timelines by five clinicians and achieved a relevancy of 33 (97% [95% CI 91-100]) of 34 for the top forecasted candidate disorder. As a generative model, Foresight can forecast follow-on biomedical concepts for as many steps as required. INTERPRETATION: Foresight is a general-purpose model for biomedical concept modelling that can be used for real-world risk forecasting, virtual trials, and clinical research to study the progression of disorders, to simulate interventions and counterfactuals, and for educational purposes. FUNDING: National Health Service Artificial Intelligence Laboratory, National Institute for Health and Care Research Biomedical Research Centre, and Health Data Research UK.

Artificial intelligence (AI) for neurologists: do digital neurones dream of electric sheep?

Artificial intelligence (AI) is routinely mentioned in journals and newspapers, and non-technical outsiders may have difficulty in distinguishing hyperbole from reality. We present a practical guide to help non-technical neurologists to understand healthcare AI. AI is being used to support clinical decisions in treating neurological disorders. We introduce basic concepts of AI, such as machine learning and natural language processing, and explain how AI is being used in healthcare, giving examples its benefits and challenges. We also cover how AI performance is measured, and its regulatory aspects in healthcare. An important theme is that AI is a general-purpose technology like medical statistics, with broad utility applicable in various scenarios, such that niche approaches are outpaced by approaches that are broadly applicable in many disease areas and specialties. By understanding AI basics and its potential applications, neurologists can make informed decisions when evaluating AI used in their clinical practice. This article was written by four humans, with generative AI helping with formatting and image generation.

AI chatbots not yet ready for clinical use.

As large language models (LLMs) expand and become more advanced, so do the natural language processing capabilities of conversational AI, or "chatbots". OpenAI's recent release, ChatGPT, uses a transformer-based model to enable human-like text generation and question-answering on general domain knowledge, while a healthcare-specific Large Language Model (LLM) such as GatorTron has focused on the real-world healthcare domain knowledge. As LLMs advance to achieve near human-level performances on medical question and answering benchmarks, it is probable that Conversational AI will soon be developed for use in healthcare. In this article we discuss the potential and compare the performance of two different approaches to generative pretrained transformers-ChatGPT, the most widely used general conversational LLM, and Foresight, a GPT (generative pretrained transformer) based model focused on modelling patients and disorders. The comparison is conducted on the task of forecasting relevant diagnoses based on clinical vignettes. We also discuss important considerations and limitations of transformer-based chatbots for clinical use.

Hospital-wide natural language processing summarising the health data of 1 million patients.

Electronic health records (EHRs) represent a major repository of real world clinical trajectories, interventions and outcomes. While modern enterprise EHR's try to capture data in structured standardised formats, a significant bulk of the available information captured in the EHR is still recorded only in unstructured text format and can only be transformed into structured codes by manual processes. Recently, Natural Language Processing (NLP) algorithms have reached a level of performance suitable for large scale and accurate information extraction from clinical text. Here we describe the application of open-source named-entity-recognition and linkage (NER+L) methods (CogStack, MedCAT) to the entire text content of a large UK hospital trust (King's College Hospital, London). The resulting dataset contains 157M SNOMED concepts generated from 9.5M documents for 1.07M patients over a period of 9 years. We present a summary of prevalence and disease onset as well as a patient embedding that captures major comorbidity patterns at scale. NLP has the potential to transform the health data lifecycle, through large-scale automation of a traditionally manual task.

Diagnostic signature for heart failure with preserved ejection fraction (HFpEF): a machine learning approach using multi-modality electronic health record data.

