published_or_final_version / Public Health / Master / Master of Public Health

Background: Implementations of national electronic health record (EHR) were currently underway worldwide as a core objective of eHealth strategies. It was widely believed that implementation of EHR might lead to considerable financial savings. This paper aimed to conduct a systematic review to assess return-on-investment (ROI) of HER implementation and to identify areas with greatest potential to positive ROI for ongoing deliberation on continuous development of EHR. Methodology: An inclusive string was developed to search English paper published between January 2003 and June 2013. This paper only included studies meet the following criteria 1) Primary study; 2) Involve a computerized system with electronic health record; and 3) include some form of economic evaluation. Critical appraisal was undertaken and articles with higher quality were selected. Hard ROI and soft ROI defined for EHR implementation were adopted as outcome metrics to examine both tangible and intangible return of EHR implementation. Results: A total of 18 articles were examined for data extraction and synthesis. Most of the available evidences came from pre-post evaluation or cross-sectional analysis without uniform standards for reporting. Findings of 56% of the articles indicated that there is cost saving after EHR implementation while 17% of the articles indicated loss in totalrevenue. The remaining articles concluded that there is no association between cost reduction and EHR implementation. Among the defined hard ROI, most studies mentioned the positive effect in resource reduction. Some authors argued that the resource was reallocated to other initiatives and resulted in negligible cost saving. According to the selected literatures, evidences showed that EHR was able to achieve defined soft ROI, especially for improving caring process, but the overall outcome was subject to individual practice. Authors of 12 out of 18 articles have identified the factor leading to positive return and provided recommendation toward successful EHR implementation. Other than implying helpful EHR functions and promoting practice change, additional incentive on quality improvement and performance benchmarking should be considered. The organizations and EHR systems studied in the articles examined were vastly different; it would be desirable if a controlled study adopting EHR with uniform standards can be performed to evaluate the ROI of different clinical settings. Conclusions: The benefits of EHR are not guaranteed, it requires change of practice and substantial efforts. Healthcare industries have to equip themselves for implementing the new technology and to exploit the usage for better clinical outcome. / published_or_final_version / Public Health / Master / Master of Public Health

An explorative study was done to develop an evaluation methodology. This method can be applied during the development of interactive medical record systems in order to provide information which can be used to improve user interaction with the system. Th e evaluation methodology consists of a number of interactive sessions with potential users of the interactive medical record system. During the first two sessions the subjects are trained to use the system. During the third and last session the subjects are videotaped while they are doing a set of benchmark tasks on the system under evaluation. The video recordings are analysed to obtain performance data. This performance data consists of task timings and a list of problems experienced (errors made) by the subjects. The systems evaluated during the study were a problem-oriented manual medical record and an interactive computerized medical record. The computerized record system was specifically developed for this study. The design and subsequent improvements to this system are documented in the study.

This Ph.D. project started from a broad analysis aiming at investigating the key issues in the development of Information and Communication Technologies (ICT) in the health care sector, with the aim of making an in depth investigation to evaluate the effects of Electronic Medical Record (EMR) implementation on the organizations adopting them. Furthermore the study examined two study settings which have adopted the same EMR system produced by the same provider. This comparative study aims, in particular, to analyse how EMR systems are adopted by different health organizations focusing on the antecedents of the EMR project, on the implementation processes used and on the impacts produced. Diffusion theory, through the lens of socio-technical approach, represents the theoretical framework of the analysis. The research results are based on policy evaluation and case studies. The two hospitals selected for the case study analysis are the Regional Hospital of Local Health Authority in Aosta, Italy and the Royal Infirmary of Edinburgh, Scotland. In conducting the data collection several strategies have been used: documentary analysis, interviews and observations have been carried out. This work provides an overview of the key issues arising over e-health policy development through a comparative analysis of the UK and Italy and provides an insight into how EMR systems are adopted, implemented and evaluated within acute care organizations. The thesis is a comparative international research about the development of e-health and the use of ICT in health care sector. This approach makes a both a theoretical and methodological contribution. By focusing, in particular, on EMR systems, it offers to practitioners and policy makers a better basis of analysing ICT usage and its impacts on health care service delivery.

Objectives: To investigate the impact of electronic health record (EHR) on diabetes management through examination of the effectiveness of implementation of EHR and to improve the quality of care and the cost-effectiveness on the use of EHR. Methods: Three databases, PubMed, Ovid Medline and Google Scholar, were searched with specific combination keywords including electronic medical record and electronic health record, and diabetes. Quality appraisal and extraction of data were conducted on literature that met with the inclusion criteria. Results: 10 literature studies, a total of 204,251 participants with diabetes, were included in this study. All subjects, with similar demographic and clinical characteristics, were from clinic and primary care setting with the use of EHR. Different outcome measures were compared and to evaluate the effectiveness of EHR on quality of care and cost-effectiveness. Discussion: The impact of EHR on effectiveness of diabetes management, potential factors of barrier for adoption and the limitation for implementation of EHR were discussed. These suggested that further research is needed to have stronger evidence to widespread the use of EHR in Hong Kong as a future direction on public health issue. Conclusion: In this systematic review, EHR showed potential benefit in improving the quality of care and reduce the health care expenditure for long term running. Patient safety and efficiency are yet to be covered in the studies. Further research is needed on the acceptability and applicability of the use of EHR in Hong Kong. / published_or_final_version / Public Health / Master / Master of Public Health

No description available.

