Skip to menu | Skip to content |

Centre for Health Informatics

Current PhD Projects

Restructuring Healthcare Patient Flow Process on a National Level.  A Patient Oriented Approach.

Student: Akis Papagiannis

Supervisors : Prof Abdul Roudsari and Prof Ewart Carson

Most of the current patients, especially in Greece, are highly concerned about three issues: First is to receive the safest treatment possible. Second is to receive it on time. Third is to pay the least amount of money possible if not covered by a health insurance program. The aim of this study is to formulate patient flow with these three mutually exclusive, both internal and external facts, into comprehensive standards defining and measuring patient satisfaction.

It is to the health care organization's best interest to satisfy its patients with the best quality service possible at a reasonable price.  Most probably health care organizations are spending the necessary budgets to ensure best service practices.  Assuming that these budgets are available the objectives of this study are to:

  1. Reconstruct the patient flow process from healthcare oriented to patient oriented for all three levels of healthcare.
  2. Restructure a health care organization's cross-functional patient flow process implemented through a single national repository system, which will provide fertile environment for patient oriented strategies.
  3. Measure the Patient flow reconstructed process with some indicative patient oriented measures as they will occur from the fundamental difference that exists between patient and non patient oriented processes.Implement an integrated systems design based on this project patient focus strategy.
  4. Clearly define the strategy's origin as well as its management information system parameters used for the process design.
  5. Demonstrate a basic flexible OLAP (On Line Analytical Processing) real time MIS (management Information System) of patient records to be used as a tool to audit the health care organization's quality management.
  6. Further elaborate on national or even international level the EPR (Electronic Patient Record) central spine factual information about the health care organization's ability to offer best treatment service based on this strategy.
  7. Improve efficiency in the health care system through competent management of institution resources and through strategic alliances among patients, health care providers and other stakeholders.
  8. Develop and maintain current field activity based driven standards that improve patient quality services.
  9. Establish gradually trends that will serve as thresholds for international accreditation for health care institutions to serve better the patients.

Telecare Applications in Saudi Arabia

Student: Khulud Al-Kadi

Supervisors: Prof Abdul Roudsari, Prof Ewart Carson

Khulud is researching the area of Telecare Applications in Saudi Arabia and factors contributing to the success or failure of such applications, with a concentration on Telecare to Manage Patients with Chronic Diseases, in particular Diabetics.

  

How can the improved AMIS affect the efficiency and effectiveness of decisions taken by senior level managers working at crisis conditions?

Student:Yousef Al-Mimi

Supervisors: Prof Ewart Carson, Ayesha Al-Rafai

This study will discuss the effect of improved automated management information systems (AMIS) on the efficiency and effectiveness of decisions taken by senior level managers working at hospitals in crisis. Medical and administrative managers of all hospitals working in West Bank will be involved in the study. The Focus group technique and In-depth interviews will be the tools employed for data collection.

Design and evaluation of a remote health monitoring kiosk

Student: Hamed Ezzatizadeh

Supervisor : Prof Abdul Roudsari, Prof Ewart Carson

It is widely accepted that healthcare providers are coming under increasing pressure to improve the quality of care delivered to patients. This comes at a time when there is a need to curtail healthcare spending fuelled by an increasing incidence of revenue sapping ageing populations, and an explosion of modern diseases such as obesity, diabetes, chronic heart failure etc. The healthcare delivery system will be forced to treat larger numbers of patient and resources will come under severe strain. This problem lessens the health care quality and its efficiency.

One of the possible methods for improvement of health care delivery and its quality is filtering unnecessary referrals. This can be achieved either by improving public health level (e.g. preventive medicine) or assigning a part of health checks to the members of the public (e.g. regular health checks). This Project is aiming to design, develop and implement a platform making it possible for people to play a greater role in maintaining their own health and letting them prevent unexpected health problems in a quick and easy way. In the same time this device could help to increase patients comfort level, reduce the number of check ups as in hospital and enhance the effectiveness of remedy by monitoring the healing progress.

A Decision Support System for Modelling Telecare

Student: Juan Adriano Moran

Supervisors: Prof Abdul Roudsari, Prof Jonathan Kay

The use of Information and Communication Technologies (ICTs) has been seen for the last decades as a method to improve and supply equal access to health care services. However, their integration has not been an easy task. To be successful in the operation of home based telemedicine systems (or other variant telemedicine applications), it is required that feasibility studies are to be carried out in advance. Anticipating the ways in which the medical practice will be affected in terms of their operation and resources are crucial elements that decision makers require to analyse before any new initiative takes place within the healthcare system.

