Distributional semantics allows models of linguistic meaning to be derived from observations of language use in large amounts of text. By modeling the meaning of words in semantic (vector) space on the basis of co-occurrence information, distributional semantics permits a quantitative interpretation of (relative) word meaning in an unsupervised setting, i.e., human annotations are not required. The ability to obtain inexpensive word representations in this manner helps to alleviate the bottleneck of fully supervised approaches to natural language processing, especially since models of distributional semantics are data-driven and hence agnostic to both language and domain. All that is required to obtain distributed word representations is a sizeable corpus; however, the composition of the semantic space is not only affected by the underlying data but also by certain model hyperparameters. While these can be optimized for a specific downstream task, there are currently limitations to the extent the many aspects of semantics can be captured in a single model. This dissertation investigates the possibility of capturing multiple aspects of lexical semantics by adopting the ensemble methodology within a distributional semantic framework to create ensembles of semantic spaces. To that end, various strategies for creating the constituent semantic spaces, as well as for combining them, are explored in a number of studies. The notion of semantic space ensembles is generalizable across languages and domains; however, the use of unsupervised methods is particularly valuable in low-resource settings, in particular when annotated corpora are scarce, as in the domain of Swedish healthcare. The semantic space ensembles are here empirically evaluated for tasks that have promising applications in healthcare. It is shown that semantic space ensembles – created by exploiting various corpora and data types, as well as by adjusting model hyperparameters such as the size of the context window and the strategy for handling word order within the context window – are able to outperform the use of any single constituent model on a range of tasks. The semantic space ensembles are used both directly for k-nearest neighbors retrieval and for semi-supervised machine learning. Applying semantic space ensembles to important medical problems facilitates the secondary use of healthcare data, which, despite its abundance and transformative potential, is grossly underutilized. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 4 and 5: Unpublished conference papers.</p> / High-Performance Data Mining for Drug Effect Detection

Racial/ethnic disparities in cancer care are well documented in the research literature; however, less is known about the extent and potential source of cancer care disparities in the Veterans Health Administration (VA). In my first paper, I use logistic regression and hospital fixed effects models to examine racial disparities in 20 cancer-related quality measures and the extent to which racial differences in site of care explain VA cancer care disparities. I found evidence of racial disparities in 7 out of 20 cancer-related quality measures. In general, these disparities were primarily driven by racial differences in care for black and white patients within the same VA hospital, rather than racial differences in site of care.

Despite rapid advances in technology, there is currently a complex, and somewhat disjointed approach to the way health information is collected, stored, and organized for both healthcare consumers and professionals. Incompatible electronic medical records from various healthcare providers add to the complexity of a system tasked with delivering a patient’s relevant medical information in a timely manner to the appropriate point of care. Personal health records (PHR) grew out of the efforts to produce an integrated electronic record to manage the multifaceted aspects of healthcare required by both healthcare consumers and professionals. PHRs are a transformative technology with the potential to alter patient-provider relationships in a way that produces a more efficient and cost effective healthcare system as a result of better patient outcomes. PHRs can potentially include a wide variety of users ranging from the lay public to clinical professionals. As such, it is important to identify potential user groups and their corresponding health information needs in order to design PHRs that maximize accessibility, usability, and clinical relevance. This study focused on laypeople who represented a wide age-range of individuals, evenly split in gender, with an above average level of computer literacy. Most of the participants had not used an electronic PHR prior to this study. However, after a hands-on session with PHR software, most participants found it to be easy to use, accompanied with the functionality they expected from such a system. Most participants were satisfied that an electronic PHR would meet their health information needs and would recommend the use of PHRs to family and friends. Anyone in the general public is a potential PHR user. However, this study found that individuals with chronic conditions and those with complex health needs had the most to gain from using a PHR as an integral part of their healthcare routine. This study also demonstrated that an individual’s health condition has a stronger influence on their perceptions about the usefulness of PHRs than does their demographic background (age, education, computer literacy). Finally, this study established that PHRs are considered by participants of the study to be useful tools in meeting their health information needs. / Graduate / 0723 / 0769 / 0984 / dguarin@uvic.ca

