Ebook: Driving Reform: Digital Health is Everyone's Business

Healthcare planners and governments across the globe are involved in efforts to stimulate innovation and to reform health systems as a means of enhancing the delivery of safe, effective and quality care to patients. Information and communication technology (ICT) is a critical component of health system innovation and a key stimulus for reform. This is because the introduction of ICT is much more than just a technical taskit involves new ways of working, networking and organising globally. The Health Informatics Society of Australia (HISA) with the support of the Australasian College of Health Informatics (ACHI) have continued to be in the forefront of efforts to promote ICT-enabled healthcare reform and innovation in Australia. The Australian National Health Informatics Conference (HIC) is Australia's premier health informatics event bringing together researchers, industry groups and healthcare providers from Australian and internationally, providing the opportunity to display and share cutting edge research evidence, technology updates and innovations.

The theme for HIC 2015 is “Driving reform: Digital health is everyone's business”. The conference highlights the role that health informatics can play in the development of a smart digital healthcare system to drive reform and to meet increasing demands and financial pressures. In today's environment health services are being increasingly technology-enabled so as to:

• Address increasing consumer demands and expectations

• Provide relevant, appropriate and effective real-time information to clinical practitioners

• Enhance the management and monitoring of healthcare delivery performance

The papers in this volume provide valuable research evidence and information about the diverse role that ICT plays in the health, aged and community care sectors, across Australia, New Zealand and internationally. The papers represents a wide spectrum of work encompassing major theoretical concepts, examples of key applications of new technologies and important new developments in the field of health informatics. Taken together the volume makes an important contribution to the health informatics evidence base and the quality development of ICT.

This year's Health Informatics Conference continues and augments the double blind peer review process established in 2011 in a previous volume. All papers were reviewed by three experts in the field of health informatics, selected as prominent academics and industry specialists. The assistance of the Australasian College of Health Informatics in supporting this process through the voluntary efforts of its Fellows is gratefully acknowledged, as is the contribution made by many senior and experienced members of the Health Informatics Society of Australia. This phase of reviewing resulted in the provisional acceptance of 26 from a much expanded submission field of 51 papers. The Scientific Program Committee then undertook a validation process for all such papers that were resubmitted in amended form, to ensure that reviewers' recommendations were appropriately addressed or rebutted. In total 26 papers were included in this volume. Congratulations to all the authors of the papers in this volume.

Andrew Georgiou

Heather Grain

Louise K. Schaper

LabSurv is an electronic notification system developed to support laboratories to directly notify the results of notifiable disease testing to public health services in New Zealand. A direct laboratory notification middleware framework was developed to manage the information flow between laboratories and public health services. The framework uses an HL7 messaging standard to receive the laboratory results and windows services to integrate the results with the cases of notifiable diseases within a national electronic surveillance system. This paper presents the system design and implementation details of direct laboratory notification system in LabSurv. It presents the HL7 messages structure implemented in the system. Finally, the performance of the system based on implemented framework is analysed and presented to evaluate the efficiency of our design.

Clinical decisions rely on expert knowledge that draws on quality patient phenotypic and physiological data. In this regard, systems that can support patient-centric care are essential. Patient registries are a key component of patient-centre care and can come in many forms such as disease-specific, recruitment, clinical, contact, post market and surveillance. There are, however, a number of significant challenges to overcome in order to maximise the utility of these information management systems to facilitate improved patient-centred care. Registries need to be harmonised regionally, nationally and internationally. However, the majority are implemented as standalone systems without consideration for data standards or system interoperability. Hence the task of harmonisation can become daunting. Fortunately, there are strategies to address this. In this paper, a disease registry framework is outlined that enables efficient deployment of national and international registries that can be modified dynamically as registry requirements evolve. This framework provides a basis for the development and implementation of data standards and enables patients to seamlessly belong to multiple registries. Other significant advances include the ability for registry curators to create and manage registries themselves without the need to contract software developers, and the concept of a registry description language for ease of registry template sharing.

The use of online technologies for supporting participants of behaviour change and diet program is a timely and important research direction. We present HealthierU, adaptive online portal offering a suite of interactive support tools. The portal was evaluated in a 24-week study, which shows that regular reminders trigger increased interaction with the portal. We also analyse interaction patters conducive to weight loss and discuss possible factors of the attrition rates observed in the study.

