Real-world Electronic Health Records Research Articles

Towards artificial intelligence-based disease prediction algorithms that comprehensively leverage and continuously learn from real-world clinical tabular data systems.

This manuscript presents a proof-of-concept for a generalizable strategy, the full algorithm, designed to estimate disease risk using real-world clinical tabular data systems, such as electronic health records (EHR) or claims databases. By integrating classic statistical methods and modern artificial intelligence techniques, this strategy automates the production of a disease prediction model that comprehensively reflects the dynamics contained within the underlying data system. Specifically, the full algorithm parses through every facet of the data (e.g., encounters, diagnoses, procedures, medications, labs, chief complaints, flowsheets, vital signs, demographics, etc.), selects which factors to retain as predictor variables by evaluating the data empirically against statistical criteria, structures and formats the retained data into time-series, trains a neural network-based prediction model, then subsequently applies this model to current patients to generate risk estimates. A distinguishing feature of the proposed strategy is that it produces a self-adaptive prediction system, capable of evolving the prediction mechanism in response to changes within the data: as newly collected data expand/modify the dataset organically, the prediction mechanism automatically evolves to reflect these changes. Moreover, the full algorithm operates without the need for a-priori data curation and aims to harness all informative risk and protective factors within the real-world data. This stands in contrast to traditional approaches, which often rely on highly curated datasets and domain expertise to build static prediction models based solely on well-known risk factors. As a proof-of-concept, we codified the full algorithm and tasked it with estimating 12-month risk of initial stroke or myocardial infarction using our hospital's real-world EHR. A 66-month pseudo-prospective validation was conducted using records from 558,105 patients spanning April 2015 to September 2023, totalling 3,424,060 patient-months. Area under the receiver operating characteristic curve (AUROC) values ranged from .830 to .909, with an improving trend over time. Odds ratios describing model precision for patients 1-100 and 101-200 (when ranked by estimated risk) ranged from 15.3 to 48.1 and 7.2 to 45.0, respectively, with both groups showing improving trends over time. Findings suggest the feasibility of developing high-performing disease risk calculators in the proposed manner.

Risk Assessment and Diagnosis Code Prediction from Electronic Health Records (EHR) using deep learning

Knowledge in the medical field is often expressed by distinct and subjective norms. Much research and development efforts have focused on using deep learning algorithms to predict the likelihood of illnesses from EHR in recent years. When it comes to risk prediction, deep learning-based techniques outperform more conventional machine learning models. But nothing in the literature fully accounts for what doctors already know, such as the connections between illnesses and their risk factors. This research examines the use of Multi-layer Perceptron models for the categorisation of diagnoses in electronic health records. The raw data and a modified version of the EHR dataset are used to train two MLPs with distinct topologies. For comparative purposes, a Random Forest is used as a baseline. To phenotype patients using their electronic health records, we provide a deep learning method. Predictive modelling of chronic illnesses is the particular scenario used to verify the suggested model on a real-world EHR data warehouse. Many deep learning applications on EHRs have been effective, and there is still a lot of potential to be realised. It has been discovered that deep learning models can learn from the limited EHR dataset, but not to a level where they outperform the baseline model.

Evaluation of score-based tertiary triage policies during the COVID-19 pandemic: simulation study with real-world intensive care data.