The two longitudinal case studies that make up this dissertation sought to explain and predict the relationship between usability and clinician acceptance of a health information system. The overall aim of the research study was to determine what role usability plays in the acceptance or rejection of systems used by clinicians in a healthcare context. The focus was on the end users (the clinicians) rather than the views of the system designers and managers responsible for implementation and the clients of the clinicians. A mixed methods approach was adopted that drew on both qualitative and quantitative research methods. This study followed the implementation of a community health information system from early beginnings to its established practice. Users were drawn from different health service departments with distinctly different organisational cultures and attitudes to information and communication technology used in this context. This study provided evidence that a usability analysis in this context would not necessarily be valid when the users have prior reservations on acceptance. Investigation was made on the initial training and post-implementation support together with a study on the nature of the clinicians to determine factors that may influence their attitude. This research identified that acceptance of a system is not necessarily a measure of its quality, capability and usability, is influenced by the user’s attitude which is determined by outside factors, and the nature and quality of training. The need to recognise the limitations of the current methodologies for analysing usability and acceptance was explored to lay the foundations for further research.

Background: UK primary care accounts for 90% of patient contacts in the NHS, and over 300 million consultations every year. Consequently, when primary is suboptimal it has important impacts on population health. At the same time, virtually all general practices use electronic health records (EHR) to capture patient data. Clinical Decision Support (CDS) systems use it to highlight when individual patients do not receive care consistent with clinical guidelines, though ignore the wider population. Electronic Audit and Feedback (e-A&F) systems address the wider population, but their results are difficult to interpret. EHR data has the richness to suggest ways in which care quality could be improved, though this is currently not exploited. The aim of this thesis was to make progress towards better use of primary care EHR data for the purposes of quality improvement (QI) by focusing on e-A&F as a vehicle. Research Objectives were: 1) Develop a model and recommendations to guide EHR data analysis and its communication to health professionals; 2) Use these models and recommendations to develop a system for UK primary care; 3) Implement and evaluate the system to test the models and recommendations, and derive generalisable knowledge. Methods: The overall approach of this thesis was informed by guidance from the Medical Research Council on the development of complex interventions, and Boyrcki et al.’s evidence-based framework for the development of health information technologies (Chapter 2). Theory was first identified through a critical examination of the empirical and theoretical literature regarding CDS and e-A&F systems (Chapter 3), then built upon in a systematic literature search and metasynthesis of qualitative studies of A&F (and e-A&F) interventions (Chapter 4). This resulted in the development a new theory of A&F (Clinical Performance Feedback Intervention Theory; CP-FIT), which was used to inform the development of an e-A&F system for UK primary care – the Performance Improvement plaN GeneratoR (PINGR; version 1). PINGR was then iteratively optimised through a series of three empirical studies. First, its usability was evaluated by software experts using Heuristic Evaluation and Cognitive Walkthrough methodologies (Chapter 5). GPs then performed structured tasks using the system in a laboratory whilst their on-screen interactions and eye movements were recorded (Chapter 6). Finally, PINGR was implemented in 15 GP practices, and CP-FIT used to guide the mixed methods evaluation including examinations of usage records, and interviews with 38 health professionals. Results: There are both empirical and theoretical arguments for combining features from CDS and e-A&F systems to increase their effectiveness; a key recommendation is that e-A&F systems should suggest clinical actions to health professionals (Chapter 3). This is supported by CP-FIT, which has three core propositions: 1) A&F interventions exert their effects through health professionals taking action; 2) Health care organisations have limited capacity to engage with A&F; and 3) Health care professionals and organisations have a strong set of beliefs and behaviours regarding how they provide patient care (Chapter 4). Based on these findings, the unique feature of PINGR is that it suggests improvement actions to users based on EHR data analysis (‘decision-supported feedback’). Key findings from PINGR’s usability evaluation with software experts translated into a set of design guidelines for e-A&F interfaces regarding: summarising clinical performance, patient lists, patient-level information, and suggested actions (Chapter 5). When tested with GPs, these guidelines were found to impact: user engagement; actionability; and information prioritisation (Chapter 6). Following its implementation in practice, PINGR was used on 227 occasions to facilitate the care of 725 patients. These patients were 1.8 (95% CI 1.6-1.9) times more likely to receive improved care according to at least one clinical guideline. Barriers and facilitators to its success included: the resources available to use it; its perceived relative advantages; how compatible it was with pre-existing beliefs and ways of working; the credibility of its data; the complexity of the clinical problems it highlighted; and the ability to act on its recommendations (Chapter 7). Conclusion: It is both feasible and acceptable to health professionals to make better use of EHR data for QI by enabling e-A&F systems to suggest actions for them to take. When designing e-A&F interfaces, attention should be paid to how they summarise clinical performance, and present patient lists and detailed patient-level information. Implementation of e-A&F interventions is influenced by availability of resources, compatibility with existing workflows, and ability to take action based on their feedback results. Unresolved tensions exist regarding how they may deal with patient complexity. Policymakers should consider the relevance of these findings for National Clinical Audits and pay-for-performance initiatives.

