Academic Research

Large language models (LLMs) have emerged as a popular resource for generating information relevant to a user query. Such models are created through a resource-intensive training process utilizing an extensive, static corpus of textual data. This static nature results in limitations for adoption in domains with rapidly changing knowledge, proprietary information, and sensitive data. In this work, methods are outlined for augmenting general-purpose LLMs, known as foundation models, with domain-specific information using an embeddings-based approach for incorporating up-to-date, peer-reviewed scientific manuscripts.

Transitions of care are pivotal, vulnerable times as patients are discharged from the hospital. Telephonic care coordination is standard care, but labor intensive. We implemented a patient postdischarge digital engagement (PDDE) program to scale coordination. We hypothesized that PDDE could reduce readmissions for low-risk patients and supplement care coordination for medium- and high-risk patients.

Despite the widespread adoption of early warning systems (EWSs), it is uncertain if their implementation improves patient outcomes. The authors report a pre-post quasi-experimental evaluation of a commercially available EWS on patient outcomes at a 700-bed academic medical center. The EWS risk scores were visible in the electronic medical record by bedside clinicians. The EWS risk scores were also monitored remotely 24/7 by critical care trained nurses who actively contacted bedside nurses when a patient's risk levels increased. The primary outcome was inpatient mortality. Secondary outcomes were rapid response team calls and activation of cardiopulmonary arrest (code-4) response teams. The study team conducted a regression discontinuity analysis adjusting for age, gender, insurance, severity of illness, risk of mortality, and hospital occupancy at admission. The analysis included 53,229 hospitalizations. Adjusted analysis showed no significant change in inpatient mortality, rapid response team call, or code-4 activations after implementing the EWS. This study confirms the continued uncertainty in the effectiveness of EWSs and the need for further rigorous examinations of EWSs.

Interdisciplinary research teams can be extremely beneficial when addressing difficult clinical problems. The incorporation of conceptual and methodological strategies from a variety of research disciplines and health professions yields transformative results. In this setting, the long-term goal of team science is to improve patient care, with emphasis on population health outcomes.

The problem of unaffordable prescription medications in the United States is complex and can result in poor patient adherence to therapy, worse clinical outcomes, and high costs to the healthcare system. While providers are aware of the financial burden of healthcare for patients, there is a lack of actionable price transparency at the point of prescribing. Real-time prescription benefit (RTPB) tools are new electronic clinical decision support tools that retrieve patient- and medication-specific out-of-pocket cost information and display it to clinicians at the point of prescribing.

In this narrative case, we identify issues related to patients’ use of technology, make comparisons between telehealth adoption and the deployment of electronic health records, and propose that building intuitive and supported digital care experiences for patients is required to make virtual care sustainable.

This quality improvement study of 137 846 video visits showed an overall 90% success rate. Patient rather than clinician factors were more systematically associated with successful completion of video visits, and clinician comfort with technology was associated with successful video visits or conversion to telephone visits. The findings suggest that, as policy makers consider expanding telehealth coverage and hospital systems focus on investments, consideration of patient support, equity, and friction should be kept in the forefront.

Metformin, an antidiabetic drug, triggers anti-aging cellular responses. Aging is the principal risk factor for dementia, but previous observational studies of the diabetes drugs metformin vs. sulfonylureas have been mixed. We tested the hypotheses that metformin improves survival and reduces the risk of dementia, relative to the sulfonylureas, by emulating target trials in electronic health records of diabetic patients at an academic-centered healthcare system in the US and a wide-ranging group of primary care practices in the UK. To address metformin’s potentially dual influences on dementia risk—that it might reduce the hazard of death and put more people at risk of developing dementia while reducing the hazard of dementia by slowing biological aging, we used a competing risks approach and carefully grounded that within a causal inference emulated trial framework. To identify candidate biomarkers of metformin’s actions in the brain that might mediate reduced dementia risk, we conducted an in-vitro systems pharmacology evaluation of metformin and glyburide on differentiated human neural cells through differential gene expression. We named our multi-dimensional approach DRIAD-EHR (Drug Repurposing in Alzheimer’s Disease-Electronic Health Records). In intention-to-treat analyses, metformin was associated with a lower hazard of all-cause mortality than sulfonylureas in both cohorts. In competing risks analyses, there was also a lower cause-specific hazard of dementia onset among metformin initiators. In in-vitro studies, metformin reduced human neural cell expression of SPP1 and APOE, two secreted proteins that have been implicated in Alzheimer’s disease pathogenesis and whose levels can be quantified in the CSF. Together, our findings suggest that metformin might prevent dementia in patients without type II diabetes. In addition, our results inform the design of clinical trials of metformin in non-diabetics and suggest a pharmacodynamic CSF biomarker, SPP1, for metformin’s action in the brain.

The outbreak of coronavirus disease 2019 (COVID-19) further accelerated the deployment and utilization of telemedicine services. An analysis of the socioeconomic characteristics of telemedicine users to understand potential socioeconomic gaps and disparities is critical for improving the adoption of telemedicine services among patients.. This study aims to measure the correlation of socioeconomic determinants with the use of telemedicine services in Milwaukee metropolitan area.

Online triage tools are increasingly being adopted in health care to aid patients in identifying the appropriate care level. However, there is a lack of empirical evidence on how patients use virtual triage and whether these tools influence care-seeking behavior. Using data from a free online triage tool, we describe the common symptoms queried by users and analyze whether the tool was associated with the level of care that patients intended to seek.

The ever-increasing pace of technological advancements, rising costs, and new entrants into the health care marketplace are part of the challenge health care incumbents face today. With no alternative but to adapt, health care organizations must find effective methods to embrace innovation, which we define as the delivery of new patient and clinician value. Embedding and accelerating innovation in health care, however, has proven to be difficult. In health care, most current processes of governance, business planning, and information technology implementation are designed to minimize risk to organizations and are often inflexible to adapt quickly to new technological changes, netting incremental changes that fail to deliver much needed transformation.

Advances in predictive analytics and machine learning supported by an ever-increasing wealth of data and processing power are transforming almost every industry. Accuracy and precision of predictive analytics have significantly increased over the past few years and are evolving at an exponential pace. There have been significant breakthroughs in using Predictive Analytics in healthcare where it is held as the foundation of precision medicine. Yet, although the research in the field is expanding with the profuse volume of papers applying machine learning algorithms to medical data, very few have contributed meaningfully to clinical care. This lack of impact stands in stark contrast to the enormous relevance of machine learning to many other industries. Regardless of the status of its current contribution, the field of predictive analytics is expected to fundamentally change the way we diagnose and treat diseases, as well as the conduct of biomedical science research. In this review, we describe the main tools and techniques in predictive analytics and will analyze the trends in application of these techniques over the recent years. We will also provide examples of its application in medicine and more specifically in stroke and neurovascular research and outline current limitations.

The UK is one of the largest funders of health research in the world, but little is known about how health funding is spent. Our study explores whether major UK public and charitable health research funders support the research of UK-based scientists producing the most highly-cited research.