Matching Items (3)
Filtering by
- All Subjects: Electronic Health Records
- Creators: Anwar, Saadat
- Creators: Grando, Adela Maria
- Resource Type: Text
Description
Medical records are increasingly being recorded in the form of electronic health records (EHRs), with a significant amount of patient data recorded as unstructured natural language text. Consequently, being able to extract and utilize clinical data present within these records is an important step in furthering clinical care. One important aspect within these records is the presence of prescription information. Existing techniques for extracting prescription information — which includes medication names, dosages, frequencies, reasons for taking, and mode of administration — from unstructured text have focused on the application of rule- and classifier-based methods. While state-of-the-art systems can be effective in extracting many types of information, they require significant effort to develop hand-crafted rules and conduct effective feature engineering. This paper presents the use of a bidirectional LSTM with CRF tagging model initialized with precomputed word embeddings for extracting prescription information from sentences without requiring significant feature engineering. The experimental results, run on the i2b2 2009 dataset, achieve an F1 macro measure of 0.8562, and scores above 0.9449 on four of the six categories, indicating significant potential for this model.
ContributorsRawal, Samarth Chetan (Author) / Baral, Chitta (Thesis director) / Anwar, Saadat (Committee member) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
This thesis concerns the adoption of health information technology in the medical sector, specifically electronic health records (EHRs). EHRs have been seen as a great benefit to the healthcare system and will improve the quality of patient care. The federal government, has seen the benefit EHRs can offer, has been advocating the use and adoption of EHR for nearly a decade now. They have created policies that guide medical providers on how to implement EHRs. However, this thesis concerns the attitudes medical providers in Phoenix have towards government implementation. By interviewing these individuals and cross-referencing their answers with the literature this thesis wants to discover the pitfalls of federal government policy toward EHR implementation and EHR implementation in general. What this thesis found was that there are pitfalls that the federal government has failed to address including loss of provider productivity, lack of interoperability, and workflow improvement. However, the providers do say there is still a place for government to be involved in the implementation of EHR.
ContributorsKaldawi, Nicholas Emad (Author) / Lewis, Paul (Thesis director) / Cortese, Denis (Committee member) / Jones, Ruth (Committee member) / Barrett, The Honors College (Contributor) / School of Politics and Global Studies (Contributor) / School of Human Evolution and Social Change (Contributor)
Created2013-05
Description
The traditional model of assessing and treating behavioral health (BH) and physical health (PH) in silos is inadequate for supporting whole-person health and wellness. The integration of BH and PH may result in better care quality, patient-provider experiences, outcomes, and reduced costs. Cross-organizational health data sharing between BH and PH providers is critical to patients with BH conditions (BHCs). In the last few decades, many initiatives -including health information exchange organizations- have facilitated cross-organizational health data sharing. The current challenge is affording meaningful consent and ensuring patient privacy, two of the core requirements for advancing the adoption and use of health information technology (HIT) in the US. The Office of the National Coordinator for HIT (ONC) recommends that patients should be given granular control beyond the “share all” or “share none” approach widely used currently in consent practices. But there is no consensus on the variables relevant to promote granularity in data sharing to honor privacy satisfaction for patients. As a result, existing granular data sharing (GDS) studies use ad-hoc and non-standardized approaches to implement or investigate patient data sharing preferences. Novel informatics methods were proposed and piloted to support patient-driven GDS and to validate the suitability and applicability of such methods in clinical environments. The hypotheses were: H1) the variables recommended by the ONC are relevant to support GDS; H2) there is diversity in medical record sharing preferences of individuals with BHCs; and H3) the most frequently used sensitive data taxonomy captures sensitive data sharing preferences of patients with BHCs. Findings validated the study hypotheses by proposing an innovative standards-based GDS framework, validating the framework with the design and pilot testing of a clinical decision support system with 209 patients with BHCs, validating with patients the adequacy of the most frequently used sensitive data taxonomy, and systematically exploring data privacy views and data sharing perceptions of patients with BHCs. This research built the foundations for a new generation of future data segmentation methods and tools that advances the vision of the ONC of creating standards-based, interoperable models to share sensitive health information in compliance with patients’ data privacy preferences.
ContributorsKarway, George K (Author) / Grando, Adela Maria (Thesis advisor) / Murcko, Anita C (Committee member) / Franczak, Michael (Committee member) / Arizona State University (Publisher)
Created2022