Complex pilot projects can be implemented to capture such changes; however these projects may not be replicable and they imply extra costs and time. Furthermore, diverse policies and strategies can be difficult, impossible, or even dangerous to explore within the actual service running in reality. If telemedicine and telecare are not well planned, the consequences could affect negatively the performance of the whole service. For instance, they could interfere with other activities of health providers, increment workloads, and promote resistance to adoption, among others.

So far, we sustain that ill-defined projects, technology-driven approaches, and lack of understanding complex system are factors that lead to non-satisfactory telecare results. A possible solution to those problems would be to design a tool for capturing the behaviour of telecare within the healthcare system. With such tool, decision makers will be able to perform service evaluations before the actual system is implemented, and hopefully as result, better decision making would take place.

The general aim of this research is to designing a Decision Support System (DSS) tool that helps in the decision making process for planning telecare services under different healthcare scenarios. As a first step towards developing the DSS tool, a System Dynamics (SD) approach is proposed.

Wearable Computers In the Operating Room Environment

Student: Qi Ma

Supervisor: Dr Peter Weller, Prof Abdul Roudsari

The modern operating theatre is a busy environment with many people and any accidents that will happen at any time. To make out a best solution in a quick time, clinicians will need first hand information about patients as quick as possible. Wearable computer, worn much as glasses or clothing are worn, can interact with the user anytime anywhere and also according to the situation. Using wearable computer technology, a portable system can be provided to clinicians to not only allow access data quickly but also make visualized medical images projected same time as looking at patients.

Based on wearable computer technology, the ultimate goal is to provide a practical system platform, that can be convenient wearable, easy and comfort to interact with (input/output), dedicated in medical environment. Based on this platform, we can seamlessly integrate this node with existing hospital information system to allow a real-time information changing, and also implement other applications to ease and enhance clinicians' work, such as: X-Ray/MRI image displaying, 3-D visualized medical image projection, remote patient monitor, and so on.

The research work will involve researching, building and configuring a wearable computer system, investigating and implementing the input/output aspects of the system to make it easy to be use. Also, researching and providing clinical data in an efficient and interaction way will also be included. Finally, the work of evaluating the system, ultimately in a clinical environment, is also an important part of the whole job.

Considering this system will dedicated to be used in a modern operating theatre environment, we will need to solve many specified problems related to this special environment:

A Decision Support System for Clinical Guide Lines

Student: Ramon Garcia Rochin

Supervisor: Dr Peter Weller, Prof. Abdul Roudsari, J Mancilla

The general aim of this research is to design a Decision Support System (DSS) or Agent tool to aid in the creation of Clinical Guidelines (GLs) using Guide Line Interchange Format (GLIF) represented in Protege. GLs can be coupled into the Electronic Health Record (EHR) in a GLIF specification with a variety of guide line execution engines (GLEE). GLs are created by the healthcare organizations (HCOs) based on “evidence” obtained from patient outcomes. Agent technology can be used to evaluate the healthcare provider (HCP) and patient outcomes, comparing these to the “Gold Standard” and providing feedback to the GLs, HCPs, and patients.

Drug safety decision support system

Student: Sanaa Henni

Supervisor: Prof Abdul Roudsari, Prof Jonathan Kay

Medication errors are significant in hospitals, errors in drugs ordering, dosage, prescribing and administration. The majority of medication errors in health care are the prescribing errors; in the United States hospitals approximately 1-2% inpatients are harmed by medication error. Also 1000 patients complain in the United Kingdom in 1996 related to medication errors.

In the 1960’s errors were already a problem, but it was considered only as human error, then scientists started to find ways to detect it then to reduce it. First of all, we investigated the problem by understanding prospective and observational studies made in hospitals in order to identify causes and types of errors. This part of my project was based on writing a systematic review, collecting data on types and sources of errors, and understanding the concept of using computerized systems to decrease this problem. I have used system dynamics to describe the mismanagement of medication process, and linked it with human error, demonstrated by causal loop diagrams.

