Close
Principal Investigators: Dr Robert Aldridge and Dr Alireza Moayyeri , UCL and Farr London
Investigators: Prof Harry Hemingway, Dr Spiros Denaxas and Ms Hannah Evans, UCL and Farr London; Prof Liam Smeeth, Dr Krishnan Bhaskaran and Dr Rohini Mathur, London School of Hygiene & Tropical Medicine and Farr London; Dr Jürgen Schmidt, Dr Andrew Hughes and Dr Julian Flowers, Public Health England
The Challenge
Having a reliable and detailed estimate for the burden of different diseases affecting the population is important for management of healthcare resources. Different regions of a country may have different priorities and they can change over time. Effective investment in healthcare provision, public health and research needs a common way to compare current disease across health systems. To have maximum value, these comparators need to be compatible at regional, national and international scales.
The Research
A team from the Farr Institute at University College London and Public Health England aims to use electronic health records from all regions of England to estimate burden of various diseases in different regions of the country over time. As the starting project, they estimated burden of three common and important diseases: cancers, diabetes and low back pain. The data included linked GP and hospital data with mortality records. Results were compared to the current best methods for population estimates of diseases, i.e. using national cancer and diabetes registries or from health surveys. The research also estimated burden of low back pain, for which there is no available data from national registries.
The Results
The work developed valuable new methods for rapid comparative measures of disease burden. Detailed estimates for cancers and diabetes were generally consistent with comparator estimates. Burden estimates for each disease can be viewed by years (2000-2014), age groups, sex, ethnicity, socio-economic deprivation, and geographic regions. It shows that linked electronic health records have the potential to improve disease burden estimates for various diseases and conditions including patients with multiple morbidities, multiple risk factors and diseases with low death rates.
The Impact
The results hold significance for regional and national policy makers on resourcing healthcare and public health activities. The approach exploits existing electronic data sources, reducing the need for data gathering. The methods can apply to where registry data does not exist and provides new knowledge where disease is rare at the local level. It also offers new opportunities for studying sub-populations where people have more than one disease (multi-morbidity). The methodology is scalable for comparing local to international disease burden estimates (as required by the Global Burden of Disease Project).
For more information: Newton et al. (2015) Lancet 386: 2257-74; Herrett et al. (2015) Int J Epidemiol;44:827-36.