Abstract Clinical notes in Electronic Health Records (EHRs) capture rich temporal information on events, clinician reasoning, and lifestyle factors often missing from structured data. Leveraging them for predictive modeling can be impactful for timely identification of chronic diseases. However, they present core natural language processing (NLP) challenges: long text, irregular event distribution, complex temporal dependencies, privacy constraints, and resource limitations. We present two complementary methods for temporally and contextually grounded risk prediction from longitudinal notes. First, we introduce HITGNN, a hierarchical temporal graph neural network that integrates intranote temporal event structures, inter-visit dynamics, and medical knowledge to model patient trajectories with fine-grained temporal granularity. Second, we propose REVEAL, a lightweight test-time framework that distills LLMs’ reasoning into smaller verifier models. Applied to opportunistic screening for Type 2 Diabetes (T2D) using temporally realistic cohorts curated from private and public hospital corpora, HITGNN achieves the highest predictive accuracy—especially for near-term risk—while preserving privacy and limiting reliance on large proprietary models. REVEAL enhances sensitivity to true T2D cases and retains explanatory reasoning. Our ablations confirm the value of temporal structure and knowledge augmentation, and fairness analysis shows HITGNN performs more equitably across subgroups.

Type 2 diabetes (T2D) is a major public health problem, and opportunistic screening to detect T2D at an early stage can help initiate interventions that delay or prevent the disease and its complications. In this study, we use electronic health records (EHR) and concepts extracted from clinical notes to predict future T2D risk. Our deep neural network-based model captures the temporal sequence of patient visits. We use explainable AI algorithms to assess the model decisions and observe alignment with the domain knowledge of clinical experts.

Embedded pragmatic clinical trials (ePCTs) play a vital role in addressing current population health problems, and their use of electronic health record (EHR) systems promises efficiencies that will increase the speed and volume of relevant and generalizable research. However, as the number of ePCTs using EHR-derived data grows, so does the risk that research will become more vulnerable to biases due to differences in data capture and access to care for different subsets of the population, thereby propagating inequities in health and the healthcare system. We identify 3 challenges-incomplete and variable capture of data on social determinants of health, lack of representation of vulnerable populations that do not access or receive treatment, and data loss due to variable use of technology-that exacerbate bias when working with EHR data and offer recommendations and examples of ways to actively mitigate bias.

OBJECTIVE: To determine factors that influence the adoption and use of patient-reported outcomes (PROs) in the electronic health record (EHR) among users. MATERIALS AND METHODS: Q methodology, supported by focus groups, semistructured interviews, and a review of the literature was used for data collection about opinions on PROs in the EHR. An iterative thematic analysis resulted in 49 statements that study participants sorted, from most unimportant to most important, under the following condition of instruction: "What issues are most important or most unimportant to you when you think about the adoption and use of patient-reported outcomes within the electronic health record in routine clinical care?" Using purposive sampling, 50 participants were recruited to rank and sort the 49 statements online, using HTMLQ software. Principal component analysis and Varimax rotation were used for data analysis using the PQMethod software. RESULTS: Participants were mostly physicians (24%) or physician/researchers (20%). Eight factors were identified. Factors included the ability of PROs in the EHR to enable: efficient and reliable use; care process improvement and accountability; effective and better symptom assessment; patient involvement for care quality; actionable and practical clinical decisions; graphical review and interpretation of results; use for holistic care planning to reflect patients' needs; and seamless use for all users. DISCUSSION: The success of PROs in the EHR in clinical settings is not dependent on a "one size fits all" strategy, demonstrated by the diversity of viewpoints identified in this study. A sociotechnical approach for implementing PROs in the EHR may help improve its success and sustainability. CONCLUSIONS: PROs in the EHR are most important to users when the technology is used to improve patient outcomes. Future research must focus on the impact of embedding this EHR functionality on care processes.

Diabetes intensive care unit (ICU) patients are at increased risk of complications leading to in-hospital mortality. Assessing the likelihood of death is a challenging and time-consuming task due to a large number of influencing factors. Healthcare providers are interested in the detection of ICU patients at higher risk, such that risk factors can possibly be mitigated. While such severity scoring methods exist, they are commonly based on a snapshot of the health conditions of a patient during the ICU stay and do not specifically consider a patient's prior medical history. In this paper, a process mining/deep learning architecture is proposed to improve established severity scoring methods by incorporating the medical history of diabetes patients. First, health records of past hospital encounters are converted to event logs suitable for process mining. The event logs are then used to discover a process model that describes the past hospital encounters of patients. An adaptation of Decay Replay Mining is proposed to combine medical and demographic information with established severity scores to predict the in-hospital mortality of diabetes ICU patients. Significant performance improvements are demonstrated compared to established risk severity scoring methods and machine learning approaches using the Medical Information Mart for Intensive Care III dataset.