Patients typically enter the healthcare systems at the primary care level from where they are further referred to specialists or hospitals as necessary. In the private healthcare system, primary care is provided by a general practitioner (GP). A GP will refer a patient to a specialist for treatment when necessary, while the GP remains the main healthcare provider. The provision of care is, thus, fragmented which results in continuity of care becoming a challenge. Furthermore, the majority of healthcare providers continue to use paper-based systems to capture and store patient medical data. However, capturing and storing patient medical data via electronic methods, such as Electronic Medical Records (EMRs), has been found to improve continuity of care. Despite this benefit, research reveals that smaller practices are slow to adopt electronic methods of record keeping. Hence this explorative research attempts to identify factors that affect the lack of adoption and meaningful use of EMRs in general practices. Four general practices are surveyed through patient and staff questionnaires, as well as GP interviews. Socio-Technical Systems (STS) theory is used as a theoretical lens to formulate the resulting factors. The findings of the research indicate specific factors that relate to either the social, environmental or technical sub-systems of the socio-technical system, or an overlap between these sub-systems. It is significant to note that within these sub-systems, the social sub-system plays a key role. This is due to various reasons revealed by this research. Furthermore, multiple perceptions emerged from the GP and patient participants during the analysis of the findings. These perceptions may have an influence on the adoption and potential meaningful use of an EMR in a general practice. Additionally, the socio-technical factors identified from this research highlight the challenges related to encouraging the adoption and meaningful use of EMRs. These challenges are introduced by the complexities represented by these factors. Nevertheless, addressing the factors will contribute towards improving the rate of adoption and meaningful use of EMRs in small practices.

No description available.

The theme of this dissertation focuses on methods for estimating personalized treatment using machine learning algorithms leveraging information from electronic health records (EHRs). Current guidelines for medical decision making largely rely on data from randomized controlled trials (RCTs) studying average treatment effects. However, RCTs are usually conducted under specific inclusion/exclusion criteria, they may be inadequate to make individualized treatment decisions in real-world settings. Large-scale EHR provides opportunities to fulfill the goals of personalized medicine and learn individualized treatment rules (ITRs) depending on patient-specific characteristics from real-world patient data. On the other hand, since patients' electronic health records (EHRs) document treatment prescriptions in the real world, transferring information in EHRs to RCTs, if done appropriately, could potentially improve the performance of ITRs, in terms of precision and generalizability. Furthermore, EHR data domain usually consists text notes or similar structures, thus topic modeling techniques can be adapted to engineer features. In the first part of this work, we address challenges with EHRs and propose a machine learning approach based on matching techniques (referred as M-learning) to estimate optimal ITRs from EHRs. This new learning method performs matching method instead of inverse probability weighting as commonly used in many existing methods for estimating ITRs to more accurately assess individuals' treatment responses to alternative treatments and alleviate confounding. Matching-based value functions are proposed to compare matched pairs under a unified framework, where various types of outcomes for measuring treatment response (including continuous, ordinal, and discrete outcomes) can easily be accommodated. We establish the Fisher consistency and convergence rate of M-learning. Through extensive simulation studies, we show that M-learning outperforms existing methods when propensity scores are misspecified or when unmeasured confounders are present in certain scenarios. In the end of this part, we apply M-learning to estimate optimal personalized second-line treatments for type 2 diabetes patients to achieve better glycemic control or reduce major complications using EHRs from New York Presbyterian Hospital (NYPH). In the second part, we propose a new domain adaptation method to learn ITRs in by incorporating information from EHRs. Unless assuming no unmeasured confounding in EHRs, we cannot directly learn the optimal ITR from the combined EHR and RCT data. Instead, we first pre-train “super" features from EHRs that summarize physicians' treatment decisions and patients' observed benefits in the real world, which are likely to be informative of the optimal ITRs. We then augment the feature space of the RCT and learn the optimal ITRs stratifying by these features using RCT patients only. We adopt Q-learning and a modified matched-learning algorithm for estimation. We present theoretical justifications and conduct simulation studies to demonstrate the performance of our proposed method. Finally, we apply our method to transfer information learned from EHRs of type 2 diabetes (T2D) patients to improve learning individualized insulin therapies from an RCT. In the last part of this work, we report M-learning proposed in the first part to learn ITRs using interpretable features extracted from EHR documentation of medications and ICD diagnoses codes. We use a latent Dirichlet allocation (LDA) model to extract latent topics and weights as features for learning ITRs. Our method achieves confounding reduction in observational studies through matching treated and untreated individuals and improves treatment optimization by augmenting feature space with clinically meaningful LDA-based features. We apply the method to extract LDA-based features in EHR data collected at NYPH clinical data warehouse in studying optimal second-line treatment for T2D patients. We use cross validation to show that ITRs outperforms uniform treatment strategies (i.e., assigning insulin or another class of oral organic compounds to all individuals), and including topic modeling features leads to more reduction of post-treatment complications.

Structured electronic health record (EHR) data are commonly incomplete and can lack diagnostic detail. Clinical reports, on the other hand, are typically comprehensive and contain a wealth of detailed medical information. Pathologists invest considerable time and specialized training to create information-rich pathology reports, but the necessary manual review of these reports for clinical or research use is a high barrier to their routine utilization. The automated extraction of clinical targets directly from pathology reports would allow for the structured aggregation of relevant patient data that are not currently routinely captured in the EHR. In this dissertation, I apply recently developed transformer models to predict clinical targets from cancer pathology report text. In the first chapter, I present a pathology report corpus that I fully processed and made publicly available, and perform a proof-of-concept cancer type classification. In the second chapter, I discuss a set of cancer stage classification models that I fine-tune on the pathology report corpus and then externally validate on reports from Columbia University Irving Medical Center (CUIMC). In the last chapter, I explore additional applications for this methodology, developing a generalizable model to classify prostate cancer reports into primary Gleason score categories, applying a transformer model to classify reports into diagnosis categories for a Barrett’s esophagus patient cohort in a low-data environment, and performing a proof-of-concept prediction of adverse drug events from 1D drug representations.