Journal of the American Medical Informatics Association最新文献

Objectives: Report the development of the patient-centered myAURA application and suite of methods designed to aid epilepsy patients, caregivers, and clinicians in making decisions about self-management and care.

Materials and methods: myAURA rests on an unprecedented collection of epilepsy-relevant heterogeneous data resources, such as biomedical databases, social media, and electronic health records (EHRs). We use a patient-centered biomedical dictionary to link the collected data in a multilayer knowledge graph (KG) computed with a generalizable, open-source methodology.

Results: Our approach is based on a novel network sparsification method that uses the metric backbone of weighted graphs to discover important edges for inference, recommendation, and visualization. We demonstrate by studying drug-drug interaction from EHRs, extracting epilepsy-focused digital cohorts from social media, and generating a multilayer KG visualization. We also present our patient-centered design and pilot-testing of myAURA, including its user interface.

Discussion: The ability to search and explore myAURA's heterogeneous data sources in a single, sparsified, multilayer KG is highly useful for a range of epilepsy studies and stakeholder support.

Conclusion: Our stakeholder-driven, scalable approach to integrating traditional and nontraditional data sources enables both clinical discovery and data-powered patient self-management in epilepsy and can be generalized to other chronic conditions.

Objective: To describe the prevalence of and trends in practices that interfere with the exchange of patient health information (potential information blocking) 2 years after implementation of information-blocking regulations.

Materials and methods: Drawing from the American Hospital Association Information Technology (IT) Supplement and a national survey of health information organizations (HIOs), we described rates and methods of potential information blocking from these organizations' perspectives in 2023 and compared them to prior years.

Results: Twenty-seven percent of hospitals sometimes or often observed potential information blocking by any actor in 2023, down from 42% in 2021 and 33% in 2022. Thirty percent of HIOs routinely observed potential information blocking by health IT developers, down from 50% in 2015. 13% of HIOs routinely observed potential information blocking by hospitals and health systems, down from 25% in 2015. According to both hospitals and HIOs, the most prevalent method of potential information blocking by developers in 2023 was through price, while the most prevalent by healthcare providers/health systems was by focusing exchange on strategic affiliations. Few hospitals and HIOs that experienced potential information blocking said that they had reported it to the Department of Health and Human Services.

Discussion: Hospitals and HIOs perceived lower rates of potential information blocking in 2023 than in prior years indicating some impact of regulations addressing information blocking. However, both respondent types reported that substantial potential information blocking persisted in 2023 and negatively impacted the exchange of information.

Conclusion: While potential information-blocking practices have decreased, they have not been eliminated, indicating the value of continued and robust enforcement of information-blocking regulations.

Objective: Prediction of mortality in intensive care unit (ICU) patients typically relies on black box models (that are unacceptable for use in hospitals) or hand-tuned interpretable models (that might lead to the loss in performance). We aim to bridge the gap between these 2 categories by building on modern interpretable machine learning (ML) techniques to design interpretable mortality risk scores that are as accurate as black boxes.

Material and methods: We developed a new algorithm, GroupFasterRisk, which has several important benefits: it uses both hard and soft direct sparsity regularization, it incorporates group sparsity to allow more cohesive models, it allows for monotonicity constraint to include domain knowledge, and it produces many equally good models, which allows domain experts to choose among them. For evaluation, we leveraged the largest existing public ICU monitoring datasets (MIMIC III and eICU).

Results: Models produced by GroupFasterRisk outperformed OASIS and SAPS II scores and performed similarly to APACHE IV/IVa while using at most a third of the parameters. For patients with sepsis/septicemia, acute myocardial infarction, heart failure, and acute kidney failure, GroupFasterRisk models outperformed OASIS and SOFA. Finally, different mortality prediction ML approaches performed better based on variables selected by GroupFasterRisk as compared to OASIS variables.

Discussion: Group Faster Risk's models performed better than risk scores currently used in hospitals, and on par with black box ML models, while being orders of magnitude sparser. Because GroupFasterRisk produces a variety of risk scores, it allows design flexibility-the key enabler of practical model creation.

Conclusion: Group Faster Risk is a fast, accessible, and flexible procedure that allows learning a diverse set of sparse risk scores for mortality prediction.