Computer-based decision support information systems have been promoted for their potential to improve physician performance and patient outcomes and support clinical decision making. The current case study reported design and implementation of a high-level decision support system (DSS) which facilitated the flow of data from operational level to top managers and leadership level of hospitals. The results shows that development of a DSS improve data connectivity, timing, and responsiveness issues via centralised sourcing and storing of principal health-related information in the hospital. The implementation of the system has resulted in significant enhancements in outpatient waiting times management.

People having access to their medical records could have a transformative improvement effect on healthcare delivery and use. Our research aimed to explore the concerns and attitudes of giving people electronic access to their medical records through patient portals. We conducted 28 semi-structured interviews with 30 people, asking questions about portal design, organisational implications and governance. We report the findings of the governance considerations raised during the interviews. These revealed that (1) there is uncertainty about the possible design and extent of giving people access to their medical records to view/use, (2) existing policies about patient authentication, proxy, and privacy require modification, and (3) existing governance structures and functions require further examination and adjustment. Future research should include more input from patients and health informaticians.

Interpreters are required to aid communication between clinicians and culturally and linguistically diverse (CALD) patients to ensure appropriate and timely care. Demand for interpreting services however, often exceeds supply. A mobile app to translate clinical assessment questions in 10 common languages using pictorial, written and voice-over prompts to assist patient assessments when interpreters are unavailable has been developed. This paper reports on the User Needs Analysis that informed the app. The analysis consisted of focus groups with allied health clinicians to understand pertinent aspects of initial allied health assessments and the communication needs to be addressed in the design of an app-based patient assessment tool. Outcomes show that of primary importance to clinicians was the ability to not only ask the patients questions, but to communicate information to increase understanding of, and ensure compliance with, treatments and interventions to promote patient function and comfort.

This paper presents the StepKinnection game, a Kinect-driven stepping game for the elderly that delivers stepping exercises to train specific cognitive and physical abilities associated with falls. This system combines a set of suitable age-related features, meaningful exercise routines and an embedded clinical test for fall risk assessment. The combination of these three aspects makes the game potentially useful in practice as the game is appealing to the elderly cohort, trains one of the most important abilities to prevent falls and at the same time allows for a continuous assessment of health outcomes; characteristics not available in the literature nor in current commercial games.

Background: The effective care and well-being of a community is a challenging task especially in an emergency situation. Traditional technology-based silos between health and emergency services are challenged by the changing needs of the community that could benefit from integrated health and safety services. Low-cost smart-home automation solutions, wearable devices and Cloud technology make it feasible for communities to interact with each other, and with health and emergency services in a timely manner.

Objectives: This paper proposes a new community-based care model, supported by technology, that aims at reducing healthcare and emergency services costs while allowing community to become resilient in response to health and emergency situations.

Methods: We looked at models of care in different industries and identified the type of technology that can support the suggested new model of care. Two prototypes were developed to validate the adequacy of the technology.

Results: The result is a new community-based model of care called ‘Le Bon Samaritain’. It relies on a network of people called ‘Bons Samaritains’ willing to help and deal with the basic care and safety aspects of their community. Their role is to make sure that people in their community receive and understand the messages from emergency and health services. The new care model is integrated with existing emergency warning, community and health services.

Conclusion: Le Bon Samaritain model is scalable, community-based and can help people feel safer, less isolated and more integrated in their community. It could be the key to reduce healthcare cost, increase resilience and drive the change for a more integrated emergency and care system.

Dynamic and automatic patient specific prediction of the risk associated with ICU mortality may facilitate timely and appropriate intervention of health professionals in hospitals. In this work, patient information and time series measurements of vital signs and laboratory results from the first 48 hours of ICU stays of 4000 adult patients from a publicly available dataset are used to design and validate a mortality prediction system. An ensemble of decision trees are used to simultaneously predict and associate a risk score against each patient in a k-fold validation framework. Risk assessment prediction accuracy of 87% is achieved with our model and the results show significant improvement over a baseline algorithm of SAPS-I that is commonly used for mortality prediction in ICU. The performance of our model is further compared to other state-of-the-art algorithms evaluated on the same dataset.

Home monitoring of chronically ill or elderly patient can reduce frequent hospitalisations and hence provide improved quality of care at a reduced cost to the community, therefore reducing the burden on the healthcare system. Activity recognition of such patients is of high importance in such a design. In this work, a system for automatic human physical activity recognition from smart-phone inertial sensors data is proposed. An ensemble of decision trees framework is adopted to train and predict the multi-class human activity system. A comparison of our proposed method with a multi-class traditional support vector machine shows significant improvement in activity recognition accuracies.

