Automated Decision Support for Drug-Induced Long QT.

Advancements in clinical informatics and translational genomics are changing the way we practice medicine. Automated decision support currently helps providers adjust prescribing patterns to reduce the likelihood of QT prolongation based upon drug-drug interaction. A similar approach is being explored for drug-gene interaction. Like many adverse drug reactions, QT prolongation can be influenced by variability in genetic factors. However, drug-induced QT prolongation can occur in the absence of any known ion channel gene abnormalities. We therefore review differences between congenital long QT syndrome and drug-induced long QT syndrome, and we underscore the need for decision support that integrates EKG data.

Implementation of wide-scale pharmacogenetic testing in primary care.

The convergence of translational genomics and biomedical informatics has changed healthcare delivery. Institutional consortia have begun implementing lab testing and decision support for drug-gene interactions. Aggregate datasets are now revealing the impact of clinical decision support for drug-gene interactions. Given the pleiotropic nature of pharmacogenes, interdisciplinary teams and robust clinical decision support tools must exist within an informatics framework built to be flexible and capable of cross-talk between clinical specialties. Navigation of the challenges presented with the implementation of five steps to build a genetics program infrastructure requires the expertise of multiple healthcare professionals. Ultimately, this manuscript describes our efforts to place pharmacogenomics in the hands of the primary care provider integrating this information into a patient's healthcare over their lifetime.

Chronic Kidney Disease in Agricultural Communities.

Patients residing in agricultural communities have a high risk of developing chronic kidney disease. In the Great Plains, geo-environmental risk factors (eg, variable climate, temperature, air quality, water quality, and drought) combine with agro-environmental risk factors (eg, exposure to fertilizers, soil conditioners, herbicides, fungicides, and pesticides) to increase risk for toxic nephropathy. However, research defining the specific influence of agricultural chemicals on the progression of kidney disease in rural communities has been somewhat limited. By linking retrospective clinical data within electronic medical records to environmental data from sources like US Environmental Protection Agency, analytical models are beginning to provide insight into the impact of agricultural practices on the rate of progression for kidney disease in rural communities.

Learning from Longitudinal Data in Electronic Health Record and Genetic Data to Improve Cardiovascular Event Prediction.

Current approaches to predicting a cardiovascular disease (CVD) event rely on conventional risk factors and cross-sectional data. In this study, we applied machine learning and deep learning models to 10-year CVD event prediction by using longitudinal electronic health record (EHR) and genetic data. Our study cohort included 109, 490 individuals. In the first experiment, we extracted aggregated and longitudinal features from EHR. We applied logistic regression, random forests, gradient boosting trees, convolutional neural networks (CNN) and recurrent neural networks with long short-term memory (LSTM) units. In the second experiment, we applied a late-fusion approach to incorporate genetic features. We compared the performance with approaches currently utilized in routine clinical practice - American College of Cardiology and the American Heart Association (ACC/AHA) Pooled Cohort Risk Equation. Our results indicated that incorporating longitudinal feature lead to better event prediction. Combining genetic features through a late-fusion approach can further improve CVD prediction, underscoring the importance of integrating relevant genetic data whenever available.

Decreasing the critical value of hemoglobin required for physician notification reduces the rate of blood transfusions.

Red blood cell transfusions have been cited as one of the most overused therapeutic interventions in the USA. Excessively aggressive transfusion practices may be driven by mandatory physician notification of critical hemoglobin values that do not generally require transfusion. We examined the effect of decreasing the critical value of hemoglobin from 8 to 7 g/dL at our institution. Along with this change, mandatory provider notification for readings between 7 and 8 g/dL was rescinded. Transfusion rates were compared retrospectively during paired 5-month periods for patients presenting in three key hemoglobin ranges (6.00-6.99, 7.00-7.99, and 8.00-8.99 g/dL). A change in transfusion practices was hypothesized in the 7-8 g/dL range, which was no longer labeled critical and for which mandated physician calls were rescinded. Transfusion rates showed a statistically significant 8% decrease (P≤0.0001) during the 5-month period post change in our transfusion practices. This decrease in the 7.00-7.99 g/dL range was significantly greater than the 2% decrease observed in either the 6-6.99 g/dL (P=0.0017) or 8-8.99 g/dL (P≤0.0001) range. Cost savings of up to $700,000/year were extrapolated from our results showing 491 fewer units of red blood cells transfused during the 5-month post change. These cost savings do not take into account the additional impact of complications associated with blood transfusions.