BACKGROUND: Heart failure with preserved ejection fraction (HFpEF) is thought to be highly prevalent yet remains underdiagnosed. Evidence-based treatments are available that increase quality of life and decrease hospitalization. We sought to develop a data-driven diagnostic model to predict from electronic health records (EHR) the likelihood of HFpEF among patients with unexplained dyspnea and preserved left ventricular EF. METHODS AND RESULTS: The derivation cohort comprised patients with dyspnea and echocardiography results. Structured and unstructured data were extracted using an automated informatics pipeline. Patients were retrospectively diagnosed as HFpEF (cases), non-HF (control cohort I), or HF with reduced EF (HFrEF; control cohort II). The ability of clinical parameters and investigations to discriminate cases from controls was evaluated by extreme gradient boosting. A likelihood scoring system was developed and validated in a separate test cohort. The derivation cohort included 1585 consecutive patients: 133 cases of HFpEF (9%), 194 non-HF cases (Control cohort I) and 1258 HFrEF cases (Control cohort II). Two HFpEF diagnostic signatures were derived, comprising symptoms, diagnoses and investigation results. A final prediction model was generated based on the averaged likelihood scores from these two models. In a validation cohort consisting of 269 consecutive patients [with 66 HFpEF cases (24.5%)], the diagnostic power of detecting HFpEF had an AUROC of 90% (P < 0.001) and average precision of 74%. CONCLUSION: This diagnostic signature enables discrimination of HFpEF from non-cardiac dyspnea or HFrEF from EHR and can assist in the diagnostic evaluation in patients with unexplained dyspnea. This approach will enable identification of HFpEF patients who may then benefit from new evidence-based therapies.

Mapping multimorbidity in individuals with schizophrenia and bipolar disorders: evidence from the South London and Maudsley NHS Foundation Trust Biomedical Research Centre (SLAM BRC) case register.

OBJECTIVES: The first aim of this study was to design and develop a valid and replicable strategy to extract physical health conditions from clinical notes which are common in mental health services. Then, we examined the prevalence of these conditions in individuals with severe mental illness (SMI) and compared their individual and combined prevalence in individuals with bipolar (BD) and schizophrenia spectrum disorders (SSD). DESIGN: Observational study. SETTING: Secondary mental healthcare services from South London PARTICIPANTS: Our maximal sample comprised 17 500 individuals aged 15 years or older who had received a primary or secondary SMI diagnosis (International Classification of Diseases, 10th edition, F20-31) between 2007 and 2018. MEASURES: We designed and implemented a data extraction strategy for 21 common physical comorbidities using a natural language processing pipeline, MedCAT. Associations were investigated with sex, age at SMI diagnosis, ethnicity and social deprivation for the whole cohort and the BD and SSD subgroups. Linear regression models were used to examine associations with disability measured by the Health of Nations Outcome Scale. RESULTS: Physical health data were extracted, achieving precision rates (F1) above 0.90 for all conditions. The 10 most prevalent conditions were diabetes, hypertension, asthma, arthritis, epilepsy, cerebrovascular accident, eczema, migraine, ischaemic heart disease and chronic obstructive pulmonary disease. The most prevalent combination in this population included diabetes, hypertension and asthma, regardless of their SMI diagnoses. CONCLUSIONS: Our data extraction strategy was found to be adequate to extract physical health data from clinical notes, which is essential for future multimorbidity research using text records. We found that around 40% of our cohort had multimorbidity from which 20% had complex multimorbidity (two or more physical conditions besides SMI). Sex, age, ethnicity and social deprivation were found to be key to understand their heterogeneity and their differential contribution to disability levels in this population. These outputs have direct implications for researchers and clinicians.

Inpatient COVID-19 mortality has reduced over time: Results from an observational cohort.