The explicit prohibition of discontinuing intensive care unit (ICU) treatment that has already begun by the newly established German Triage Act in favor of new patients with better prognoses (tertiary triage) under crisis conditions may prevent saving as many patients as possible and therefore may violate the international well-accepted premise of undertaking the "best for the most" patients. During the COVID-19 pandemic, authorities set up lockdown measures and infection-prevention strategies to avoid an overburdened health-care system. In cases of situational overload of ICU resources, when transporting options are exhausted, the question of atertiary triage of patients arises. We provide data-driven analyses of score- and non-score-based tertiary triage policies using simulation and real-world electronic health record data in aCOVID-19 setting. Ten different triage policies, for example, based on the Simplified Acute Physiology Score (SAPSII), are compared based on the resulting mortality in the ICU and inferential statistics. Our study shows that score-based tertiary triage policies outperform non-score-based tertiary triage policies including compliance with the German Triage Act. Based on our simulation model, aSAPSII score-based tertiary triage policy reduces mortality in the ICU by up to 18percentage points. The longer the queue of critical care patients waiting for ICU treatment and the larger the maximum number of patients subject to tertiary triage, the greater the effect on the reduction of mortality in the ICU. ASAPSII score-based tertiary triage policy was superior in our simulation model. Random allocation or "first come, first served" policies yield the lowest survival rates, as will adherence to the new German Triage Act. An interdisciplinary discussion including an ethical and legal perspective is important for the social interpretation of our data-driven results.

Semi-supervised Double Deep Learning Temporal Risk Prediction (SeDDLeR) with Electronic Health Records

Background:Risk prediction plays a crucial role in planning for prevention, monitoring, and treatment. Electronic Health Records (EHRs) offer an expansive repository of temporal medical data encompassing both risk factors and outcome indicators essential for effective risk prediction. However, challenges emerge due to the lack of readily available gold-standard outcomes and the complex effects of various risk factors. Compounding these challenges are the false positives in diagnosis codes, and formidable task of pinpointing the onset timing in annotations. Objective:We develop a Semi-supervised Double Deep Learning Temporal Risk Prediction (SeDDLeR) algorithm based on extensive unlabeled longitudinal Electronic Health Records (EHR) data augmented by a limited set of gold standard labels on the binary status information indicating whether the clinical event of interest occurred during the follow-up period. Methods:The SeDDLeR algorithm calculates an individualized risk of developing future clinical events over time using each patient’s baseline EHR features via the following steps: (1) construction of an initial EHR-derived surrogate as a proxy for the onset status; (2) deep learning calibration of the surrogate along gold-standard onset status; and (3) semi-supervised deep learning for risk prediction combining calibrated surrogates and gold-standard onset status. To account for missing onset time and heterogeneous follow-up, we introduce temporal kernel weighting. We devise a Gated Recurrent Units (GRUs) module to capture temporal characteristics. We subsequently assess our proposed SeDDLeR method in simulation studies and apply the method to the Massachusetts General Brigham (MGB) Biobank to predict type 2 diabetes (T2D) risk. Results:SeDDLeR outperforms benchmark risk prediction methods, including Semi-parametric Transformation Model (STM) and DeepHit, with consistently best accuracy across experiments. SeDDLeR achieved the best C-statistics ( 0.815, SE 0.023; vs STM +.084, SE 0.030, P-value .004; vs DeepHit +.055, SE 0.027, P-value .024) and best average time-specific AUC (0.778, SE 0.022; vs STM + 0.059, SE 0.039, P-value .067; vs DeepHit + 0.168, SE 0.032, P-value <0.001) in the MGB T2D study. Conclusion:SeDDLeR can train robust risk prediction models in both real-world EHR and synthetic datasets with minimal requirements of labeling event times. It holds the potential to be incorporated for future clinical trial recruitment or clinical decision-making.

DNA-T: Deformable Neighborhood Attention Transformer for Irregular Medical Time Series.

The real-world Electronic Health Records (EHRs) present irregularities due to changes in the patient's health status, resulting in various time intervals between observations and different physiological variables examined at each observation point. There have been recent applications of Transformer-based models in the field of irregular time series. However, the full attention mechanism in Transformer overly focuses on distant information, ignoring the short-term correlations of the condition. Thereby, the model is not able to capture localized changes or short-term fluctuations in patients' conditions. Therefore, we propose a novel end-to-end Deformable Neighborhood Attention Transformer (DNA-T) for irregular medical time series. The DNA-T captures local features by dynamically adjusting the receptive field of attention and aggregating relevant deformable neighborhoods in irregular time series. Specifically, we design a Deformable Neighborhood Attention (DNA) module that enables the network to attend to relevant neighborhoods by drifting the receiving field of neighborhood attention. The DNA enhances the model's sensitivity to local information and representation of local features, thereby capturing the correlation of localized changes in patients' conditions. We conduct extensive experiments to validate the effectiveness of DNA-T, outperforming existing state-of-the-art methods in predicting the mortality risk of patients. Moreover, we visualize an example to validate the effectiveness of the proposed DNA.

