Ebook: Informatics for Health: Connected Citizen-Led Wellness and Population Health

Founded in 1976, the European Federation for Medical Informatics (EFMI) is the leading organisation in medical informatics in Europe, representing 32 countries. EFMI is a non-profit organisation concerned with the theory and practice of information science and technology within health and health science in a European context. EFMI's objectives are:

• To advance international co-operation and dissemination of information in medical informatics at the European level;

• To promote high standards in the application of medical informatics;

• To promote research and development in medical informatics;

• To encourage high standards in education in medical informatics; and

• To function as the autonomous European Regional Council of IMIA, the International Medical Informatics Association.

In recent years, Europe has seen major investments in research infrastructure for harnessing the potential of routinely collected health data. In the UK, this has led to the establishment in 2013 of The Farr Institute of Health Informatics Research. Funded by the UK's Medical Research Council (MRC) and nine other funders, The Farr Institute is comprised of 21 academic institutions, two MRC units and public bodies such as Public Health England, Public Health Wales and NHS National Services Scotland. Together they form a national research collaboration led by four regional centres: Farr Institute CIPHER, Farr Institute HeRC, Farr Institute London, and Farr Institute Scotland. The Farr Institute aims to be a global leader in health informatics research through scientific discovery and the enhancement of patient and public health. By analysing data from multiple sources and collaborating with the government, public sector, academia and industry, The Farr Institute will unleash the value of vast sources of clinical, biological, population and environmental data for public benefit.

The Medical Informatics Europe (MIE) conference, organised by EFMI, is a key event in the EFMI calendar. The first conference took place in Cambridge, UK in 1978 and now takes place annually. The Farr Institute has also been establishing its own conference series, with the first Farr International Conference taking place in St. Andrews, Scotland, in 2015, followed in 2016 by a second conference held in Swansea, Wales in collaboration with the International Population Data Linkage Network. For 2017, the decision was made to combine the power and established reputational excellence of EFMI with the emerging and innovative research within The Farr Institute community. EFMI, The Farr Institute, and the British Computer Society have worked together to organise Informatics for Health 2017, a joint conference that combines MIE and the Farr International Conference, creating a scientific forum that allows these two communities to share knowledge, insights, and experience, to advance cross-disciplinary thinking, and to stimulate creativity. The conference took place in the city of Manchester in the UK from the 24th to the 26th of April.

The conference received a total of 404 submissions, in the form of both full papers and abstracts, for oral presentation at the conference. The current volume presents the 116 full papers that were presented at the conference, with contributions from 28 different countries. Abstracts that were presented at the conference can be found in Scott, P.J. et al. (2017). Informatics for Health 2017: Advancing both science and practice, J. Innov. Health Inform., 24(1).

We would like to thank all of those who contributed to these proceedings and the success of this important event, in particular the authors who chose to share their work and the reviewers who generously gave their time and expertise. Special thanks go to the Local Organising Committee for their work in organising such a great event.

Rebecca Randell

Ronald Cornet

Colin McCowan

Niels Peek

Philip J. Scott

Emerging technologies show great potential in the field of patient care. One such technology is mobile heath applications (mhealth apps), which have exploded in number and variety in recent years, and offer great promise in the ability to collect and monitor patient health data. Despite their apparent success in proliferation and user adoption, these applications struggle to integrate into the primary care system and there is scant information regarding their efficacy to effect patient behavior and consequently health outcomes. In this paper we investigate the potential of a promising clinical evaluation methodology, response adaptive randomized clinical trials, to rapidly and effectively evaluate the efficacy and effectiveness of mhealth apps and to personalize mhealth app selection to individualize patient benefit. Diabetes prevention provides the use case for evaluating the case for and against response-adaptive randomized trials.

Case-based reasoning and data interpretation is an artificial intelligence approach that capitalizes on past experience to solve current problems and this can be used as a method for practical intelligent systems. Case-based data reasoning is able to provide decision support for experts and clinicians in health systems as well as lifestyle systems. In this project we were focusing on developing a solution for healthy ageing considering daily activities, nutrition as well as cognitive activities. The data analysis of the reasoner followed state of the art guidelines from clinical practice. Guidelines provide a general framework to guide clinicians, and require consequent background knowledge to become operational, which is precisely the kind of information recorded in practice cases; cases complement guidelines very well and helps to interpret them. It is expected that the interest in case-based reasoning systems in the health.

