Simultaneous trend analysis for evaluating outcomes in patient-centred health monitoring services

Abstract

The research aim underpinning the Healthcare@Home (HH) information system described here was to enable ‘near real time’ risk analysis for disease early detection and prevention. To this end, we are implementing a family of prototype web services to ‘push’ or ‘pull’ individual’s health-related data via an system of clinical hubs, mobile communication devices and/or dedicated home-based network computers. We are examining more efficient methods for ethical use of such data in timeline-based (i.e. ‘longitudinal’) data analysis systems. A consistent data collation infrastructure is being created for use along the ‘patient path’—accessible wherever patients happen to be. This ‘patient-centred’ infrastructure can be applied in the evaluation of disease progression risk (in the light of clinical understanding of disease processes). In this paper we describe the requirements for making multi-data trend management ‘scale-up’, together with some requirements of an ‘end-to-end’ functioning data collection system. A Service-Oriented Architecture (SOA) approach is used to maximise benefits from (1) clinical evidence and (2) computational models of disease progression that can be made available elsewhere on the SOA. We discuss the implications of this so-called ‘closed loop’ approach for improving healthcare intervention outcomes, patient safety, decision support, objective measurement of service quality and in providing inputs for quantitative healthcare (predictive) modelling.

Glossary of (Jargon) terms in the context of the project for non-specialists

Aggregate analytical framework :

A data collection system that has consistent data formats and collection policies to permit combination of data to suit an analysis task.

Algorithmic modelling of outcomes :

A method for discovering the complex dependencies of measured variables to a specified healthcare outcome.

Closed loop outcomes analysis :

a key design point of Healthcare@Home, where a high quality risk analysis step takes place prior to and following a specific intervention so to better understand effectiveness.

Composite Process Functions :

where information services are separated into small chunks which work together to support a business process.

Domain :

in the project’s context, the real-world, day-to-day activities and concerns of ‘healthcare’ (as contrasted with technical computing concerns).

Domain Knowledge Expert :

someone who has deep understanding of knowledge and its structure associated with the domain.

Domain-Driven Business Process Modelling :

A methodology that captures understanding of domain knowledge experts and other roles associated with the domain. The modelling is useful for the construction of information services that serve the needs of the domain.

Grid Technologies :

in general, the use of resources of many separate computers (working in parallel connected by a network that is usually the Internet) to solve large-scale computational problems.

HH :

Abbreviation for Healthcare@Home, the name of the research phase demonstrator project being described in this paper—see http://www.ehealthnews.eu/content/view/837/27/

Integrated Care Pathway :

in general, a multidisciplinary outline of anticipated care, placed in an appropriate timeframe, to help a patient with a specific condition or set of symptoms move progressively through a clinical experience to positive outcomes.

Inter-Enterprise Computing :

creation and use of a wide range of environments, equipment and infrastructure in a way that allows secure, appropriate sharing of knowledge and powerful computing resources. The sharing can occur over a wide area using a high bandwidth communications infrastructure (sometimes but not exclusively referred to as grids).

Interoperable :

in general the ability of systems to exchange and use information, the term is often applied to products and systems from multiple vendors that can be used together without modification or development of custom interfaces and tools.

Logical Domain Model :

in the context used here, a model created by domain stakeholders that organises functional requirements as a framework for technical implementation choices.

OGSA-DAI :

Open Grid Services Architecture—Data Access and Integration—A project conceived by the UK Database Task Force with the aim to develop middleware assisting with access and integration of data from separate sources via the grid—see http://www.ogsadai.org.uk

Portal :

In the context used here, a Web site that provides a single point of access to specific applications and information (there may be many portals created for different purposes).

Risk Monitoring :

In the Healthcare@Home context, continuous or discontinuous analysis of measured variables (e.g. blood glucose concentration) that may indicate specified outcome risk as part of a legitimate healthcare ‘trend’ service. Monitoring can be extended anywhere a telecommunications network can support the devices and applications model.

Scalability :

in general, the ability to expand a computing solution to support large numbers of users (i.e. increasing demand) without impacting performance. Scaleable services have potential for reducing cost, though this is untested.

Service-Oriented Architecture (SOA) :

put simply, a concept for a software infrastructure that defines how collections of distributed information services can interact to communicate and interoperate via agreed standards. Collections of services are so-called ‘loosely-coupled’ to promote re-use of components and when combined, support complex processes.

Time Series Analysis :

an approach to identifying underlying behaviour from a timeline-based sequence of observations, e.g. sensor-derived biomedical data. TSA may forecast (predict) future values of the time series variable, but this requires that patterns of observed time series data are identified and formally described.

Web Services :

in the context used here, WS represent simple, self contained applications which perform specific functions. Web Services describe a standardized way of integrating Web-based applications using the XML, SOAP, WSDL and UDDI open standards over an Internet protocol backbone.