TEE4EHR: Transformer event encoder for better representation learning in electronic health records

Irregular sampling of time series in electronic health records (EHRs) is one of the main challenges for developing machine learning models. Additionally, the pattern of missing values in certain clinical variables is not at random but depends on the decisions of clinicians and the state of the patient. Point process is a mathematical framework for analyzing event sequence data consistent with irregular sampling patterns. Our model, TEE4EHR, is a transformer event encoder (TEE) with point process loss that encodes the pattern of laboratory tests in EHRs. The utility of our TEE has been investigated in various benchmark event sequence datasets. Additionally, we conduct experiments on two real-world EHR databases to provide a more comprehensive evaluation of our model. Firstly, in a self-supervised learning approach, the TEE is jointly learned with an existing attention-based deep neural network, which gives superior performance in negative log-likelihood and future event prediction. Besides, we propose an algorithm for aggregating attention weights to reveal the events’ interactions. Secondly, we transfer and freeze the learned TEE to the downstream task for the outcome prediction, where it outperforms state-of-the-art models for handling irregularly sampled time series. Furthermore, our results demonstrate that our approach can improve representation learning in EHRs and be useful for clinical prediction tasks.

OntoMedRec: Logically-pretrained model-agnostic ontology encoders for medication recommendation

Recommending medications with electronic health records (EHRs) is a challenging task for data-driven clinical decision support systems. Most existing models learnt representations for medical concepts based on EHRs and make recommendations with the learnt representations. However, most medications appear in EHR datasets for limited times (the frequency distribution of medications follows power law distribution), resulting in insufficient learning of their representations of the medications. Medical ontologies are the hierarchical classification systems for medical terms where similar terms will be in the same class on a certain level. In this paper, we propose OntoMedRec, the logically-pretrained and model-agnostic medical Ontology Encoders for Medication Recommendation that addresses data sparsity problem with medical ontologies. We conduct comprehensive experiments on real-world EHR datasets to evaluate the effectiveness of OntoMedRec by integrating it into various existing downstream medication recommendation models. The result shows the integration of OntoMedRec improves the performance of various models in both the entire EHR datasets and the admissions with few-shot medications. We provide the GitHub repository for the source code. (https://github.com/WaicongTam/OntoMedRec)

MixEHR-SurG: A joint proportional hazard and guided topic model for inferring mortality-associated topics from electronic health records

Survival models can help medical practitioners to evaluate the prognostic importance of clinical variables to patient outcomes such as mortality or hospital readmission and subsequently design personalized treatment regimes. Electronic Health Records (EHRs) hold the promise for large-scale survival analysis based on systematically recorded clinical features for each patient. However, existing survival models either do not scale to high dimensional and multi-modal EHR data or are difficult to interpret. In this study, we present a supervised topic model called MixEHR-SurG to simultaneously integrate heterogeneous EHR data and model survival hazard. Our contributions are three-folds: (1) integrating EHR topic inference with Cox proportional hazards likelihood; (2) integrating patient-specific topic hyperparameters using the PheCode concepts such that each topic can be identified with exactly one PheCode-associated phenotype; (3) multi-modal survival topic inference. This leads to a highly interpretable survival topic model that can infer PheCode-specific phenotype topics associated with patient mortality. We evaluated MixEHR-SurG using a simulated dataset and two real-world EHR datasets: the Quebec Congenital Heart Disease (CHD) data consisting of 8211 subjects with 75,187 outpatient claim records of 1767 unique ICD codes; the MIMIC-III consisting of 1458 subjects with multi-modal EHR records. Compared to the baselines, MixEHR-SurG achieved a superior dynamic AUROC for mortality prediction, with a mean AUROC score of 0.89 in the simulation dataset and a mean AUROC of 0.645 on the CHD dataset. Qualitatively, MixEHR-SurG associates severe cardiac conditions with high mortality risk among the CHD patients after the first heart failure hospitalization and critical brain injuries with increased mortality among the MIMIC-III patients after their ICU discharge. Together, the integration of the Cox proportional hazards model and EHR topic inference in MixEHR-SurG not only leads to competitive mortality prediction but also meaningful phenotype topics for in-depth survival analysis. The software is available at GitHub: https://github.com/li-lab-mcgill/MixEHR-SurG.