Achieving interoperability in health is a challenge and requires standardization. The newly developed HL7 standard: Fast Healthcare Interoperability Resources (FHIR) promises both flexibility and interoperability. This study investigates the feasibility of expressing a Danish microbiology message model content in FHIR to explore whether complex in-use legacy models can be migrated and what challenges this may pose. The Danish microbiology message model (the DMM) is used as a case to illustrate challenges and opportunities accosted with applying the FHIR standard. Mapping of content from DMM to FHIR was done as close as possible to the DMM to minimize migration costs except when the structure of the content did not fit into FHIR. From the DMM a total of 183 elements were mapped to FHIR. 75 (40.9%) elements were modeled as existing FHIR elements and 96 (52.5%) elements were modeled as extensions and 12 (6.6%) elements were deemed unnecessary because of build-in FHIR characteristics. In this study, it was possible to represent the content of a Danish message model using HL7 FHIR.

Introduction: The aim of the paper is to establish the requirements and methodology for the development process of GreyMatters, a memory clinic system, outlining the conceptual, practical, technical and ethical challenges, and the experiences of capturing clinical and research oriented data along with the implementation of the system. Methods: The methodology for development of the information system involved phases of requirements gathering, modeling and prototype creation, and ‘bench testing’ the prototype with experts. The standard Institute of Electrical and Electronics Engineers (IEEE) recommended approach for the specifications of software requirements was adopted. An electronic health record (EHR) standard, EN13606 was used, and clinical modelling was done through archetypes and the project complied with data protection and privacy legislation. Results: The requirements for GreyMatters were established. Though the initial development was complex, the requirements, methodology and standards adopted made the construction, deployment, adoption and population of a memory clinic and research database feasible. The electronic patient data including the assessment scales provides a rich source of objective data for audits and research and to establish study feasibility and identify potential participants for the clinical trials. Conclusion: The establishment of requirements and methodology, addressing issues of data security and confidentiality, future data compatibility and interoperability and medico-legal aspects such as access controls and audit trails, led to a robust and useful system. The evaluation supports that the system is an acceptable tool for clinical, administrative, and research use and forms a useful part of the wider information architecture.

In-home monitoring systems have been proposed to support aging in place and facilitate home care service. Through a qualitative approach the study explores nurses' existing challenges and perspectives of an in-home monitoring system. Results indicate that nurses base care decisions on multiple, and sometimes, unreliable information sources. However, access to information about elderlies' physical motion could support the care planning process by reducing ambiguity and raising attention. Hence, a simple and affordable system that largely relies on nurses to interpret the sensed data could bring additional value.

A recent trend in healthcare is to motivate patients to self-manage their health conditions in home-based settings. Medication adherence is an important aspect in disease self-management since sub-optimal medication adherence by the patient can lead to serious healthcare costs and discomfort for the patient. In order to alleviate the limitations of self-reported medication adherence, we can use ambient assistive living (AAL) technologies in smart environments. Activity recognition services allow to retrieve self-management information related to medication adherence in a less intrusive way. By remotely monitor compliance with medication adherence, self-management program's interventions can be tailored and adapted based on the observed patient's behaviour. To address this challenge, we present an AAL framework that monitor activities related to medication adherence.

Clinical reading centers provide expertise for consistent, centralized analysis of medical data gathered in a distributed context. Accordingly, appropriate software solutions are required for the involved communication and data management processes. In this work, an analysis of general requirements and essential architectural and software design considerations for reading center information systems is provided. The identified patterns have been applied to the implementation of the reading center platform which is currently operated at the Center of Ophthalmology of the University Hospital of Tübingen.

The DICOM Standard has been fundamental for ensuring the interoperability of Picture Archive and Communications Systems (PACS). By compiling rigorously to the standard, medical imaging equipment and applications from different vendors can share their data, and create integrated workflows which contributes to better quality healthcare services. However, DICOM is a complex, flexible and very extensive standard. Thus, it is difficult to attest the conformity of data structures produced by DICOM applications resulting in unexpected behaviors, errors and malfunctions. Those situations may be critical for regular PACS operation, resulting in serious losses to the healthcare enterprise. Therefore, it is of paramount importance that application vendors and PACS administrators are confident that their applications follow the standard correctly. In this regard, we propose a method for validating the compliance of PACS application with the DICOM Standard. It can capture the intricate dependency structure of DICOM modules and data elements using a relatively simple description language. The modular nature of our method allows describing each DICOM module, their attributes, and dependencies on a re-usable basis. As a result, our validator is able to encompass the numerous modules present in DICOM, as well as keep up with the emergence of new ones.