BACKGROUND: The Covid-19 pandemic in the United Kingdom has seen two waves; the first starting in March 2020 and the second in late October 2020. It is not known whether outcomes for those admitted with severe Covid were different in the first and second waves. METHODS: The study population comprised all patients admitted to a 1,500-bed London Hospital Trust between March 2020 and March 2021, who tested positive for Covid-19 by PCR within 3-days of admissions. Primary outcome was death within 28-days of admission. Socio-demographics (age, sex, ethnicity), hypertension, diabetes, obesity, baseline physiological observations, CRP, neutrophil, chest x-ray abnormality, remdesivir and dexamethasone were incorporated as co-variates. Proportional subhazards models compared mortality risk between wave 1 and wave 2. Cox-proportional hazard model with propensity score adjustment were used to compare mortality in patients prescribed remdesivir and dexamethasone. RESULTS: There were 3,949 COVID-19 admissions, 3,195 hospital discharges and 733 deaths. There were notable differences in age, ethnicity, comorbidities, and admission disease severity between wave 1 and wave 2. Twenty-eight-day mortality was higher during wave 1 (26.1% versus 13.1%). Mortality risk adjusted for co-variates was significantly lower in wave 2 compared to wave 1 [adjSHR 0.49 (0.37, 0.65) p<0.001]. Analysis of treatment impact did not show statistically different effects of remdesivir [HR 0.84 (95%CI 0.65, 1.08), p = 0.17] or dexamethasone [HR 0.97 (95%CI 0.70, 1.35) p = 0.87]. CONCLUSION: There has been substantial improvements in COVID-19 mortality in the second wave, even accounting for demographics, comorbidity, and disease severity. Neither dexamethasone nor remdesivir appeared to be key explanatory factors, although there may be unmeasured confounding present.

A Knowledge Distillation Ensemble Framework for Predicting Short- and Long-Term Hospitalization Outcomes From Electronic Health Records Data.

The ability to perform accurate prognosis is crucial for proactive clinical decision making, informed resource management and personalised care. Existing outcome prediction models suffer from a low recall of infrequent positive outcomes. We present a highly-scalable and robust machine learning framework to automatically predict adversity represented by mortality and ICU admission and readmission from time-series of vital signs and laboratory results obtained within the first 24 hours of hospital admission. The stacked ensemble platform comprises two components: a) an unsupervised LSTM Autoencoder that learns an optimal representation of the time-series, using it to differentiate the less frequent patterns which conclude with an adverse event from the majority patterns that do not, and b) a gradient boosting model, which relies on the constructed representation to refine prediction by incorporating static features. The model is used to assess a patient's risk of adversity and provides visual justifications of its prediction. Results of three case studies show that the model outperforms existing platforms in ICU and general ward settings, achieving average Precision-Recall Areas Under the Curve (PR-AUCs) of 0.891 (95% CI: 0.878-0.939) for mortality and 0.908 (95% CI: 0.870-0.935) in predicting ICU admission and readmission.

Investigating the association between physical health comorbidities and disability in individuals with severe mental illness.

BACKGROUND: Research suggests that an increased risk of physical comorbidities might have a key role in the association between severe mental illness (SMI) and disability. We examined the association between physical multimorbidity and disability in individuals with SMI. METHODS: Data were extracted from the clinical record interactive search system at South London and Maudsley Biomedical Research Centre. Our sample (n = 13,933) consisted of individuals who had received a primary or secondary SMI diagnosis between 2007 and 2018 and had available data for Health of Nations Outcome Scale (HoNOS) as disability measure. Physical comorbidities were defined using Chapters II-XIV of the International Classification of Diagnoses (ICD-10). RESULTS: More than 60 % of the sample had complex multimorbidity. The most common organ system affected were neurological (34.7%), dermatological (15.4%), and circulatory (14.8%). All specific comorbidities (ICD-10 Chapters) were associated with higher levels of disability, HoNOS total scores. Individuals with musculoskeletal, skin/dermatological, respiratory, endocrine, neurological, hematological, or circulatory disorders were found to be associated with significant difficulties associated with more than five HoNOS domains while others had a lower number of domains affected. CONCLUSIONS: Individuals with SMI and musculoskeletal, skin/dermatological, respiratory, endocrine, neurological, hematological, or circulatory disorders are at higher risk of disability compared to those who do not have those comorbidities. Individuals with SMI and physical comorbidities are at greater risk of reporting difficulties associated with activities of daily living, hallucinations, and cognitive functioning. Therefore, these should be targeted for prevention and intervention programs.

Natural language word embeddings as a glimpse into healthcare language and associated mortality surrounding end of life.