The project reported in this paper models a new approach to making health informatics and e-health education widely available to students in a range of Australian clinical health profession degrees. The development of a Masters level subject uses design-based research to apply educational quality assurance practices which are consistent with university qualification frameworks, and with clinical health profession education standards; at the same time it gives recognition to health informatics as a specialised profession in its own right. The paper presents details of (a) design with reference to the Australian Qualifications Framework and CHIA competencies, (b) peer review within a three-university teaching team, (c) external review by experts from the professions, (d) cross-institutional interprofessional online learning, (e) methods for evaluating student learning experiences and outcomes, and (f) mechanisms for making the curriculum openly available to interested parties. The project has sought and found demand among clinical health professionals for formal health informatics and e-health education that is designed for them. It has helped the educators and organisations involved to understand the need for nuanced and complementary health informatics educational offerings in Australian universities. These insights may aid in further efforts to address substantive and systemic challenges that clinical informatics faces in Australia.

Introduction: Skin and subcutaneous tissue infections (SSTI) are common conditions that cause avoidable hospitalisation in New Zealand. As part of a program to improve the management of SSTI in primary care, electronic medical records (EMR) of four Auckland general practices were analysed to identify SSTI occurrences in the last three years.

Methods: An ontology for SSTI risks, manifestation and treatment was created based on literature and guidelines. An SSTI identification algorithm was developed examining EMR data for skin swab tests, diagnoses (READ codes) and textual clinical notes.

Results: High occurrence and recurrence rates in those aged 20 or younger were found. Due to low usage of READ coding and laboratory tests, 65% of SSTI occurrences were identified by notes. However, 91% of all identified SSTI occurrences were appropriately treated with oral/topical antibiotics according to prescription records in the EMR. The F₁ score of the analysis algorithm is 0.76 using manual review as gold standard.

Discussion and Conclusion: The SSTI identification algorithm shows a reasonable accuracy suggesting the feasibility of automatic detecting SSTI occurrences using clinical data that are routinely collected in healthcare delivery.

Introduction: Diabetes is a common disease affecting 9% of the adult population worldwide. People with impaired glucose tolerance (‘prediabetes’) are at high risk of progressing to type 2 diabetes.

Methods: To understand prediabetes incidence rate, we analysed the electronic medical records (EMR) from 14 New Zealand general practices regarding patients aged ≥20 years and enrolled with the practices between 2009 and 2012. Prediabetes incidence rate was calculated by the number of patients with an initial HbA1c of 41–49 mmol/mol in 2011 among those who had not been diagnosed or treated for diabetes.

Results: 28,192 adults were included in the analysis, 11% of this cohort had diabetes before 2011. 1,276 new cases of prediabetes were identified in 2011, giving a 5.0% incidence rate. The relative risk (RR) for prediabetes was increased for the Māori and Pacific groups versus non-Māori/non-Pacific people, with RR of 1.97 in the younger age groups (<50 years) and RR of 1.42 in the 50+ group. The RR for having uncontrolled HbA1c (highest HbA1c in 2011 ≥65 mmol/mol) among the whole adult population was also increased for the Māori and Pacific groups versus non-Māori/non-Pacific people (RR=3.35 among those <50 years, RR=4.35 in the 50+ group).

Discussion and Conclusion: EMR analysis identified an alarming incidence rate of prediabetes, especially among Māori and Pacific groups, highlighting the need to better prevent and manage the condition.

We consider the task of automatic classification of clinical incident reports using machine learning methods. Our data consists of 5448 clinical incident reports collected from the Incident Information Management System used by 7 hospitals in the state of New South Wales in Australia. We evaluate the performance of four classification algorithms: decision tree, naïve Bayes, multinomial naïve Bayes and support vector machine. We initially consider 13 classes (incident types) that were then reduced to 12, and show that it is possible to build accurate classifiers. The most accurate classifier was the multinomial naïve Bayes achieving accuracy of 80.44% and AUC of 0.91. We also investigate the effect of class labelling by an ordinary clinician and an expert, and show that when the data is labelled by an expert the classification performance of all classifiers improves. We found that again the best classifier was multinomial naïve Bayes achieving accuracy of 81.32% and AUC of 0.97. Our results show that some classes in the Incident Information Management System such as Primary Care are not distinct and their removal can improve performance; some other classes such as Aggression Victim are easier to classify than others such as Behavior and Human Performance. In summary, we show that the classification performance can be improved by expert class labelling of the training data, removing classes that are not well defined and selecting appropriate machine learning classifiers.