In this study health information seeking behaviour among young adults is investigated and attitudes toward available health information examined. Further, the electronic health record(EHR), ‘Journalen’, is investigated among young adults that use the health services in Sweden. The aim of this study is to gain insight in young adults’ health information seeking behaviour on the Internet.   A survey of health information seeking behaviour among young adults, 18-29 years old, reports that as well as searching for information for their own use, searching for information to help family and friends is also their practice. 95.4% say that they search for health information on the Web.   Overall, the top five categories of information searched are Personal Health, Symptoms, Healthy Diet, Mental Health and Treatment. For the Top five Information searches for themselves(’Self’), Mental Health is displaced by Disease Prevention. Health information categories sought the least by the young adults are information about Illegal Drugs, Allied Health, Alternative medicine and Support Groups. Respondents mostly report a high regard for the information they retrieve in their searches.   The majority rate the quality of health information as ‘Good’ or ‘Very Good’. Broadly speaking the young adults trust the health information found online but say that the quality of health information varies from source to source. 1177.se is often referred to as a trustworthy source among the respondents and according to them good quality information can be found there. The respondents use of support groups for themselves is reported, as well an altruistic motivation to support others. While the motivation for many searches is to avoid attending a health professional, approximately one-third report searching after their visit.   This study demonstrates that young adults are active consumers of health information on the Internet and their attitudes toward the available health information are generally positive. Of the respondents who are Swedish residents, 42% are unaware of the availability of their electronic health record in Sweden, Journalen, and of those 56.7% who know of it, 18% have never accessed their EHRs.  More research is needed to investigate the reason for the relatively low numbers of the young adults’ awareness and use of the Journelen. The aim of the study is discussed in the Introduction chapter and research questions identified. The research design is described in the methodology chapter and how data is collected through a survey. Results are presented in Tables and Figures and further analysed in the Analysis chapter in context of earlier research. Finally, the researcher concludes on the study and what could be learned from the findings.