The aim of this project is to design decision support system to detect and minimise mistakes, it should include complete patient information, treatment history, diseases, and even should be linked to a drug database which is updated each time new drug is in the market or a drug is withdrawn. An alarm should be included to alert user of any mistake and notify the user of any warning such as drug interaction or patient personal information and medication history. But the Electronic Patient Record already exists, and each patient will have its own information electronically, subsequently a model should be made to detect errors, and reduce them.

Future work will include survey construction, extend research in patient needs and safety, design of drug safety decision support system with warnings, and formal evaluation of the model.

Intelligent Ventilation Management

Student: Aris Tzavaris

Supervisor: Dr Peter Weller, Prof Abdul Roudsari

Intensive Care Unit (ICU) medical personnel, in aiming to accommodate patient ventilation needs, utilize patient physiology and pathology data to define ventilator apparatus settings. Since patient needs are changing the procedure of adapting ventilation parameters, known as ventilation management, is ongoing. The above procedure of ventilation management could be described as a search of an optimum solution through a space of possible solutions. The aim of the research is to develop and evaluate a hybrid ventilation advisor system that could support ventilation process. It focuses on ICU patients who are ventilated in Control mode rather than patients assisted by the ventilator apparatus. The proposed system is designed to utilize real patient physiology data for “training” and ultimately developing an Artificial Intelligence (A.I.)Algorithm, based on a synergism of AI methodologies, including Fuzzy Logic [1], Evolution Algorithms [2] and Neural Networks [3].

A framework for evaluating the impact of medical digital libraries.

Student: Gemma Madle

Supervisor: Prof Abdul Roudsari, Dr Patti Kostkova

Recent years have seen an explosion in the amount of money being spent on IT projects in healthcare with a specially commissioned National Program for IT (NpfIT) being established for the UK National Health Service with £2.3billion allocated for the next three years.  Alongside this investment in infrastructure is an investment in the provision of knowledge with the creation and development of the National electronic Libraries for Health, Social Care and Public Health as part of the National Knowledge Service.  Medical digital libraries such as the National electronic Library for Health are essentially life critical applications enabling professionals to stay up to date.  They have the potential to change working culture creating new types of professional relationships and communities based across distances, hierarchy and other traditional barriers.

With this huge investment of public money it is surely important to investigate these changes and the impact of the resources provided on clinical care, decision-making and patient outcomes.  But how can we measure this impact? Previous research in the 1980s and early 1990s has shown that physical hospital library services can contribute to patient care and clinical decision-making.  However these studies were often driven by the need for the library to assert its value in times of economic constraint, therefore the focus tended to be on efficiency of services, satisfaction with the library and its impact on decision-making through self-reporting by clinicians', rather than measuring any actual outcomes e.g. knowledge and attitudes, length of hospital stay, prescribing rates, patient satisfaction.

 

This research aims to develop a framework for digital library impact evaluation, that will provide a true picture of the impact of a digital library on its user community. A review of the literature will identify dimensions of digital library work and produce a definition of a digital library to reflect these dimensions and a model to show interactions between the dimensions.  The review will also identify and review the current status of evaluation and impact research in digital libraries. Using the digital library definition and digital library work dimensions model a framework for digital library impact evaluation will be created to fit within the model.  This framework will be piloted on two medical digital libraries and evaluated.

Cardiac Arrhythmia Diagnosis in Implantable Devices, Development of New Algorithms using Informatics Techniques and Technologies.

Student: Julian Bostock

Supervisor: Dr Peter Weller, Prof Abdul Roudsari

Cardiac rhythm disturbances (arrhythmias) are among the commonest reasons for hospital admission. Many require rapid and accurate diagnosis to guide treatment. Very slow (bradyarrhythmias) and very rapid (tachyarrhythmias) arrhythmias are important as they may result in haemodynamic compromise and hypoxia and are considered life-threatening. Implantable cardiac pacemakers, defibrillators and monitoring devices require rapid accurate diagnosis to automatically deliver electrical therapies for arrhythmia and reduce inappropriate therapy.

 

Automatic arrhythmia diagnosis is a “classification” problem. Typical current algorithms use intracardiac electrograms as inputs to decision tree algorithms. Newer diagnostic algorithms offering improved accuracy over current technology may reduce inappropriate therapy. A preliminary study suggested that use of artificial intelligence (AI) might lead to significant further improvements in diagnostic accuracy.