This paper proposes a decision support system for screening pediatric cardiac disease in primary healthcare centres relying on the heart sound time series analysis. The proposed system employs our processing method which is based on the hidden Markov model for extracting appropriate information from the time series. The binary output resulting from the method is discriminative for the two classes of time series existing in our databank, corresponding to the children with heart disease and the healthy ones. A total 90 children referrals to a university hospital, constituting of 55 healthy and 35 children with congenital heart disease, were enrolled into the study after obtaining the informed consent. Accuracy and sensitivity of the method was estimated to be 86.4% and 85.6%, respectively, showing a superior performance than what a paediatric cardiologist could achieve performing auscultation. The method can be easily implemented using mobile and web technology to develop an easy-to-use tool for paediatric cardiac disease diagnosis.

Mobile health is fast evolving into a practical solution to remotely monitor high-risk patients and deliver timely intervention in case of emergencies. Building upon our previous work on a fast and power efficient summarization framework for remote health monitoring applications, called RASPRO (Rapid Alerts Summarization for Effective Prognosis), we have developed a real-time criticality detection technique, which ensures meeting physician defined interventional time. We also present the results from initial testing of this technique.

In order to provide for best possible child health care, timely access to all relevant medical data is of vital importance. The aim of this study is to investigate the use of unique identifiers, a key instrument in this regard, in the countries of Europe. A survey was carried out in all 28 European Member States plus 2 European Economic Area countries in 2015, and refreshed in 2016. In 23 countries unique identifiers are used to link children's health records. Five countries indicated they currently do not link child health records, and two have no such plans. There is variety as regards the type of number and the issuing process.

Harnessing the power of IT solutions in child primary care requires strategic thought at national level, and good health care delivery needs this support. The aim of this study was to investigate whether children's needs are considered in national e-health strategies in Europe. In 2016, a survey was carried out in all 28 European Member States plus 2 European Economic Area countries. Sixteen countries fail to mention children's needs at all. Only eleven of 27 countries mention children and adolescents in their national e-health strategy documents ranging from mere data protection concerns to comprehensive IT approaches for the improvement of child primary care.

Engaging patients in the self-management decision-making provides opportunities for positive health outcomes. The process of shared decision making (SDM) is touted as the pinnacle of patient-centred care, yet it has been difficult to implement in practice. Access to tools resulting from the integration of all health data and clinical evidence, and an ease of communications with care providers are needed to engage patients in decision making. Personal health record (PHR) technology is a promising approach for overcoming such barriers. Yet there is a scarcity of studies on system design for SDM via PHR. This paper describes a study protocol to identify functional requirements of PHR for facilitating SDM and factors that would influence the embedding of the proposed system in clinical practice.

Successful medication adherence particularly in elderly with chronic diseases will improve their self-management. Medication reminder systems could be useful to improve this adherence. This study consists of two phases, designing a mobile medical app based on Android platform and then its evaluation. To develop this application, first, the use case scenarios have been hypothesized in partnership with health professionals and patients used to take medications daily. Unified Modeling Language was used to model the use cases. The evaluation was performed with usability testing and efficacy testing. The results show that the app was well accepted both in young people and older adults. Engaging target users and health professionals in the conception and development of a health-related app could have better results in the usability and the efficacy of the app.

Recently a new buzzword has slowly but surely emerged, namely the Internet of Things (IoT). The importance of IoT is identified worldwide both by organisations and governments and the scientific community with an incremental number of publications during the last few years. IoT in Health is one of the main pillars of this evolution, but limited research has been performed on future visions and trends. Thus, in this study we investigate the longitudinal trends of Internet of Things in Health through bibliometrics and use of text mining. Seven hundred seventy eight (778) articles were retrieved form The Web of Science database from 1998 to 2016. The publications are grouped into thirty (30) clusters based on abstract text analysis resulting into some eight (8) trends of IoT in Health. Research in this field is obviously obtaining a worldwide character with specific trends, which are worth delineating to be in favour of some areas.

mHealth and Telehealth technologies are increasingly used to provide personalised, interactive and timely access to health data, thereby helping patients take a more active role in their care process. However, similar to any intervention, the use of these technologies has to be assured to justify that they do not compromise patient safety. In this paper, we discuss the development of a safety case for MediPi; a research prototype for a low-cost open-source digital platform that collects physiological data from patients, at home, and makes it available to decision-support systems used by clinicians. We identify potential hazardous failures associated with the use of MediPi and examine current risk controls. We also explore the modular structure of the overall safety case of the platform. We conclude with a discussion of patient safety challenges related to the unsupervised nature of the care setting and the use of commercial off-the-shelf personal devices.

