Identifying admission patterns and targeting interventions in COPD

north west icon North West Coast
Researchers involved: Led by Professor Mike Pearson (Clinical Lead for COPD)
Disease area impacted: Unplanned admissions in COPD
Key partners: Prof Peter Diggle, Distinguished University Professor of Statistics in the Faculty of Health and Medicine, Lancaster University
Start and end dates: Jan 2016 to Dec 2018

Project Overview

Chronic Obstructive Pulmonary Disease (COPD) affects around 3 million people in the UK and is the second most common cause of emergency medical admission.  COPD is a long-term condition that can progressively worsen and can have a significant impact on the quality of life for those living with it. Flare-ups, often called exacerbations, occur frequently in people with COPD and often result in emergency hospital admissions.  The key to minimising emergency admissions is understanding the factors that increase the risk of acute exacerbations and ensuring that those patients most at risk are well supported by the health professionals and services involved in the provision of their care.

Better coordinated health and social care and better use of health and social care data, has the potential to improve patient outcomes and efficiency, reduce emergency department attendances, hospital admissions, and duration of stay.

Reducing data latency

The University of Liverpool has established a Healthcare Data Laboratory to undertake clinically driven analysis of emergency admissions for COPD across the whole of the North West Coast. We have engaged the clinical communities involved in the provision of care for people with COPD in the production of metrics to make care safer and more effective.  The data will be refreshed on a quarterly basis and shared with clinical colleagues in cycle of report generation, feedback and refinement to allow healthcare teams to make the most of the data available to them.

Improved planning and more integrated provision of services for COPD

The overlaying of aggregated health data relating to patient journeys at small area level (starting with metrics of emergency admissions in COPD, follow-up, and readmission) with a range of extra information on community assets such as GP practices, community clinics, and voluntary sector organisations will provide insight into the level of out of hospital support available to patients with COPD in defined geographical areas. We will flag GP practices according to the levels of prescribing of relevant medications for COPD and identify opportunities for supporting community services.  Geo-mapping techniques and mapping of local service provision will enable us to identify those areas with high disease burden or poor outcomes with a mismatch of service provision. This will enable more effective service planning and targeting of NHS resources.

The activities for our CHC Healthcare Data Laboratory and Pathway Profiling Team undertaking this work will include:

  • Developing clinically validated algorithms that accurately capture all interactions that COPD patients have with provider Trusts across the patch
  • Engaging clinical communities in the development of clinically relevant metrics for COPD
  • Reducing data latency (delays) for frontline teams – providing up to date information that can inform service design and improve point of care decision making
  • Applying smart analytics to routinely collected anonymised administrative data to segment COPD patients according to higher or lower risk of unplanned admission. Identify health, social, and demographic factors associated with higher risk for unplanned admission and implement appropriate interventions to minimise their impact.
  • Mapping hot-spots in emergency admissions for COPD patients
  • Mapping services to highlight any mismatch in service provision and patient demand

Why is the research important?

People with long terms conditions such as COPD, have overlapping health and social needs and while it might seem obvious that care should be better coordinated across health and social care, it has been difficult to achieve.     

There are huge amounts of data collected and we believe that we could use technology to bring data together in such a way it could then be used by care services, for the benefit of the patient.  This should improve patient outcomes, and if people are supported to manage conditions better themselves, there would be a reduced need for emergency department visits and fewer hospital admissions.    

What data is being used in this project?

We currently have access to anonymised commissioning datasets, including all A & E attendances, inpatient hospital admissions, and outpatient attendances for the whole of the Northwest Coast. In the early phases we will be linking this data to open source aggregated data such as indexes of deprivation and air pollution data.

Is the data anonymised?

The data we are currently working with is anonymised patient data which means a patient can’t be identified from it.  If use of identifiable data is needed in later phases of the project, we will seek consent from the patients to whom the data relates.

What methods are you using to conduct this work?

There will be two strands to the approach we take in conducting this work and a combination of research methods will be employed.  One strand of the work will be undertaken by the Healthcare Data Laboratory established within the University of Liverpool’s Department of Biostatistics. This team will be analysing anonymised commissioning datasets for the whole of the North West Coast; developing algorithms that accurately capture all interactions that patients have with provider Trusts across the patch; and seeking to understand patterns in admission frequencies and any significant trends and correlations within the data.

By linking our analytical expertise with a programme of engagement with front-line NHS clinicians across the North West Coast region, the team will support the co-production of clinically validated analytics and data-visualisation for defined patient pathways with NHS staff and stakeholders.  This will drive the development of a Learning Health System linking information systems to innovative health informatics research.

In parallel to the work undertaken by analysts within the Healthcare Data Laboratory, a Pathway Profiling team comprising two researchers will work with healthcare teams within provider organisations to build a detailed profile of the people, data, and systems involved in the care of patients across the various pathways. This team will map out the patient data trails that are created within and across health and social care providers and assess points along the patient pathway at which better systems or processes may facilitate better data capture and accessibility.

Who will benefit from your research?

The ultimate aim of the project is to optimise patient pathways and to improve outcomes for patients who have COPD. We will do this by identifying data-driven solutions to the problems we find in patient pathways and equipping the professionals involved in the provision of care to these patients with the tools and knowledge needed to make better decisions about patient care.