Background: Hospital administrative data commonly consist of hundreds of variables with many consisting of hundreds, if not thousands, of distinct categories, especially for disease groups. Conventional approaches to develop regression models for prediction either fail completely due to multicollinearity or sparsity issues or take too long and consume too many computer resources.

Methods: We demonstrate how regularisation and variable aggregation techniques such as Elastic Net can overcome some of these problems. Parameter estimates from univariate generalised linear models (GLM) and Elastic Net models were used to aggregate disease groups into a more manageable number and predict the probability of mortality for a given patient.

Results: When employed for variable aggregation and variable selection, Elastic Net models ran at least four times faster than GLMs, though producing a less discriminative model. When applied to final models for predicting hospital mortality, though, both Elastic Net and GLM models demonstrated similar predictive power and efficiently solved an otherwise complex problem.

Conclusion: Elastic Net regularisation and variable aggregation provide an efficient mechanism for solving healthcare modelling problems.

To date, there is little information in the literature to guide the provision of supports for using the Personally Controlled Electronic Health Record (PCEHR) in populations with severe communication impairments associated with a range of disabilities. In this paper we will (a) outline the rationale for use of PCEHR in these populations by providing an overview of relevant research to date, and (b) present results of three integrated pilot studies aiming to investigate the barriers to and facilitators for PCEHR use by people with severe communication impairments and their service providers. Finally, we will present directions for future research on use of PCEHR by people with severe communication impairments.

Health data has long been scrutinised in relation to data quality and integrity problems. Currently, no internationally accepted or “gold standard” method exists measuring data quality and error rates within datasets. We conducted a source data verification (SDV) audit on a prospective clinical trial dataset. An audit plan was applied to conduct 100% manual verification checks on a 10% random sample of participant files. A quality assurance rule was developed, whereby if >5% of data variables were incorrect a second 10% random sample would be extracted from the trial data set. Error was coded: correct, incorrect (valid or invalid), not recorded or not entered. Audit-1 had a total error of 33% and audit-2 36%. The physiological section was the only audit section to have <5% error. Data not recorded to case report forms had the greatest impact on error calculations. A significant association (p=0.00) was found between audit-1 and audit-2 and whether or not data was deemed correct or incorrect. Our study developed a straightforward method to perform a SDV audit. An audit rule was identified and error coding was implemented. Findings demonstrate that monitoring data quality by a SDV audit can identify data quality and integrity issues within clinical research settings allowing quality improvement to be made. The authors suggest this approach be implemented for future research.

In a health system increasingly driven by cost constraints, there is a focus on improved electronic transfer of information to support healthcare delivery. One area of healthcare that has moved more quickly than others to achieve this is prescribing in the primary care environment. Whilst the move to electronic transfer of prescriptions has reduced transcription errors, the regulatory environment persists with handwritten signatures. This constraint, whilst addressed slowly with technology solutions, needs support from legislative change. The ultimate step is to have a secure mobile model, which would support the move to a fully-electronic, paperless transaction model.

Over 380 million adults worldwide are currently living with diabetes and the number has been projected to reach 590 million by 2035. Uncontrolled diabetes often lead to complications, disability, and early death. In the management of diabetes, dietary intervention to control carbohydrate intake is essential to help manage daily blood glucose level within a recommended range. The intervention traditionally relies on a self-report to estimate carbohydrate intake through a paper based diary. The traditional approach is known to be inaccurate, inconvenient, and resource intensive. Additionally, patients often require a long term of learning or training to achieve a certain level of accuracy and reliability. To address these issues, we propose a design of a smartphone application that automatically estimates carbohydrate intake from food images. The application uses imaging processing techniques to classify food type, estimate food volume, and accordingly calculate the amount of carbohydrates. To examine the proof of concept, a small fruit database was created to train a classification algorithm implemented in the application. Consequently, a set of fruit photos (n=6) from a real smartphone were applied to evaluate the accuracy of the carbohydrate estimation. This study demonstrates the potential to use smartphones to improve dietary intervention, although further studies are needed to improve the accuracy, and extend the capability of the smartphone application to analyse broader food contents.

