The rapid adoption of Electronic Health Records (EHRs)--electronic versions of a patient's medical history--has been instrumental in streamlining administrative tasks, increasing transparency, and enabling continuity of care across providers. An unintended consequence of the increased documentation burden, however, has been reduced face-time with patients and, concomitantly, a dramatic rise in clinician burnout. Time spent maintaining and making sense of a patient's electronic record is a leading cause of burnout. In this thesis, we pinpoint a particularly time-intensive, yet critical, documentation task: generating a summary of a patient's hospital admissions, and propose and evaluate automated solutions. In particular, we focus on faithfulness, i.e., accurately representing the patient record, and completeness, i.e., representing the full context, as the sine qua non for safe deployment of a hospital-course summarization tool in a clinical setting. The bulk of this thesis is broken up into four chapters: §2 Creating and Analyzing the Data, §3 Improving the Faithfulness of Summaries, §4 Measuring the Faithfulness of Summaries, and, finally, §5 Generating Grounded, Complete Summaries with LLMs. Each chapter links back to the core themes of faithfulness and completeness, while the chapters are linked to each other in that the findings from each chapter shape the direction of subsequent chapters. Given the documentation authored throughout a patient's hospitalization, hospital-course summarization requires generating a lengthy paragraph that tells the story of the patient admission. In § 2, we construct a dataset based on 109,000 hospitalizations (2M source notes) and perform exploratory analyses to motivate future work on modeling and evaluation [NAACL 2021]. The presence of highly abstractive, entity dense references, coupled with the high stakes nature of text generation in a clinical setting, motivates us to focus on faithfulness and adequate coverage of salient medical entities. In § 3, we address faithfulness from a modeling perspective by revising noisy references [EMNLP 2022] and, to reduce the reliance on references, directly calibrating model outputs to metrics [ACL 2023]. These works relied heavily on automatic metrics as human annotations were limited. To fill this gap, in §4, we conduct a fine-grained expert annotation of system errors in order to meta-evaluate existing metrics and better understand task-specific issues of domain adaptation and source-summary alignments. We find that automatically generated summaries can exhibit many errors, including incorrect claims and critical omissions, despite being highly extractive. These errors are missed by existing metrics. To learn a metric which is less correlated to extractiveness (copy-and-paste), we derive noisy faithfulness labels from an ensemble of existing metrics and train a faithfulness classifier on these pseudo labels [MLHC 2023]. Finally, in § 5, we demonstrate that fine-tuned LLMs (Mistral and Zephyr) are highly prone to entity hallucinations and cover fewer salient entities. We improve both coverage and faithfulness by performing sentence-level entity planning based on a set of pre-computed salient entities from the source text, which extends our work on entity-guided news summarization ([ACL, 2023] and [EMNLP, 2023]).

No description available.

The purpose of this project was to develop a document imaging application. By scanning the documents into an electronic repository, medical staff will be able to more easily store and locate these records. To make the application user friendly and facilitate staff access to patient medical records, the application is wed-based and uses the Oracle Application Server to implement a multitiered model.

A Dissertation in fulfillment of the requirements for the Degree of Master of Commerce by Research in Information Systems / Electronic Medical Records (EMR) systems aim to provide an integrated solution to a number of hospital processes, including supporting administrative functions such as patient billing, providing clinical guidelines, and allowing clinicians to order and view lab reports. Yet with all these possible benefits, like many other types of information systems and technologies, their impact on hospital performance has been a key area of interest and a source of debate. This study examines the extent to which EMR systems are available across South African hospitals as well as how well they have been integrated into the processes of the said hospitals. The diffusion of EMR systems means they are available in the relevant clinical units whilst their infusion suggests they are comprehensively embedded into the processes that they are meant to support. The most important question answered in this study was whether the diffusion and infusion of these systems had any impact on the performance of hospitals. Answering this question would aid decision makers on whether more effort is indeed needed to ensure their availability across units and their embedding into hospital processes. The study collected data from a number of hospitals in the private and public sector with multiple informants providing data on diffusion and infusion of these systems as well as on hospital performance. Performance measures included the hospitals’ economic efficiency, their clinical effectiveness, patient safety, and patient satisfaction. Key findings were that while there are performance benefits in the diffusion of EMR systems across units, the impacts are not always positive for some hospitals. Large hospitals were less likely to benefit from the diffusion of these systems. However the comprehensive use of these systems provided benefits for all types of hospitals, suggesting that while availability may be important, it is the comprehensive embedding of the EMR system into the hospital’s processes that is more beneficial to all types of hospitals. These results present further opportunity for research into why the impact of EMR systems is not always positive for some hospitals. / MT2017