Physiological data is collected from patients undergoing routine cardiac electrophysiological investigations to create a “library” of expert-validated arrhythmias for development, testing and validation of algorithms. Physiological parameters associated with cardiac arrhythmia are: heart rate, blood pressure; the surface ECG; intracardiac electrograms; skin temperature; body motion; thoracic bioimpedance; blood oxygen saturation and clinical history (predisposition to a condition) as well as additional derived parameters.

From the large volume of data available, feature selection is performed using principal components analysis (PCA),retaining its important characteristics, while reducing inputs to experimental algorithms.

Algorithms will be developed and tested in vitro, initially using MATLAB® “toolkits” suitable for approaching a classification problem (including linear discriminant analysis, signal processing, decision trees, Expert Systems, ANN, Fuzzy Logic, Genetic Algorithms, Wavelets, Chaotic Systems) as simple, hybrid or modular systems. Classification algorithms will be used separately, in series, in parallel, in modular or hybrid form and the most successful selected for further development.

Further validation is planned, with additional controlled clinical trials, comparing any new algorithm against current production algorithms, using parallel testing in a clinical environment.

 

The Use of Data and Computers in the Management of Diabetes

Student: Michael Hill

Supervisor:Prof Abdul Roudsari, Prof Ewart Carson

Academic research into the development of computer systems to support the management of diabetes has been taking place for over two decades now. The aims, scope and complexity of the systems developed have covered a wide range. A number of decision support systems providing insulin dose adjustment have been shown to have positive effects in clinical trials. However, it is not apparent that any of the systems available are widely used. There is little data available about the extent or methods of usage of computer systems by health care professionals and patients in the real world. The purpose of this study is to collect some quantitative data about computer usage in the task of diabetes management. In the process, some insight into the barriers preventing more widespread usage may be gained.

The Application of Health Informatics and Visualisation to Rehabilitation Goal Setting

Student: Gary Derwent

Supervisor: Prof Abdul Roudsari, Wendy Magee

Goal-setting is a recognised technique in neurological rehabilitation and is also used in many other fields such as mental health, learning disabilities and care of the elderly. Goal setting has been shown to contribute to outcome measurement, patient motivation, multi-disciplinary team organisation and communication. However there are a number of problems with current goal-setting practice. A wide variety of goal-setting methods are used across rehabilitation units and there are few methods of analysing and utilising this information to compare between services and few tools to help realise the potential of this information to organise multi-disciplinary team working. The aim of the present study is to analyse goal-setting information captured by different rehabilitation units and ascertain whether a common data-structure can accommodate the information needs of a variety of goal-setting methods. The study will also examine classification methods and possible visual tools that could make the analysis of goal-setting information more useful to rehabilitation clinicians and other stakeholders.

Computerized Clinical Decision Support System( CDDS) for Management of HIV and AIDS in South Africa

Student: Letsholo Mojanaga

Supervisor: Prof Abdul Roudsari

The complexity of treating patients infected with HIV requires consideration of multiple factors like appropriate combination of drugs from different classes, avoidance of combinations of medications known to cause adverse reactions, make adjustments when patients become resistant to specific drugs and monitor adverse reactions to individual drugs. In addition, routine HIV care is becoming more complicated as many new pieces of clinical data have to be considered on an ongoing basis, and the knowledge of HIV care continues to change much more rapidly than in most other medical disciplines.

The proposed study aims to develop a clinical decision support system(cdds) that will improve HIV and AIDS treatment by providing automated alerts, consistency checks, increase treatment efficacy, provide best practice guidelines, and increase knowledge by providing information when and where it is needed.

CareNet: A web based solution augmented by web 2.0 for diabetes education of adolescent patients

Student: Omid L. Shabestari

Supervisor: Prof Abdul Roudsari

Diabetes is a common chronic disease involving 3.66% of UK population with incidence rate of 0.33%. Because of multi-organ involvement of diabetes, this disease costs more than £5 million per day for NHS.

In the UK, this disease has got two important patterns in recent years. First is the increase in incidence and the second is the shift to the youth in type II. Diabetic patients become more complicated in adolescence because of the change in hormonal balance due to puberty and the service transfer from paediatric diabetes services to adult services.

Education has proved to have a short-time effect on diabetes care, but there is a big need to extend and renew the diabetes education to cover more patients (11% structured education in 2007 based on Healthcare Watchdog report). Considering the 61% coverage of Internet in UK and the computer-native nature of the young generation, E-learning via web can be a good solution for this purpose.

This project is based on the application of web 2.0 for diabetes education of the adolescent patients.