People with Parkinson's disease are known to have difficulties in language and communication. This paper proposes the use of an artificial conversational agent, commonly known as a chat-bot that runs on a smart-phone device and performs two-way conversation with the user. In this paper, initial work on a Parkinson's disease themed chat-bot that interacts with the user relative to their symptoms is presented. Potential dialogues are provided to illustrate the various roles chat-bots can play in the management of Parkinson's disease. The chat-bot can be used for measuring voice and communication outcomes during the daily life of the user, and for gaining information about challenges encountered. Moreover, it is anticipated that it may also have an educational and support role. The chat-bot is now ready for usability testing with a clinical population.

This study aims to identify the benefits of using electronic health records (EHR) for client safety in residential aged care (RAC) homes. The aged care accreditation reports published between 27 April 2011 and 3 December 2013 were downloaded and analysed. It could be seen from these reports that only 1,031(37.45%) RAC homes in Australia had adopted an EHR system by 2013. 13 RAC homes failed one or more accreditation standards. Only one of these was using an EHR system and this one met the accreditation standards on information systems. Our study provides empirical evidence to suggest that adopting and using EHR can be one of the effective organisational mechanisms to meeting accreditation standards in RAC homes.

Objective: To inform about a very unique and first of its kind telehealth pilot study in India that has provided virtual telehealth consultation to eye care patients in low resource at remote villages.

Background: Provision of Access to eye care services in remote population is always challenging due to pragmatic reasons. Advances in Telehealth technologies have provided an opportunity to improve access to remote population. However, current Telehealth technologies are limited to face-to-face video consultation only. We inform about a pilot study that illustrates real-time imaging access to ophthalmologists. Our innovative software led technology solution allowed screening of patients with varying ocular conditions.

Methods: Eye camps were conducted in 2 districts in South India over a 12-month period in 2014. Total of 196 eye camps were conducted. Total of 19,634 patients attended the eye camps. Innovative software was used to conduct consultation with the ophthalmologist located in the city hospital. The software enabled virtual visit and allowed instant sharing of fundus camera images for assessment and diagnosis.

Results: About 71% of the patients were found to have Refractive Error problems, 15% of them were found to have cataract, 7% of the patients were diagnosed to have Retina problems and 7% of the patients were found to have other ocular diseases. The patients requiring cataract surgery were immediately transferred to city hospital for treatment. Software led assessment of fundus camera images assisted in identifying retinal eye diseases.

Conclusion: Our real-time virtual visit software assisted in specialist care provision and illustrated a novel tele health solution for low resource population.

Non-contact measurements of cardiac pulse can provide robust measurement of heart rate (HR) without the annoyance of attaching electrodes to the body. In this paper we explore a novel and reliable method to carry out video-based HR estimation and propose various performance improvement over existing approaches. The investigated method uses Independent Component Analysis (ICA) to detect the underlying HR signal from a mixed source signal present in the RGB channels of the image. The original ICA algorithm was implemented and several modifications were explored in order to determine which one could be optimal for accurate HR estimation. Using statistical analysis, we compared the cardiac pulse rate estimation from the different methods under comparison on the extracted videos to a commercially available oximeter. We found that some of these methods are quite effective and efficient in terms of improving accuracy and latency of the system. We have made the code of our algorithms openly available to the scientific community so that other researchers can explore how to integrate video-based HR monitoring in novel health technology applications. We conclude by noting that recent advances in video-based HR monitoring permit computers to be aware of a user's psychophysiological status in real time.

There is a substantial variation in healthcare spending and readmission rate for individuals having admissions to different hospitals. This study assessed how the community structure of physician collaboration networks that evolve during the period of providing healthcare services to hospitalised patients contribute to this variation. A physician collaboration network is said to have a community structure if the nodes (i.e. physicians) of that network can be easily grouped into sets of nodes such that each set of nodes is densely connected internally but sparsely connected between groups. This study constructed physician collaboration networks based on patient-sharing ties among physicians who provided healthcare services to hospitalised patients. An administrative health insurance claim dataset was utilised to extract patient-sharing ties among physicians. Simple linear regression models were estimated to assess the impact of the community structure of physician collaboration networks on the healthcare outcome measures (i.e. readmission rate and hospitalisation cost). From these models, this study found that the structure of a physician community has significant impact on readmission rate and hospitalisation cost. Healthcare administrators or managers could consider this finding in developing effective and efficient healthcare environments in their respective healthcare organisations.