Design and implementation of electronic health record-based tools to support a weight management program in primary care.

This paper reports on a mixed methods formative evaluation to support the design and implementation of information technology (IT) tools for a primary care weight management intervention delivered through the patient portal using primary care staff as coaches. We performed a qualitative needs assessment, designed the IT tools to support the weight management program, and developed implementation tracking metrics. Implementation tracking metrics were designed to use real world electronic health record (EHR) data. The needs assessment revealed IT requirements as well as barriers and facilitators to implementation of EHR-based weight management interventions in primary care. We developed implementation metrics for the IT tools. These metrics were used in weekly project team calls to make sure that project resources were allocated to areas of need. This study identifies the important role of IT in supporting weight management through patient identification, weight and activity tracking in the patient portal, and the use of the EHR as a population management tool. An intensive multi-level implementation approach is required for successful primary care-based weight management interventions including well-designed IT tools, comprehensive involvement of clinic leadership, and implementation tracking metrics to guide the process of workflow integration. This study helps to bridge the gap between informatics and implementation by using socio-technical formative evaluation methods early in order to support the implementation of IT tools. clinicaltrials.gov, NCT04420936. Registered June 9, 2020.

Analyzing the Impact of Personalization on Fairness in Federated Learning for Healthcare.

As machine learning (ML) usage becomes more popular in the healthcare sector, there are also increasing concerns about potential biases and risks such as privacy. One countermeasure is to use federated learning (FL) to support collaborative learning without the need for patient data sharing across different organizations. However, the inherent heterogeneity of data distributions among participating FL parties poses challenges for exploring group fairness in FL. While personalization within FL can handle performance degradation caused by data heterogeneity, its influence on group fairness is not fully investigated. Therefore, the primary focus of this study is to rigorously assess the impact of personalized FL on group fairness in the healthcare domain, offering a comprehensive understanding of how personalized FL affects group fairness in clinical outcomes. We conduct an empirical analysis using two prominent real-world Electronic Health Records (EHR) datasets, namely eICU and MIMIC-IV. Our methodology involves a thorough comparison between personalized FL and two baselines: standalone training, where models are developed independently without FL collaboration, and standard FL, which aims to learn a global model via the FedAvg algorithm. We adopt Ditto as our personalized FL approach, which enables each client in FL to develop its own personalized model through multi-task learning. Our assessment is achieved through a series of evaluations, comparing the predictive performance (i.e., AUROC and AUPRC) and fairness gaps (i.e., EOPP, EOD, and DP) of these methods. Personalized FL demonstrates superior predictive accuracy and fairness over standalone training across both datasets. Nevertheless, in comparison with standard FL, personalized FL shows improved predictive accuracy but does not consistently offer better fairness outcomes. For instance, in the 24-h in-hospital mortality prediction task, personalized FL achieves an average EOD of 27.4% across racial groups in the eICU dataset and 47.8% in MIMIC-IV. In comparison, standard FL records a better EOD of 26.2% for eICU and 42.0% for MIMIC-IV, while standalone training yields significantly worse EOD of 69.4% and 54.7% on these datasets, respectively. Our analysis reveals that personalized FL has the potential to enhance fairness in comparison to standalone training, yet it does not consistently ensure fairness improvements compared to standard FL. Our findings also show that while personalization can improve fairness for more biased hospitals (i.e., hospitals having larger fairness gaps in standalone training), it can exacerbate fairness issues for less biased ones. These insights suggest that the integration of personalized FL with additional strategic designs could be key to simultaneously boosting prediction accuracy and reducing fairness disparities. The findings and opportunities outlined in this paper can inform the research agenda for future studies, to overcome the limitations and further advance health equity research.