OBJECTIVES: To clarify real-world linguistic nuances around dying in hospital as well as inaccuracy in individual-level prognostication to support advance care planning and personalised discussions on limitation of life sustaining treatment (LST). DESIGN: Retrospective cross-sectional study of real-world clinical data. SETTING: Secondary care, urban and suburban teaching hospitals. PARTICIPANTS: All inpatients in 12-month period from 1 October 2018 to 30 September 2019. METHODS: Using unsupervised natural language processing, word embedding in latent space was used to generate phrase clusters with most similar semantic embeddings to 'Ceiling of Treatment' and their prognostication value. RESULTS: Word embeddings with most similarity to 'Ceiling of Treatment' clustered around phrases describing end-of-life care, ceiling of care and LST discussions. The phrases have differing prognostic profile with the highest 7-day mortality in the phrases most explicitly referring to end of life-'Withdrawal of care' (56.7%), 'terminal care/end of life care' (57.5%) and 'un-survivable' (57.6%). CONCLUSION: Vocabulary used at end-of-life discussions are diverse and has a range of associations to 7-day mortality. This highlights the importance of correct application of terminology during LST and end-of-life discussions.

Regional performance variation in external validation of four prediction models for severity of COVID-19 at hospital admission: An observational multi-centre cohort study.

BACKGROUND: Prediction models should be externally validated to assess their performance before implementation. Several prediction models for coronavirus disease-19 (COVID-19) have been published. This observational cohort study aimed to validate published models of severity for hospitalized patients with COVID-19 using clinical and laboratory predictors. METHODS: Prediction models fitting relevant inclusion criteria were chosen for validation. The outcome was either mortality or a composite outcome of mortality and ICU admission (severe disease). 1295 patients admitted with symptoms of COVID-19 at Kings Cross Hospital (KCH) in London, United Kingdom, and 307 patients at Oslo University Hospital (OUH) in Oslo, Norway were included. The performance of the models was assessed in terms of discrimination and calibration. RESULTS: We identified two models for prediction of mortality (referred to as Xie and Zhang1) and two models for prediction of severe disease (Allenbach and Zhang2). The performance of the models was variable. For prediction of mortality Xie had good discrimination at OUH with an area under the receiver-operating characteristic (AUROC) 0.87 [95% confidence interval (CI) 0.79-0.95] and acceptable discrimination at KCH, AUROC 0.79 [0.76-0.82]. In prediction of severe disease, Allenbach had acceptable discrimination (OUH AUROC 0.81 [0.74-0.88] and KCH AUROC 0.72 [0.68-0.75]). The Zhang models had moderate to poor discrimination. Initial calibration was poor for all models but improved with recalibration. CONCLUSIONS: The performance of the four prediction models was variable. The Xie model had the best discrimination for mortality, while the Allenbach model had acceptable results for prediction of severe disease.

Multi-domain clinical natural language processing with MedCAT: The Medical Concept Annotation Toolkit.

Electronic health records (EHR) contain large volumes of unstructured text, requiring the application of information extraction (IE) technologies to enable clinical analysis. We present the open source Medical Concept Annotation Toolkit (MedCAT) that provides: (a) a novel self-supervised machine learning algorithm for extracting concepts using any concept vocabulary including UMLS/SNOMED-CT; (b) a feature-rich annotation interface for customizing and training IE models; and (c) integrations to the broader CogStack ecosystem for vendor-agnostic health system deployment. We show improved performance in extracting UMLS concepts from open datasets (F1:0.448-0.738 vs 0.429-0.650). Further real-world validation demonstrates SNOMED-CT extraction at 3 large London hospitals with self-supervised training over ∼8.8B words from ∼17M clinical records and further fine-tuning with ∼6K clinician annotated examples. We show strong transferability (F1 > 0.94) between hospitals, datasets and concept types indicating cross-domain EHR-agnostic utility for accelerated clinical and research use cases.

