The wide implementation of electronic health record (EHR) systems facilitates the collection of large-scale health data from real clinical settings. Despite the significant increase in adoption of EHR systems, this data remains largely unexplored, but presents a rich data source for knowledge discovery from patient health histories in tasks such as understanding disease correlations and predicting health outcomes. However, the heterogeneity, sparsity, noise, and bias in this data present many complex challenges. This complexity makes it difficult to translate potentially relevant information into machine learning algorithms. In this paper, we propose a computational framework, Patient2Vec, to learn an interpretable deep representation of longitudinal EHR data which is personalized for each patient. To evaluate this approach, we apply it to the prediction of future hospitalizations using real EHR data and compare its predictive performance with baseline methods. Patient2Vec produces a vector space with meaningful structure and it achieves an AUC around 0.799 outperforming baseline methods. In the end, the learned feature importance can be visualized and interpreted at both the individual and population levels to bring clinical insights.
Aims: To examine factors that affect cost-related medication non-adherence (CRN), defined as taking medication less than as prescribed because of cost, among adults with diabetes and to determine their relative contribution in explaining CRN. Methods: Behavioral Risk Factor Surveillance System data for 2013–2014 were used to identify individuals with diabetes and their CRN. We modeled CRN as a function of financial factors, regimen complexity, and other contextual factors including diabetes care, lifestyle, and health factors. Dominance analysis was performed to rank these factors by relative importance. Results: CRN among U.S. adults with diabetes was 16.5%. Respondents with annual income <$50,000 and without health insurance were more likely to report CRN, compared to those with income ≥$50,000 and those with insurance, respectively. Insulin users had 1.24 times higher risk of CRN compared to those not on insulin. Contextual factors that significantly affected CRN included diabetes care factors, lifestyle factors, and comorbid depression, arthritis, and COPD/asthma. Dominance analysis showed health insurance was the most important factor for respondents <65 and depression was the most important factor for respondents ≥65. Conclusions: In addition to traditional risk factors of CRN, compliance with annual recommendations for diabetes and healthy lifestyle were associated with lower CRN. Policies and social supports that address these contextual factors may help improve CRN.
ObjectiveBurnout is a public health crisis that impacts 1 in 3 registered nurses in the United States and the safe provision of patient care. This study sought to understand the cost of nurse burnout-attributed turnover using hypothetical hospital scenarios.MethodsA cost-consequence analysis with a Markov model structure was used to assess nurse burnout-attributed turnover costs under the following scenarios: (1) a hospital with “status quo” nurse burnout prevalence and (2) a hospital with a “burnout reduction program” and decreased nurse burnout prevalence. The model evaluated turnover costs from a hospital payer perspective and modeled a cohort of nurses who were new to a hospital. The outcome measures were defined as years in burnout among the nurse cohort and years retained/employed in the hospital. Data inputs derived from the health services literature base.ResultsThe expected model results demonstrated that at status quo, a hospital spends an expected $16,736 per nurse per year employed on nurse burnout-attributed turnover costs. In a hospital with a burnout reduction program, such costs were $11,592 per nurse per year employed. Nurses spent more time in burnout under the status quo scenario compared with the burnout reduction scenario (1.5 versus 1.1 y of employment) as well as less time employed at the hospital (2.9 versus 3.5 y of employment).ConclusionsGiven that status quo costs of burnout are higher than those in a hospital that invests in a nurse burnout reduction program, hospitals should strongly consider proactively supporting programs that reduce nurse burnout prevalence and associated costs.
Several non-hand hygiene activities took place regularly in ICU handwashing sinks; these may provide a mechanism for nosocomial transmission and promotion of bacterial growth in the drain. Redesigning hospital workflow and sink usage may be necessary as it becomes apparent that sink drains may be a reservoir for transmission of multidrug-resistant bacteria.
OBJECTIVE We sought to evaluate the role healthcare providers play in carbapenem-resistant Enterobacteriaceae (CRE) acquisition among hospitalized patients. DESIGN A 1:4 case-control study with incidence density sampling. SETTING Academic healthcare center with regular CRE perirectal screening in high-risk units. PATIENTS We included case patients with ≥1 negative CRE test followed by positive culture with a length of stay (LOS) >9 days. For controls, we included patients with ≥2 negative CRE tests and assignment to the same unit set as case patients with a LOS >9 days. METHODS Controls were time-matched to each case patient. Case exposure was evaluated between days 2 and 9 before positive culture and control evaluation was based on maximizing overlap with the case window. Exposure sources were all CRE-colonized or -infected patients. Nonphysician providers were compared between study patients and sources during their evaluation windows. Dichotomous and continuous exposures were developed from the number of source-shared providers and were used in univariate and multivariate regression. RESULTS In total, 121 cases and 484 controls were included. Multivariate analysis showed odds of dichotomous exposure (≥1 source-shared provider) of 2.27 (95% confidence interval [CI], 1.25-4.15; P=.006) for case patients compared to controls. Multivariate continuous exposure showed odds of 1.02 (95% CI, 1.01-1.03; P=.009) for case patients compared to controls. CONCLUSIONS Patients who acquire CRE during hospitalization are more likely to receive care from a provider caring for a patient with CRE than those patients who do not acquire CRE. These data support the importance of hand hygiene and cohorting measures for CRE patients to reduce transmission risk. Infect Control Hosp Epidemiol 2017;38:1329-1334.
Objective: The study aimed to examine the association based on objective estimates of sleep duration and quality and aortic stiffness while accounting for the potential confounding effect of SDB. Method: Participants were part of the Multi-Ethnic Study of Atherosclerosis (MESA) Sleep study. Sleep duration and quality were assessed by 7-day wrist actigraphy, SDB by home polysomnography, and aortic stiffness by magnetic resonance imaging (MRI)-based aortic pulse wave velocity (aPWV), ascending and descending aorta distensibility (AAD and DAD). Aortic stiffness of participants with ‘normal’ sleep duration (6–8 hours) were compared with those of ‘short’ (<6 hours) and ‘long’ sleep duration (>8 hours) adjusting for common cardiovascular risk factors and apnea hypopnea index (AHI). Results: The sample consisted of 908 participants (mean age 68.4±9.1 years, 55.3% female). There was a significant linear trend of increased aPWV across short (n=252), normal (n=552), and long sleep durations (n=104) (p for trend=0.008). Multivariable analysis showed that people with short sleep duration had 0.94 m/s lower aPWV (95% CI: −1.54, −0.35), compared with those with normal sleep duration. Conclusion: In this ethnically diverse community cohort, habitual short sleep duration as estimated by actigraphy was associated with lower aortic stiffness.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.