Abstract 6467: Causal personalized treatment estimation framework using real-world electronic health record (EHR) data to inform cancer care decisions

Abstract Background: The average treatment effects (ATEs) from randomized controlled trials often do not generalize across populations, or they are not sufficiently personalized to capture the benefits and risks to individual patients. Real-world data aids in validating and extending trial results, but can introduce biases that invalidate the causal interpretation. Here, we propose a robust statistical framework, which uses real-world EHR data to evaluate the efficacy of cancer treatments at both population and individual levels. We demonstrate the ability of our framework to deliver personalized insights in metastatic prostate cancer treatments. Methods: We assembled a retrospective cohort of metastatic prostate cancer patients from a large academic medical center. We gathered demographic, comorbidity, cancer treatment, and outcome data from the EHR or natural language processing of clinician narratives, and social determinants of health from Census data. We used doubly robust machine learning with bootstrap standard errors to compute the ATE and personalized treatment estimates (PTEs), comparing the impact of ADT (androgen depravation therapy) + chemotherapy (ACT) and ADT + combination anti-androgen therapy (ACAT) on percentage reduction in PSA levels in metastatic prostate cancer. Results: The baseline characteristics of our cohort are summarized in Table 1a. Our approach found that ACAT reduces PSA levels significantly more than ACT, and captured the heterogeneity in treatment effects across individuals (Table 1b). Through simulation studies, we demonstrated that any reasonable bias from unobserved confounding or missing information was small compared to the estimated treatment effects and their standard errors. Conclusion: We presented a doubly robust framework and demonstrated its utility in analyzing real-world data, enabling us to uncover the diversity in treatment responses among individuals with metastatic prostate cancer. Results summary a. Baseline demographic characteristics. ADT + chemotherapy N = 302 ADT + combination therapy N = 114 Treatments Any ADT + Docetaxel or Cabazitaxel Any ADT + Abiraterone or apalutamide or Enzalutamide Age Mean (SD) 69.3 (8.76) 66.7 (8.17) Median [Min, Max] 69.0 [47.0, 94.0] 66.5 [46.0, 86.0] Race Asian 8 (2.6%) 2 (1.8%) Black or African American 48 (15.9%) 7 (6.1%) White or Caucasian 224 (74.2%) 99 (86.8%) Other 22 (7.3%) 6 (5.3%) Max PSA treatment start (ng/mL) Mean (SD) 34.9 (98.5) 83.3 (448) b. Treatment effects estimation ATE (95% CI) 32.0% [24.6%, 40.8%] PTEs: 25th and 75th percentiles 29.0%, 34.3% Citation Format: Sumana Srivatsa, David Walsh, Mayada Aljehani, Jason Weinreb, Nathan Becker, Benjamin H. Ellis, Renee George, Xingyao Chen, Naim Matasci, Reva Basho, Mitchell E. Gross, Christine L. Swisher. Causal personalized treatment estimation framework using real-world electronic health record (EHR) data to inform cancer care decisions [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6467.

Multi-institutional Assessment of Otitis Media Epidemiology Using Real-world Data