Evaluation and improvement of the National Early Warning Score (NEWS2) for COVID-19: a multi-hospital study.

BACKGROUND: The National Early Warning Score (NEWS2) is currently recommended in the UK for the risk stratification of COVID-19 patients, but little is known about its ability to detect severe cases. We aimed to evaluate NEWS2 for the prediction of severe COVID-19 outcome and identify and validate a set of blood and physiological parameters routinely collected at hospital admission to improve upon the use of NEWS2 alone for medium-term risk stratification. METHODS: Training cohorts comprised 1276 patients admitted to King's College Hospital National Health Service (NHS) Foundation Trust with COVID-19 disease from 1 March to 30 April 2020. External validation cohorts included 6237 patients from five UK NHS Trusts (Guy's and St Thomas' Hospitals, University Hospitals Southampton, University Hospitals Bristol and Weston NHS Foundation Trust, University College London Hospitals, University Hospitals Birmingham), one hospital in Norway (Oslo University Hospital), and two hospitals in Wuhan, China (Wuhan Sixth Hospital and Taikang Tongji Hospital). The outcome was severe COVID-19 disease (transfer to intensive care unit (ICU) or death) at 14 days after hospital admission. Age, physiological measures, blood biomarkers, sex, ethnicity, and comorbidities (hypertension, diabetes, cardiovascular, respiratory and kidney diseases) measured at hospital admission were considered in the models. RESULTS: A baseline model of 'NEWS2 + age' had poor-to-moderate discrimination for severe COVID-19 infection at 14 days (area under receiver operating characteristic curve (AUC) in training cohort = 0.700, 95% confidence interval (CI) 0.680, 0.722; Brier score = 0.192, 95% CI 0.186, 0.197). A supplemented model adding eight routinely collected blood and physiological parameters (supplemental oxygen flow rate, urea, age, oxygen saturation, C-reactive protein, estimated glomerular filtration rate, neutrophil count, neutrophil/lymphocyte ratio) improved discrimination (AUC = 0.735; 95% CI 0.715, 0.757), and these improvements were replicated across seven UK and non-UK sites. However, there was evidence of miscalibration with the model tending to underestimate risks in most sites. CONCLUSIONS: NEWS2 score had poor-to-moderate discrimination for medium-term COVID-19 outcome which raises questions about its use as a screening tool at hospital admission. Risk stratification was improved by including readily available blood and physiological parameters measured at hospital admission, but there was evidence of miscalibration in external sites. This highlights the need for a better understanding of the use of early warning scores for COVID.

Angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers are not associated with severe COVID-19 infection in a multi-site UK acute hospital trust.

AIMS: The SARS-CoV-2 virus binds to the angiotensin-converting enzyme 2 (ACE2) receptor for cell entry. It has been suggested that angiotensin-converting enzyme inhibitors (ACEi) and angiotensin II receptor blockers (ARB), which are commonly used in patients with hypertension or diabetes and may raise tissue ACE2 levels, could increase the risk of severe COVID-19 infection. METHODS AND RESULTS: We evaluated this hypothesis in a consecutive cohort of 1200 acute inpatients with COVID-19 at two hospitals with a multi-ethnic catchment population in London (UK). The mean age was 68 ± 17 years (57% male) and 74% of patients had at least one comorbidity. Overall, 415 patients (34.6%) reached the primary endpoint of death or transfer to a critical care unit for organ support within 21 days of symptom onset. A total of 399 patients (33.3%) were taking ACEi or ARB. Patients on ACEi/ARB were significantly older and had more comorbidities. The odds ratio for the primary endpoint in patients on ACEi and ARB, after adjustment for age, sex and co-morbidities, was 0.63 (95% confidence interval 0.47-0.84, P < 0.01). CONCLUSIONS: There was no evidence for increased severity of COVID-19 in hospitalised patients on chronic treatment with ACEi or ARB. A trend towards a beneficial effect of ACEi/ARB requires further evaluation in larger meta-analyses and randomised clinical trials.