Objectives: Explanations help to understand why anomaly detection algorithms identify data as anomalous. This study evaluates whether robustly standardized explanation scores correctly identify the implausible variables that make cancer data anomalous.

Materials and methods: The dataset analyzed consists of 18 587 truncated real-world cancer registry records containing 8 categorical variables describing patients diagnosed with bladder and lung tumors. We identified 800 anomalous records using an autoencoder's per-record reconstruction error, which is a common neural network-based anomaly detection approach. For each variable of a record, we determined a robust explanation score, which indicates how anomalous the variable is. A variable's robust explanation score is the autoencoder's per-variable reconstruction error measured by cross-entropy and robustly standardized across records; that is, large reconstruction errors have a small effect on standardization. To evaluate the explanation scores, medical coders identified the implausible variables of the anomalous records. We then compare the explanation scores to the medical coders' validation in a classification and ranking setting. As baselines, we identified anomalous variables using the raw autoencoder's per-variable reconstruction error, the non-robustly standardized per-variable reconstruction error, the empirical frequency of implausible variables according to the medical coders' validation, and random selection or ranking of variables.

Results: When we sort the variables by their robust explanation scores, on average, the 2.37 highest-ranked variables contain all implausible variables. For the baselines, on average, the 2.84, 2.98, 3.27, and 4.91 highest-ranked variables contain all the variables that made a record implausible.

Discussion: We found that explanations based on robust explanation scores were better than or as good as the baseline explanations examined in the classification and ranking settings. Due to the international standardization of cancer data coding, we expect our results to generalize to other cancer types and registries. As we anticipate different magnitudes of per-variable autoencoder reconstruction errors in data from other medical registries and domains, these may also benefit from robustly standardizing the reconstruction errors per variable. Future work could explore methods to identify subsets of anomalous variables, addressing whether individual variables or their combinations contribute to anomalies. This direction aims to improve the interpretability and utility of anomaly detection systems.

Conclusions: Robust explanation scores can improve explanations for identifying implausible variables in cancer data.

Objective: To identify distinct patterns in consumer willingness to share health data with various stakeholders and analyze characteristics across consumer groups.

Materials and methods: Data from the Rock Health Digital Health Consumer Adoption Survey from 2018, 2019, 2020, and 2022 were analyzed. This study comprised a Census-matched representative sample of U.S. adults. Latent class analysis (LCA) identified groups of respondents with similar data-sharing attitudes. Groups were compared by sociodemographics, health status, and digital health utilization.

Results: We identified three distinct LCA groups: (1) Wary (36.8%), (2) Discerning (47.9%), and (3) Permissive (15.3%). The Wary subgroup exhibited reluctance to share health data with any stakeholder, with predicted probabilities of willingness to share ranging from 0.07 for pharmaceutical companies to 0.34 for doctors/clinicians. The Permissive group showed a high willingness, with predicted probabilities greater than 0.75 for most stakeholders except technology companies and government organizations. The Discerning group was selective, willing to share with healthcare-related entities and family (predicted probabilities >0.62), but reluctant to share with other stakeholders (predicted probabilities <0.29). Individual characteristics were associated with LCA group membership.

Discussion: Findings highlight a persistent trust in traditional healthcare providers. However, the varying willingness to share with non-traditional stakeholders suggests that while some consumers are open to sharing, others remain hesitant and selective. Data privacy policies and practices need to recognize and respond to multifaceted and stakeholder-specific attitudes.

Conclusion: LCA reveals significant heterogeneity in health data-sharing attitudes among U.S. consumers, providing insights to inform the development of data privacy policies.

Objective: To develop a distributed algorithm to fit multi-center Cox regression models with time-varying coefficients to facilitate privacy-preserving data integration across multiple health systems.

Materials and methods: The Cox model with time-varying coefficients relaxes the proportional hazards assumption of the usual Cox model and is particularly useful to model time-to-event outcomes. We proposed a One-shot Distributed Algorithm to fit multi-center Cox regression models with Time varying coefficients (ODACT). This algorithm constructed a surrogate likelihood function to approximate the Cox partial likelihood function, using patient-level data from a lead site and aggregated data from other sites. The performance of ODACT was demonstrated by simulation and a real-world study of opioid use disorder (OUD) using decentralized data from a large clinical research network across 5 sites with 69 163 subjects.