ObjectivesTo determine rates and risk factors of pediatric otitis media (OM) using real-world electronic health record (PEDSnet) data from January 2009 through May 2021. Study designRetrospective cohort study. SettingSeven pediatric academic health systems that participate in PEDSnet. MethodsChildren <6 months-old at time of first outpatient, Emergency Department, or inpatient visit were included and followed longitudinally. A time-to-event analysis was performed using a Cox proportional hazards model to estimate hazard ratios for OM incidence based on sociodemographic factors and specific health conditions. ResultsThe PEDSnet cohort included 910,265 children, 54.3% male, mean age (months) 1.3 [standard deviation (SD) 1.6] and mean follow up (years) 4.3 (SD 3.2). By age 3 years, 39.6% of children had evidence of one OM episode. OM rates decreased following pneumococcal-13 vaccination (PCV-13) and the COVID-19 pandemic. Along with young age, non-Hispanic Black/African American or Hispanic race/ethnicity, public insurance, higher family income, and male sex had higher incidence rates. Health conditions that increased OM risk included cleft palate [adjusted hazard ratio (aHR) 4.0 [95% confidence interval (CI) 3.9–4.2], primary ciliary dyskinesia [aHR 2.5 (95% CI 1.8–3.5)], trisomy 21 [aHR 2.0 (95% CI 1.9–2.1)], atopic dermatitis [aHR 1.4 (95% CI 1.4–1.4)], and gastroesophageal reflux [aHR1.5 (95% CI 1.5–1.5)]. ConclusionsApproximately 20% of children by age 1 and 40% of children by age 3 years will have experienced an OM episode. OM rates decreased after PCV-13 and COVID-19. Children with abnormal ciliary function or craniofacial conditions, specifically cleft palate, carry the highest risk of OM.

Comparison of late-onset and non-late-onset systemic lupus erythematosus individuals in a real-world electronic health record cohort.

Objective: Late-onset systemic lupus erythematosus (LO-SLE) is defined as SLE diagnosed at age 50 years or later. Current studies on LO-SLE are small and have conflicting results.Methods: Using a large, electronic health record (EHR)-based cohort of SLE individuals, we compared demographics, disease characteristics, SLE-specific antibodies, and medication prescribing practices in LO (n = 123) vs. NLO-SLE (n = 402) individuals.Results: The median age (interquartile range) at SLE diagnosis was 60 (56-67) years for LO-SLE and 28 (20-38) years for NLO-SLE. Both groups were predominantly female (85% vs. 91%, p = 0.10). LO-SLE individuals were more likely to be White than NLO-SLE individuals (74% vs. 60%, p = 0.005) and less likely to have positive dsDNA (39% vs. 58%, p = 0.001) and RNP (17% vs. 32%, p = 0.02) with no differences in Smith, SSA, and SSB. Autoantibody positivity declined with increasing age at SLE diagnosis. LO-SLE individuals were less likely to develop SLE nephritis (9% vs. 29%, p < 0.001) and less likely to be prescribed multiple classes of SLE medications including antimalarials (90% vs. 95%, p = 0.04), azathioprine (17% vs. 31%, p = 0.002), mycophenolate mofetil (12% vs. 38%, p < 0.001), and belimumab (2% vs. 8%, p = 0.02).Conclusion: LO-SLE individuals may be less likely to fit an expected course for SLE with less frequent positive autoantibodies at diagnosis and lower rates of nephritis, even after adjusting for race. Understanding how age impacts SLE disease presentation could help reduce diagnostic delays in SLE.

Prediction of Effectiveness and Toxicities of Immune Checkpoint Inhibitors Using Real-World Patient Data.

Although immune checkpoint inhibitors (ICIs) have improved outcomes in certain patients with cancer, they can also cause life-threatening immunotoxicities. Predicting immunotoxicity risks alongside response could provide a personalized risk-benefit profile, inform therapeutic decision making, and improve clinical trial cohort selection. We aimed to build a machine learning (ML) framework using routine electronic health record (EHR) data to predict hepatitis, colitis, pneumonitis, and 1-year overall survival. Real-world EHR data of more than 2,200 patients treated with ICI through December 31, 2018, were used to develop predictive models. Using a prediction time point of ICI initiation, a 1-year prediction time window was applied to create binary labels for the four outcomes for each patient. Feature engineering involved aggregating laboratory measurements over appropriate time windows (60-365 days). Patients were randomly partitioned into training (80%) and test (20%) sets. Random forest classifiers were developed using a rigorous model development framework. The patient cohort had a median age of 63 years and was 61.8% male. Patients predominantly had melanoma (37.8%), lung cancer (27.3%), or genitourinary cancer (16.4%). They were treated with PD-1 (60.4%), PD-L1 (9.0%), and CTLA-4 (19.7%) ICIs. Our models demonstrate reasonably strong performance, with AUCs of 0.739, 0.729, 0.755, and 0.752 for the pneumonitis, hepatitis, colitis, and 1-year overall survival models, respectively. Each model relies on an outcome-specific feature set, though some features are shared among models. To our knowledge, this is the first ML solution that assesses individual ICI risk-benefit profiles based predominantly on routine structured EHR data. As such, use of our ML solution will not require additional data collection or documentation in the clinic.