OBJECTIVE: We identified challenges and solutions to using electronic health record (EHR) systems for the design and conduct of pragmatic research. MATERIALS AND METHODS: Since 2012, the Health Care Systems Research Collaboratory has served as the resource coordinating center for 21 pragmatic clinical trial demonstration projects. The EHR Core working group invited these demonstration projects to complete a written semistructured survey and used an inductive approach to review responses and identify EHR-related challenges and suggested EHR enhancements. RESULTS: We received survey responses from 20 projects and identified 21 challenges that fell into 6 broad themes: (1) inadequate collection of patient-reported outcome data, (2) lack of structured data collection, (3) data standardization, (4) resources to support customization of EHRs, (5) difficulties aggregating data across sites, and (6) accessing EHR data. DISCUSSION: Based on these findings, we formulated 6 prerequisites for PCTs that would enable the conduct of pragmatic research: (1) integrate the collection of patient-centered data into EHR systems, (2) facilitate structured research data collection by leveraging standard EHR functions, usable interfaces, and standard workflows, (3) support the creation of high-quality research data by using standards, (4) ensure adequate IT staff to support embedded research, (5) create aggregate, multidata type resources for multisite trials, and (6) create re-usable and automated queries. CONCLUSION: We are hopeful our collection of specific EHR challenges and research needs will drive health system leaders, policymakers, and EHR designers to support these suggestions to improve our national capacity for generating real-world evidence.

Sabita Acharya, Barbara Di Eugenio, Andrew Boyd, Richard Cameron, Karen Dunn Lopez, Pamela Martyn-Nemeth, Debaleena Chattopadhyay, Pantea Habibi, Carolyn Dickens, Haleh Vatani, Amer Ardati. Proceedings of the 20th Annual SIGdial Meeting on Discourse and Dialogue. 2019.

Our system generates summaries of hospital stays by combining information from two heterogenous sources: physician discharge notes and nursing plans of care.It extracts medical concepts from both sources; concepts that are identified as "complex" by our metric are explained by providing definitions obtained from three external knowledge sources.Finally, relevant concepts (with or without definition) are realized by SimpleNLG.

The ability of people to hear and respond to auditory medical alarms is critical to the health and safety of patients. Unfortunately, concurrently sounding alarms can perceptually interact in ways that mask one or more of them: making them impossible to hear. Because masking may only occur in extremely specific and/or rare situations, experimental evaluation techniques are insufficient for detecting masking in all of the potential alarm configurations used in medicine. Thus, a real need exists for computational methods capable of determining if masking exists in medical alarm configurations before they are deployed. In this paper, we present such a method. Using a combination of formal modeling, psychoacoustic modeling, temporal logic specification, and model checking, our method is able to prove whether a modeled of a configuration of alarms can interact in ways that produce masking. This paper provides the motivation for this method, presents its details, describes its implementation, demonstrates its power with a case study, and outlines future work.

In the United States, International Classification of Disease Clinical Modification (ICD-9-CM, the ninth revision) diagnosis codes are commonly used to identify patient cohorts and to conduct financial analyses related to disease. In October 2015, the healthcare system of the United States will transition to ICD-10-CM (the tenth revision) diagnosis codes. One challenge posed to clinical researchers and other analysts is conducting diagnosis-related queries across datasets containing both coding schemes. Further, healthcare administrators will manage growth, trends, and strategic planning with these dually-coded datasets. The majority of the ICD-9-CM to ICD-10-CM translations are complex and nonreciprocal, creating convoluted representations and meanings. Similarly, mapping back from ICD-10-CM to ICD-9-CM is equally complex, yet different from mapping forward, as relationships are likewise nonreciprocal. Indeed, 10 of the 21 top clinical categories are complex as 78% of their diagnosis codes are labeled as "convoluted" by our analyses. Analysis and research related to external causes of morbidity, injury, and poisoning will face the greatest challenges due to 41 745 (90%) convolutions and a decrease in the number of codes. We created a web portal tool and translation tables to list all ICD-9-CM diagnosis codes related to the specific input of ICD-10-CM diagnosis codes and their level of complexity: "identity" (reciprocal), "class-to-subclass," "subclass-to-class," "convoluted," or "no mapping." These tools provide guidance on ambiguous and complex translations to reveal where reports or analyses may be challenging to impossible.Web portal: http://www.lussierlab.org/transition-to-ICD9CM/Tables annotated with levels of translation complexity: http://www.lussierlab.org/publications/ICD10to9.