Results: The ODACT method precisely estimated the time-varying effects over time. In the simulation study, ODACT always achieved estimation close to that of the pooled analysis, while the meta-estimator showed considerable amount of bias. In the OUD study, the bias of the estimated hazard ratios by ODACT are smaller than those of the meta-estimator for all 7 risk factors at almost all of the time points from 0 to 2.5 years. The greatest bias of the meta-estimator was for the effects of age ≥65 years, and smoking.

Conclusion: ODACT is a privacy-preserving and communication-efficient method for analyzing multi-center time-to-event data which allows the covariates' effects to be time-varying. ODACT provides estimates close to the pooled estimator and substantially outperforms the meta-analysis estimator.

Discussion: The proposed ODACT is a privacy-preserving distributed algorithm for fitting Cox models with time-varying coefficients. The limitations of ODACT include that privacy-preserving via aggregate data does rely on relatively large number of data at each individual site, and rigorous quantification of the risk of privacy leaks requires further investigation.

Objective: We aimed to develop a highly interpretable and effective, machine-learning based risk prediction algorithm to predict in-hospital mortality, intubation and adverse cardiovascular events in patients hospitalised with COVID-19 in Australia (AUS-COVID Score).

Materials and methods: This prospective study across 21 hospitals included 1714 consecutive patients aged ≥ 18 in their index hospitalization with COVID-19. The dataset was separated into training (80%) and test sets (20%). Eight supervised ML methods were used: LASSO, ridge, elastic net (EN), decision tree, support vector machine, random forest, AdaBoost and gradient boosting. A feature selection method was used to establish informative variables, which were considered in groups of 5/10/15/20/all. The final model was selected by balancing the optimal area under the curve (AUC) score with interpretability, through the number of included variables. The coefficients of the final models were used to build the AUS-COVID Score.

Results & discussion: Among the patients, 181 (10.6%) died in-hospital, 148 (8.6%) required intubation and 90 (5.3%) had adverse cardiovascular events. The LASSO model performed best for predicting in-hospital mortality (AUC 0.85) using five variables: age, respiratory rate, COVID-19 features on chest X-ray (CXR), troponin elevation, and COVID-19 vaccination (≥1 dose). The Elastic Net model performed best for predicting intubation (AUC 0.75) and adverse cardiovascular events (AUC 0.64), each with five variables. A user-friendly web-based application was built for clinician use at the bedside.

Conclusion: The AUS-COVID Score is an accurate and practical, machine-learning-based risk score to predict in-hospital mortality, intubation, and adverse cardiovascular events in hospitalized COVID-19 patients.

Objective: A proof-of-concept study aimed at designing and implementing Visual & Interactive Engagement With Electronic Records (VIEWER), a versatile toolkit for visual analytics of clinical data, and systematically evaluating its effectiveness across various clinical applications while gathering feedback for iterative improvements.

Materials and methods: VIEWER is an open-source and extensible toolkit that employs natural language processing and interactive visualization techniques to facilitate the rapid design, development, and deployment of clinical information retrieval, analysis, and visualization at the point of care. Through an iterative and collaborative participatory design approach, VIEWER was designed and implemented in one of the United Kingdom's largest National Health Services mental health Trusts, where its clinical utility and effectiveness were assessed using both quantitative and qualitative methods.

Results: VIEWER provides interactive, problem-focused, and comprehensive views of longitudinal patient data (n = 409 870) from a combination of structured clinical data and unstructured clinical notes. Despite a relatively short adoption period and users' initial unfamiliarity, VIEWER significantly improved performance and task completion speed compared to the standard clinical information system. More than 1000 users and partners in the hospital tested and used VIEWER, reporting high satisfaction and expressed strong interest in incorporating VIEWER into their daily practice.

Discussion: VIEWER provides a cost-effective enhancement to the functionalities of standard clinical information systems, with evaluation offering valuable feedback for future improvements.

Conclusion: VIEWER was developed to improve data accessibility and representation across various aspects of healthcare delivery, including population health management and patient monitoring. The deployment of VIEWER highlights the benefits of collaborative refinement in optimizing health informatics solutions for enhanced patient care.