Submitted by Maria Suzana Diniz (msuzanad@hotmail.com) on 2015-11-13T13:37:00Z No. of bitstreams: 1 arquivototal.pdf: 3391275 bytes, checksum: 21fa5584131c69274496290dee7e6d52 (MD5) / Made available in DSpace on 2015-11-13T13:37:00Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 3391275 bytes, checksum: 21fa5584131c69274496290dee7e6d52 (MD5) Previous issue date: 2015-02-20 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Introduction: it is undeniable that the evolution of mobile computing devices, associated with the development of capacities for operation in wireless network, potencializes new ways to use them for nursing care provision, enabling to extend information systems up with the patient. Objectives: to developing an application for mobile device that allows data logging for systematization of nursing care in the Neonatal Intensive Care Unit (NICU) and implementing the prototype allowing to identify the nurses' perception about their use. Method: this is a methodological research of applied type, developed in the NICU of the Lauro Wanderley University Hospital, located in the city of João Pessoa, PB. The main technologies used for the development of software were the Ruby programming language, Ruby on Rails and JavaScript; Bootstrap framework; Ubuntu Linux production server, Nginx Webserver and PostgreSQL Database Management System. The software was developed in the period from May to September 2014 and used by nurses for thirty days, from 9th September to 9th October 2014. The study included five nurses who develop their activities in the NICU. To evaluating the application there was developed a semi-structured questionnaire and the results analyzed through the content thematic analysis of the method Bardin. There were obeyed guidelines inherent in the research protocol contained in Resolution No. 466/12, National Health Council (CNS). The project was approved and registered with the National Information System on Ethics in Research involving Human Subjects, with CAAE-25890914.5.0000.5183. Results: with respect to mobile technology used, in general, nurses do not present difficulties, because they have familiarity with this type of device. Those who experienced difficulties with the frequent use of the system, the obstacles were minimized. Overall, it is emphasized that the application includes features for SAE in NICU, but changes are suggested as the interface of screens and some terms of data collection and parameters available by the applicable. In the view of the study participants, the main contributions of software for mobile technology are: agility to develop and document the SAE, freedom of movement, evaluation of standardization of the newborn, time optimization in carrying out bureaucratic activities, ability to retrieve information and reducing the physical space occupied. Conclusion: therefore, a system for SAE in mobile technology allows flexibility in carrying out registration activities by the nurse, since the data collection can be performed at the bedside; in addition, offers numerous advantages to this professional work routine. / Introdução: É inegável que a evolução dos dispositivos de computação móvel, associada ao desenvolvimento das capacidades para operar em rede sem fios, potencializa novas formas de utilizá-los na prestação dos cuidados de enfermagem e estende os sistemas de informação até o paciente. Objetivos: Desenvolver um aplicativo para dispositivo móvel que possibilite o registro de dados para a Sistematização da Assistência de Enfermagem em Unidade de Terapia Intensiva Neonatal (UTIN) e implementar o protótipo para identificar a percepção dos enfermeiros quanto ao seu uso. Método: Trata-se de uma pesquisa metodológica do tipo aplicada, desenvolvida na UTIN do Hospital Universitário Lauro Wanderley, localizado no município de João Pessoa-PB. As principais tecnologias utilizadas para o desenvolvimento do software foram: a linguagem de programação Ruby, Ruby on Rails e JavaScript; o framework Bootstrap; o servidor de produção Ubuntu Linux, Nginx Webserver e o Sistema de Gerenciamento de Banco de Dados PostgreSQL. O software foi desenvolvido no período de maio a setembro de 2014 e utilizado pelas enfermeiras durante trinta dias - de 09 de setembro a 09 de outubro de 2014. Participaram do estudo cinco enfermeiras que desenvolvem suas atividades na UTI Neonatal. Para avaliar o aplicativo, foi elaborado um questionário semiestruturado, e os resultados, analisados através do método de análise temática de conteúdo de Bardin. A pesquisa obedeceu às orientações inerentes ao protocolo de pesquisa contido na Resolução nº 466/12 CNS. O projeto foi aprovado e registrado no Sistema Nacional de Informação sobre Ética em Pesquisa envolvendo Seres Humanos com CAAE-25890914.5.0000.5183. Resultados: No que se refere à tecnologia móvel utilizada, em geral, as enfermeiras não apresentaram dificuldades, pois se familiarizam com esse tipo de dispositivo. Quanto às que sentiram dificuldades, com o frequente uso do sistema, os obstáculos foram minimizados. De maneira geral, o aplicativo dispõe das funções necessárias para a SAE em UTI Neonatal, mas se sugerem alterações quanto à interface das telas e alguns termos da coleta de dados e parâmetros disponibilizados pelo aplicativo. Na opinião dos participantes do estudo, as principais contribuições do software para tecnologia móvel são: agilidade em desenvolver e documentar a SAE, liberdade de movimentação, padronização da avaliação do recém-nascido, otimização de tempo na realização de atividades burocráticas, possibilidade de resgatar informações e redução do espaço físico ocupado. Conclusão: Um sistema para SAE em tecnologia móvel possibilita a flexibilidade na realização das atividades de registro pelo enfermeiro, uma vez que a coleta de dados pode ser realizada à beira do leito. Além disso, proporciona inúmeras vantagens ao cotidiano de trabalho desse profissional.