Enhancing Disease Prediction with a Hybrid CNN-LSTM Framework in EHRs

This study developed a novel hybrid deep learning framework aimed at enhancing the accuracy of disease prediction using temporal data from Electronic Health Records (EHRs). The framework integrates Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks, leveraging the strength of CNNs in extracting hierarchical feature representations from complex data and the capability of LSTMs in capturing long-term dependencies in temporal information. An empirical investigation on real-world EHR datasets revealed that, compared to Support Vector Machine (SVM) models, standalone CNNs, and LSTMs, this hybrid deep learning network demonstrated significantly higher prediction accuracy in disease prediction tasks. This research not only advances the performance of predictive models in the health data analytics domain but also underscores the importance of adopting and further developing advanced deep learning technologies to address the complexity of modern medical data. Our findings advocate for a shift towards integrating complex neural network architectures in developing predictive models, potentially offering avenues for more personalized and proactive disease management and care, thereby setting new standards for future health management practices.

COVID-19 Distribution in Pregnancy, Drug Use Patterns and COVID-19 Medication during the Pandemic in Spain: Data from Real-World Electronic Health Records.

Although pregnant women were considered a risk population for COVID-19, little is known of their drug use during the pandemic. We aimed to investigate COVID-19 distribution, drug use patterns and COVID-19 medication. We conducted a retrospective cohort of validated pregnancies aged 15-49 years, from January 2020 to December 2022, using the BIFAP database. An identified cohort of pregnant women with COVID-19 was matched by age, gestational age, length of pregnancy and outcome to a cohort free of COVID-19 (8413 vs. 24,975). We performed a descriptive analysis on COVID-19 cases, estimated the drug use patterns and assessed COVID-19-specific drugs within the week prior/after diagnosis, stratified by pandemic wave and gestational week. The results showed that 72% of pregnant women with COVID-19 received at least one prescription vs. 66.6% of those free of COVID-19, with analgesics, antibiotics and thyroid hormones being the most prescribed drugs in both groups. In the COVID-19 group, they were antithrombotics (40 prescriptions per 100 women), analgesic/NSAIDs (19.64/6.29) and antibiotics (6.95). COVID-19 cases gradually increased, peaking at the fifth and second waves. Prescription rates were similar when compared to pre-pandemic studies. The use of drugs compatible with COVID-19 treatments was in line with recommendations.

A deep clustering-based state-space model for improved disease risk prediction in personalized healthcare

AbstractDecision support systems are being developed to assist clinicians in complex decision-making processes by leveraging information from clinical knowledge and electronic health records (EHRs). One typical application is disease risk prediction, which can be challenging due to the complexity of modelling longitudinal EHR data, including unstructured medical notes. To address this challenge, we propose a deep state-space model (DSSM) that simulates the patient’s state transition process and formally integrates latent states with risk observations. A typical DSSM consists of three parts: a prior module that generates the distribution of the current latent state based on previous states; a posterior module that approximates the latent states using up-to-date medical notes; and a likelihood module that predicts disease risks using latent states. To efficiently and effectively encode raw medical notes, our posterior module uses an attentive encoder to better extract information from unstructured high-dimensional medical notes. Additionally, we couple a predictive clustering algorithm into our DSSM to learn clinically useful representations of patients’ latent states. The latent states are clustered into multiple groups, and the weighted average of the cluster centres is used for prediction. We demonstrate the effectiveness of our deep clustering-based state-space model using two real-world EHR datasets, showing that it not only generates better risk prediction results than other baseline methods but also clusters similar patient health states into groups.

