Browsing by Subject "EHR"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Statin Medication Adverse Event Detection and Prediction to Help Optimize Cardiovascular Therapy(2022-08) KIOGOU, SEBASTIENAs lipid-lowering drugs, statin medications have been shown to be effective therapies, especially for cardiovascular patients. However, patients cannot take full advantage of the benefits statin drugs offer due to statin related adverse drug events (ADE) that often foster medication non-adherence and add to patients’ morbidity and mortality. On the other hand, the increasing availability of EHR data in electronic repositories along with the advancement in data management and integration represent substantial opportunities for scientists to practically investigate unanswered questions including those of statin related ADEs. In particular, we can ask ourselves: based on available electronic medical record data, how could we identify patients taking statin drugs who are at risk of developing symptomatic adverse events?The current thesis set forth methodologies and scientific artifacts to help with clinical decision support for both general adverse drug events and statin related adverse drug events in particular. We defined methods for clinical data cohort quality assessment and improvement, compiled a list of drugs with the potential of statin to drug interactions, identified common co-morbidities that affect statin users, elaborated on advanced data science methods for clinical data cohort mining and machine leaning, built predictive models for statin related adverse events detection and surveillance, and defined contexts in which these tools can be used for health care decision support. This contribution is significant because it has the advantage of not only reducing the occurrence of statin related ADEs but also propagating a feasible and replicable EHR solution for general ADE reporting. Healthcare organizations can rely on an easy to implement EHR solution to significantly reduce statin related ADEs, save on health care resources, and improve patient management. Local and state governments could reduce their healthcare expenditures. Healthcare quality indices such as number of hospitalizations, length of stay in hospital, re-admission, medication non-adherence, and CVD mortality could be improved. A greater capacity for broader ADE mitigation efforts could also be set forth, which could have a beneficial impact on human health.Item Towards A Model Representation Of Residence, Living Conditions, And Living Situation: An Evaluation Of Clinical Practice In Documentation, And Associated Standards And Informational Models(2017-03) Winden, TamaraSocial determinants of health (SDOH) play an important role in diagnosis, prevention, health outcomes, and quality of life and have long been a consideration in the provision of care. SDOH can cause illness, exacerbate chronic illness, but can also improve health. The SDOH target domains of interest for this proposal are Residence, Living Situation, and Living Conditions. Historically, these three topic areas are either not well documented in the EHR or, if they are documented, they are in unstructured or semi-structured text making the information inaccessible to clinical decision support tools or secondary use in research. Building upon previous work, the overall goal of this three-part study is to enhance current information models of these three domains through the evaluation of content and completeness of existing standards and terminologies and examine how coherent our documentation is of these target domains is in the electronic health record through analysis of specific ancillary notes as well as other structured and unstructured sections of the EHR. Examination of 27 standards and terminology sources contributed to an enhanced model representation for Residence, Living Situation, and Living Conditions. However, our findings show there are no fully comprehensive standards for EHR documentation of these topic areas. An analysis of unstructured text from the EHR demonstrated that SDOH are being documented in notes and these data contributed to further enhanced model. Lastly, an examination of flowsheet data showed many inconsistencies in flowsheet build and documentation, many of which could be solved if we have a comprehensive standard for documentation of SDOH.