Made available in DSpace on 2015-05-14T12:36:27Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2547720 bytes, checksum: 27f87fb12b3a5d38cd81b4df2f2fd067 (MD5) Previous issue date: 2011-02-07 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Health Informatics is an innovative area which presents major challenges to be overcome. As one of its main branches of research are Electronic Health Records (EHR), which are responsible, among other things, the storage, display and man-agement of the patient's clinical records. This work is motivated primarily by two main aspects that have been gaining relevance in EHR context: Application domain flexibility and multi-device presentation. Domain flexibility inherent in the field of electronic health records is a key challenge faced by user interface designers today. The use of such records on a large scale in different enviroments and the increasing need for use of tools to clinical decision support are relevant factors too. Approaches in the literature propose generic models of domain and presentation development, with manual configurable mappings between both layers, but this configuration is very costly and error-prone. This work aims at developing a tool for GUI automatic generation for health documents, titled MedViewGen, providing independence from the scope and interface technologies. By combining metadata extracted from appli-cation domain with metadata attached to user interface components, both represented in OWL, we create the possibility to decide at run time, which compo-nents are most suitable to represent each biomedical concept. As a result we have an instantiation of the tool to support the JSF technology, and integrated to OpenCTI it presented satisfactory results in forms generation, using a set of interface compo-nents suitable for documents representation and some clinical decision support agents. / A informática em saúde é uma área inovadora e que apresenta grandes desafios a serem vencidos. Como um de seus principais ramos de pesquisa encontram-se os Registros Eletrônicos em Saúde (Electronic Health Records, EHR), que são respon-sáveis, dentre outros, pelo armazenamento, exibição e manipulação de registros clí-nicos do paciente. Este trabalho é motivado por dois aspectos principais que vêm ganhando relevância no contexto de sistemas para EHR: flexibilidade do domínio da aplicação e apresentação multi-dispositivo. A flexibilidade do domínio inerente aos registros eletrônicos em saúde é um dos principais desafios enfrentados pelos proje-tistas de interface com o usuário atualmente. A utilização desses registros em larga escala em diversas plataformas de visualização e a crescente necessidade por utiliza-ção de ferramentas de suporte a decisão clínica são fatores relevantes. Abordagens presentes na literatura propõem modelos genéricos de representação de domínio e apresentação, com mapeamentos configuráveis entre as duas camadas, porém essa configuração é um processo oneroso e susceptível a erros. Este trabalho tem como objetivo o desenvolvimento de uma ferramenta de geração automática da GUI para documentos em saúde, entitulada MedViewGen, proporcionando independência do domínio de aplicação e de tecnologias de interface. Através da combinação dos me-tadados extraídos do domínio com metadados intrínsecos aos componentes de inter-face, ambos representados em ontologias OWL, criamos a possibilidade de mapear, em tempo de execução, quais componentes estão aptos para representar cada con-ceito biomédico na GUI. Como resultado tivemos uma instanciação da ferramenta para suportar a tecnologia JSF, integrando com OpenCTI e apresentando resultados satisfatórios na geração dos formulários, utilizando um conjunto de componentes de interface adequado para a representação dos documentos e alguns agentes de supor-te a decisão clínica.

The aim of my work was to find out the specific features of Czech medical reports in terms of the possibility of extracting specific information from them. For my work, I had a total of 268 anonymized narrative medical reports from two outpatient departments. I have studied standards for preserving electronic health records and for transferring clinical information between healthcare information systems. I have also participated in the process of implementing electronic medical record in the field of dentistry. First of all, I tried to process narrative medical reports using natural language processing (NLP) tools. I came to the conclusion that narrative medical reports in the Czech language are very different than a typical Czech text, especially because it mostly contains short telegraphic phrases and the texts lack typical Czech sentence structure. It also contains many misspellings, acronyms and abbreviations. Another problem was the absence of existence of the Czech translation of the main international classification systems. Therefore I decided to continue the research by developing the method for pro-processing the input text for translation and its semantic annotation. The main objective of this part of the research was to propose a method and support software for interactive correction...

As Information Communication Technologies (ICTs) are growing, there have been ceaseless efforts to develop a National Health Information Infrastructure (NHII). One of the challenges in constructing a NHII is concerned with the management of Electronic Health Records (EHRs). In particular, exchanging EHRs is an important factor in establishing interoperability within a NHII, and the reusability of the functionality for exchanging EHRs is one of major solutions to construct an NHII. In this study, we obtain several component models, and conduct empirical studies to validate the component models in terms of component reusability. Using HL7 CDA (Health Level 7 Clinical Document Architecture) as an EHR standard, we implemented three prototypes of the EHR Exchanger based on JavaBeans, the exogenous connectors and the mediator connector respectively. As shown in the experiment results, the reuse approach using a mediator connector leads to better component reusability in terms of external dependency, total coupling between objects (CBO), additional lines of codes (LOC), and performance. Thus, we believe that the reuse approach using a mediator connector yields many benefits in terms of component reusability for the EHR Exchanger implementation.