A Post-Marketing Drug Evaluation Framework Based on Real-World Electronic Health Records Data.

Real-world data (RWD) could be a new way to evaluate the safety and efficacy of post-marketing drugs, while there is no common method for how to use RWD for drug evaluation. In this paper, we present a framework for real-world drug evaluation based on electronic medical record (EHR) data. We designed a data model customized for post-marketing drug evaluation and a unified post-marketing drug evaluation pipeline. The proposed framework can be applied to drug evaluations with different study paradigms for different purposes by flexible use of the proposed data model and pipeline. A prototype system has been developed according to the framework. Real-world EHRs in a tertiary hospital in China between 2010 and 2020 were converted to the proposed data model, and as a test case, we conducted a research on the risk of allergic reactions to cefodizime and ceftriaxone using the prototype system.

EHR-KnowGen: Knowledge-enhanced multimodal learning for disease diagnosis generation

Electronic health records (EHRs) contain diverse patient information, including medical notes, clinical events, and laboratory test results. Integrating this multimodal data can improve disease diagnoses using deep learning models. However, effectively combining different modalities for diagnosis remains challenging. Previous approaches, such as attention mechanisms and contrastive learning, have attempted to address this but do not fully integrate the modalities into a unified feature space. This paper presents EHR-KnowGen, a multimodal learning model enhanced with external domain knowledge, for improved disease diagnosis generation from diverse patient information in EHRs. Unlike previous approaches, our model integrates different modalities into a unified feature space with soft prompts learning and leverages large language models (LLMs) to generate disease diagnoses. By incorporating external domain knowledge from different levels of granularity, we enhance the extraction and fusion of multimodal information, resulting in more accurate diagnosis generation. Experimental results on real-world EHR datasets demonstrate the superiority of our generative model over comparative methods, providing explainable evidence to enhance the understanding of diagnosis results.

Real-World Genomic Profile of EGFR Second-Site Mutations and Other Osimertinib Resistance Mechanisms and Clinical Landscape of NSCLC Post-Osimertinib

IntroductionThe emergence of osimertinib as standard of care for EGFR-mutant NSCLC has renewed the need to understand and overcome drug resistance. We sought to understand the genomics and real-world treatment landscape of NSCLC with EGFR C797S and other on- and off-target resistance mechanisms. MethodsComprehensive genomic profiling (CGP) results from tissue or blood samples from 93,065 patients with NSCLC were queried for osimertinib EGFR second-site resistance mutations (ssEGFRms; C797, L718, G724, G796, L792). A real-world electronic health record–derived deidentified clinicogenomic database of patients with NSCLC undergoing CGP from approximately 280 U.S. cancer clinics was queried to assess post-osimertinib resistance and clinical treatment outcomes. ResultsA ssEGFRm was identified in 239 of 8845 (2.7%) EGFR-driven (L858R or exon 19 deletion) NSCLCs, most frequently C797 (71%), L718 (15%), and G724 (9.5%). ssEGFRms were not equally distributed across drivers; C797 and G724 changes strongly favored exon 19 deletion and L718, G796 and L792 favored L858R. Post-osimertinib CGP detected ssEGFRm in 19% of the cases (39 of 205); in paired pre-/post-osimertinib samples, on- and off-target resistance was largely mutually exclusive and observed in 24% and 27% of the cases, respectively. Of 391 patients with post-osimertinib treatment data, 62% received a chemotherapy-based regimen, whereas 25% received a targeted therapy or clinical study drug. Median real-world overall survival was 11.4 months from osimertinib progression. ConclusionsThe osimertinib resistance landscape is diverse with on-target ssEGFRm and off-target resistance detected in tissue and liquid biopsy. Post-osimertinib, patients are receiving primarily chemotherapy-based regimens with poor outcomes, and CGP at resistance may offer an opportunity to inform therapeutic development and improve treatment selection.