A Data Warehouse for Diabetes, Endocrinology and Obesity

Julian Collins, Informatics Manager for Diabetes at Kings Health Partners, blogs about the creation of a 'Data Warehouse' for diabetes to build a picture of diabetes and its treatment in Lambeth and Southwark.

King’s Health Partners has developed an integrated data store covering all aspects of diabetes care for people using our services. Data is drawn from a wide range of service providers, with identifiable information removed at source, and then added to a ‘Data Warehouse’ to build a coherent picture of diabetes and its treatment across our partner organisations.

Figure 1: Number of patient data records in the Data Warehouse by main source. Where DECS: South East London Diabetic Eye Complication Screening Programme, GSTT: Guy’s and ST Thomas’ NHS Foundation Trust, KCH: King’s College Hospital NHS Foundation Trust, GP: General Practitioner, Community: Lambeth Diabetes Intermediate Care Team and Southwark Community Clinics

Figure 1 gives an idea of the sources that supply data to the Data Warehouse, and the numbers of patients involved. More than 10 years of data has been added, enabling studies to track treatments and outcomes over time. Patient numbers do overlap significantly across the sources, reflecting the connected nature of our services. Plans are in place to increase the amount of local primary care data we hold, and to incorporate mental health data from South London and Maudsley NHS Foundation Trust, also part of King’s Health Partners.

By bringing together information from outpatient, inpatient and A&E services, pathology, observations, medicines and other treatments, as well as demographic and publicly available data, we are able to build a rich and coherent data resource that enables new types of analysis. This resource can enable service planning and redesign, assist with commissioning decisions, and provide evidence for clinical audits and academic research.

Combining information from both our partner acute trusts, Guy’s and St Thomas’ and King’s College Hospital NHS Foundation Trusts, allows us to establish a full picture of how their services are being used. Figure 2 shows the range of outpatient services that are attended by local diabetes patients, broken down here by sex and count of different services, giving a clear the picture of co-morbidity (diseases that occur simultaneously with another) and multiple service usage that impacts the lives of many diabetes patients.

Figure 2: King’s Health Partners diabetes outpatients in 2015/16. Residents of Lambeth and Southwark shown as number of patients attending additional outpatient services, by sex and count of services

Information from public data sources are also brought into the Data Warehouse, examples include: the National Diabetes Audit, the Quality Outcomes Framework from the Health and Social Care Information Centre, and social deprivation indices from the Department for Communities and Local Government. Figure 3 shows the relationship between obesity and social deprivation in our local diabetes outpatient population.

Figure 3: King's Health Partners outpatients in 2015/16. Residents of Lambeth and Southwark. Obesity levels in diabetes patients with publicly available information on social deprivation.

The continued development of the King’s Health Partners diabetes Data Warehouse will enable clinical and publicly available data to be put to full use to further research and understanding, and to develop new approaches to treating people with diabetes. Using and linking diverse data sources allows us to gain a wider perspective than using clinical data alone. We can look at potential social factors that are specifically related to our patient populations, giving us better context and insight into the health and social issues our local populations face. 

Working collaboratively across King’s Health Partners Institute of Diabetes, Endocrinology and Obesity is developing a world class clinical and research led institute and network that will help us to collectively bring about significant changes to the way we manage diabetes and related conditions in the future. Data and informatics will continue to play a crucial role in this.