Reporting of hospital adverse events relies on Patient Safety Indicators (PSIs) using International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) codes. The US transition to ICD-10-CM in 2015 could result in erroneous comparisons of PSIs. Using the General Equivalent Mappings (GEMs), we compared the accuracy of ICD-9-CM coded PSIs against recommended ICD-10-CM codes from the Centers for Medicaid/Medicare Services (CMS). We further predict their impact in a cohort of 38,644 patients (1,446,581 visits and 399 hospitals). We compared the predicted results to the published PSI related ICD-10-CM diagnosis codes. We provide the first report of substantial hospital safety reporting errors with five direct comparisons from the 23 types of PSIs (transfusion and anesthesia related PSIs). One PSI was excluded from the comparison between code sets due to reorganization, while 15 additional PSIs were inaccurate to a lesser degree due to the complexity of the coding translation. The ICD-10-CM translations proposed by CMS pose impending risks for (1) comparing safety incidents, (2) inflating the number of PSIs, and (3) increasing the variability of calculations attributable to the abundance of coding system translations. Ethical organizations addressing 'data-, process-, and system-focused' improvements could be penalized using the new ICD-10-CM Agency for Healthcare Research and Quality PSIs because of apparent increases in PSIs bearing the same PSI identifier and label, yet calculated differently. Here we investigate which PSIs would reliably transition between ICD-9-CM and ICD-10-CM, and those at risk of under-reporting and over-reporting adverse events while the frequency of these adverse events remain unchanged.

Barbara Di Eugenio, Andrew Boyd, Camillo Lugaresi, Abhinaya Balasubramanian, Gail Keenan, Mike Burton, Tamara Goncalves Rezende Macieira, Jianrong Li, Yves Lussier, Yves Lussier. Proceedings of the 8th International Natural Language Generation Conference (INLG). 2014.

OBJECTIVE: Applying the science of networks to quantify the discriminatory impact of the ICD-9-CM to ICD-10-CM transition between clinical specialties. MATERIALS AND METHODS: Datasets were the Center for Medicaid and Medicare Services ICD-9-CM to ICD-10-CM mapping files, general equivalence mappings, and statewide Medicaid emergency department billing. Diagnoses were represented as nodes and their mappings as directional relationships. The complex network was synthesized as an aggregate of simpler motifs and tabulation per clinical specialty. RESULTS: We identified five mapping motif categories: identity, class-to-subclass, subclass-to-class, convoluted, and no mapping. Convoluted mappings indicate that multiple ICD-9-CM and ICD-10-CM codes share complex, entangled, and non-reciprocal mappings. The proportions of convoluted diagnoses mappings (36% overall) range from 5% (hematology) to 60% (obstetrics and injuries). In a case study of 24 008 patient visits in 217 emergency departments, 27% of the costs are associated with convoluted diagnoses, with 'abdominal pain' and 'gastroenteritis' accounting for approximately 3.5%. DISCUSSION: Previous qualitative studies report that administrators and clinicians are likely to be challenged in understanding and managing their practice because of the ICD-10-CM transition. We substantiate the complexity of this transition with a thorough quantitative summary per clinical specialty, a case study, and the tools to apply this methodology easily to any clinical practice in the form of a web portal and analytic tables. CONCLUSIONS: Post-transition, successful management of frequent diseases with convoluted mapping network patterns is critical. The http://lussierlab.org/transition-to-ICD10CM web portal provides insight in linking onerous diseases to the ICD-10 transition.

For the success of clinical and translational science, a seamless interoperation is required between clinical and research information technology. Addressing this need, the Michigan Clinical Research Collaboratory (MCRC) was created. The MCRC employed a standards-driven Web Services architecture to create the U-M Honest Broker, which enabled sharing of clinical and research data among medical disciplines and separate institutions. Design objectives were to facilitate sharing of data, maintain a master patient index (MPI), deidentification of data, and routing data to preauthorized destination systems for use in clinical care, research, or both. This article describes the architecture and design of the U-M HB system and the successful demonstration project. Seventy percent of eligible patients were recruited for a prospective study examining the correlation between interventional cardiac catheterizations and depression. The U-M Honest Broker delivered on the promise of using structured clinical knowledge shared among providers to help clinical and translational research.