As a result of increasing demand in the face of reducing resources, technology has been implemented in many social and health care services to improve service efficiency. This paper outlines the experiences of deploying a ‘Software as a Service’ application in the UK social and health care sectors. The case studies demonstrate that every implementation is different, and unique to each organisation. Technology design and integration can be facilitated by ongoing engagement and collaboration with all stakeholders, flexible design, and attention to interoperability to suit services and their workflows.

Sequencing data will become widely available in clinical practice within the near future. Uptake of sequence data is currently being stimulated within the UK through the government-funded 100,000 genomes project (Genomics England), with many similar initiatives being planned and supported internationally. The analysis of the large volumes of data derived from sequencing programmes poses a major challenge for data analysis. In this paper we outline progress we have made in the development of predictors for estimating the pathogenic impact of single nucleotide variants, indels and haploinsufficiency in the human genome. The accuracy of these methods is enhanced through the development of disease-specific predictors, trained on appropriate data, and used within a specific disease context. We outline current research on the development of disease-specific predictors, specifically in the context of cancer research.

Clinical evidence demonstrates that BRCA 1 and BRCA2 mutations can develop a gynecological cancer but genetic testing has a high cost to the healthcare system. Besides, several studies in the literature indicate that performing these genetic tests to the population is not cost-efficient. Currently, our physicians do not have a system to provide them the support for prescribing genetic tests. A Decision Support System for prescribing these genetic tests in BRCA1 and BRCA2 and preventing gynecological cancer risks has been designed, developed and deployed in the Virgen del Rocío University Hospital (VRUH). The technological architecture integrates a set of open source tools like Mirth Connect, OpenClinica, OpenCDS, and tranSMART in addition to several interoperability standards. The system allows general practitioners and gynecologists to classify patients as low risk (they do not require a specific treatment) or high risk (they should be attended by the Genetic Council). On the other hand, by means of this system we are also able to standardize criteria among professionals to prescribe these genetic tests. Finally, this system will also contribute to improve the assistance for this kind of patients.

Exploratory Clustering is a novel general purpose clustering tool which is especially appropriate for medical domains in which we need to identify subpopulations that are similar in two different data layers. The tool implements the multi-layer clustering algorithm in a framework that enables iterative experiments by the user in his search for relevant patient subpopulations. A unique property of the tool is integration of clustering and feature selection algorithms. Differences in values of most relevant attributes are used to demonstrate decisive properties of constructed clusters. Usefulness of the tool is illustrated on a task of discovering groups of patients with similar cognitive impairment.

Treatment effectiveness plays a fundamental role in patient therapies. In most observational studies, researchers often design an analysis pipeline for a specific treatment based on the study cohort. To evaluate other treatments in the data set, much repeated and multifarious work including cohort construction, statistical analysis need to be done. In addition, as treatments are often with an intrinsic hierarchical relationship, many rational comparable treatment pairs can be derived as new treatment variables besides the original single treatment one from the original cohort data set. In this paper, we propose an automatic treatment effectiveness analysis approach to solve this problem. With our approach, clinicians can assess the effect of treatments not only more conveniently but also more thoroughly and comprehensively. We applied this method to a real world case of estimating the drug effectiveness on Chinese Acute Myocardial Infarction (CAMI) data set and some meaningful results are obtained for potential improvement of patient treatments.

Metabolomic data can potentially enable accurate, non-invasive and low-cost prediction of coronary artery disease. Regression-based analytical approaches however might fail to fully account for interactions between metabolites, rely on a priori selected input features and thus might suffer from poorer accuracy. Supervised machine learning methods can potentially be used in order to fully exploit the dimensionality and richness of the data. In this paper, we systematically implement and evaluate a set of supervised learning methods (L1 regression, random forest classifier) and compare them to traditional regression-based approaches for disease prediction using metabolomic data.