In order to improve the quality of care, there is urgent need to involve patients in their own healthcare. So to make patient centered health care system Personal Health Records are proposed as viable solution. This research discusses the importance of a Patient Centric Health Record system. Such systems can empower patients to participate in improving health care quality. It would also provide an economically viable solution to the need for better healthcare without escalating costs by avoiding duplication. The proposed system is Web-based; therefore it has high accessibility and availability. The cloud computing based architecture is used which will allow consumers to address the challenge of sharing medical data. PHR would provide a complete and accurate summary of the health and medical history of an individual by gathering data from many sources. This would make information accessible online to anyone who has the necessary electronic credentials to view the information. / by Vishesh Ved. / Thesis (M.S.C.S.)--Florida Atlantic University, 2010. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2010. Mode of access: World Wide Web.

Theoretically, the use of Electronic Medical Records (EMRs) holds promise of numerous benefits in healthcare provision, including improvement in continuity of care, quality of care and safety. However, in practice, there is evidence that the adoption of electronic medical records has been slow and where adopted, often lacks meaningful use. Thus there is a clear dichotomy between the ambitions for EMR use and the reality of EMR implementation. In the USA, a legislative approach was taken to turn around the situation. Other countries such as Canada and European countries have followed suit (in their own way) to address the adoption and meaningful use of electronic medical records. The South African e-Health strategy and the National Health Normative Standards Framework for Interoperability in eHealth in South Africa documents both recommend the adoption of EMRs. Much work has been done to establish a baseline for standards to ensure interoperability and data portability of healthcare applications and data. However, even with the increased focus on e-Health, South Africa remains excessively reliant on paper-based medical records. Where health information technologies have been adopted, there is lack of coordination between and within provinces, leading to a multitude of systems and vendors. Thus there is a lack of systematic adoption and meaningful use of EMRs in South Africa. The main objective of this research is to develop the components required to engender meaningful use of electronic medical records in the South African healthcare context. The main contributors are identified as EMR certification and consistent, proper use of certified EMRs. Literature review, a Delphi study and logical argumentation are used to develop the relevant components for the South African healthcare context. The benefits of EMRs can only be realized through systematic adoption and meaningful use of EMRs, thus this research contributes to providing a road map for engendering the meaningful use of EMRs with the ultimate aim of improving healthcare in the South African healthcare landscape.

Thesis (MTech (Information Technology))--Cape Peninsula University of Technology, 2016. / Access to healthcare is regarded as a basic and essential human right. It is widely known that ICT solutions have potential to improve access to healthcare, reduce healthcare cost, reduce medical errors, and bridge the digital divide between rural and urban healthcare centres. The access to personal healthcare records is, however, an astounding challenge for both patients and healthcare professionals alike, particularly within resource-restricted environments (such as rural communities). Most rural healthcare institutions have limited or non-existent access to electronic patient healthcare records. This study explored the accessibility of personal healthcare records by patients and healthcare professionals within a rural community hospital in the Eastern Cape Province of South Africa. The case study was conducted at the St. Barnabas Hospital with the support and permission from the Faculty of Informatics and Design, Cape Peninsula University of Technology and the Eastern Cape Department of Health. Semi-structured interviews, observations, and interactive co-design sessions and focus groups served as the main data collection methods used to determine the accessibility of personal healthcare records by the relevant stakeholders. The data was qualitatively interpreted using thematic analysis. The study highlighted the various challenges experienced by healthcare professionals and patients, including time-consuming manual processes, lack of infrastructure, illegible hand-written records, missing records and illiteracy. A number of recommendations for improved access to personal healthcare records are discussed. The significance of the study articulates the imperative need for seamless and secure access to personal healthcare records, not only within rural areas but within all communities.

No description available.

This dissertation investigates how healthcare provider networks are formed and their effects on patient health outcomes. The first chapter explores three types of hospital networks that are intended to improve coordination of patient care across different hospitals: integrated delivery systems, accountable care organizations, and electronic health records. Using 2007-2017 Healthcare Information and Management Systems Society IT data and Medicare data on accountable care organizations and hospital quality, I document several interesting patterns regarding the formation and potential effects of these networks in the United States. I find correlations consistent with assortative matching where higher quality hospitals match with higher quality groups, which may be inefficient if there are peer effects that mean higher quality groups could have more substantial influence on lower quality hospitals that have more room to improve. I show that accountable care organizations appear to be strategic about the network formation process, omitting hospitals that are natural members. They may do so for anticompetitive reasons–ordinary least square regressions find that accountable care organization market concentration is negatively correlated with hospital quality. These regressions additionally point to the need for caution in advocating for a unified electronic health record, as hospital quality is positively correlated with regional electronic health record market concentration–which is related to coordination abilities–but negatively correlated with national concentration–which is related to competition. The second chapter takes inspiration from the descriptive results of the first chapter and establishes a causal effect of electronic health record networks at the patient level. I hypothesize that systematic, reliable transfer of patient medical history can improve clinical decisions and thus health outcomes, especially during medical emergencies. Thus, I identify patients who had emergency cardiovascular episodes in 2007-2014 Medicare claims and use a difference-in-differences strategy to estimate the causal effect of their primary care and emergency hospitals being in the same electronic health record network. I find that electronic health record compatibility decreases the mortality rate but increases the rate of other bad health outcomes by approximately the same amount, suggesting that compatibility makes it easier for patients to survive given poor health but does not overall improve health otherwise. This result highlights the importance of analyzing the effects of healthcare treatments on both the rates of mortality and negative outcomes in survivors. Only looking at the rate of negative outcomes in survivors, electronic health record compatibility would have appeared to be a harmful treatment, while it was actually reducing mortality. The third chapter moves from hospital networks, which have only one type of agent, to look at physician-insurer networks, represented by a two-sided many-to-many matching market. I use Healthgrades and National Committee for Quality Assurance consumer ratings data to collect physician and insurance plan characteristics, respectively. Descriptive statistics indicate that higher quality physicians are in more insurance networks, while higher quality plans tend to be more restricted in the numbers of physicians they accept. There is a mild correlation between physician and plan quality, but there are many possible explanations for it. To test if it is due to assortative matching and to better understand how physicians and insurers decide with whom to contract, I estimate a structural many-to-many matching model using the matching maximum score estimator. Data quality and quantity appear to be obstacles in obtaining precise estimates, so I leave further exploration of this topic to future research.

Complex patient are becoming more and more of a challenge to the health care system given the amount of care they require and the amount of documentation needed to keep track of their state of health and treatment. Record keeping using the EHR makes this easier but mounting amounts of patient data also means that clinicians are faced with information overload. Information overload has been shown to have deleterious effects on care, with increased safety concerns due to missed information. Patient record summarization has been a promising mitigator for information overload. Subsequently, a lot of research has been dedicated to record summarization since the introduction of EHRs. In this dissertation we examine whether unsupervised inference methods can derive patient problem-oriented summaries, that are robust to different patients. By grounding our experiments with HIV patients we leverage the data of a group of patients that are similar in that they share one common disease (HIV) but also exhibit complex histories of diverse comorbidities. Using a user-centered, iterative design process, we design an interactive, longitudinal patient record summarization tool, that leverages automated inferences about the patient's problems. We find that unsupervised, joint learning of problems using correlated topic models, adapted to handle the multiple data types (structured and unstructured) of the EHR, is successful in identifying the salient problems of complex patients. Utilizing interactive visualization that exposes inference results to users enables them to make sense of a patient's problems over time and to answer questions about a patient more accurately and faster than using the EHR alone.