Developing and implementing machine learning driven analytics for quality improvement in health care King’s College London
- Led by King’s College London, in partnership with the Sentinel Stroke National Audit Programme (SSNAP), University College London and the University of Manchester.
- Developed tools to detect and interpret variation in stroke care quality and patient outcomes, and support better use of data for quality improvement in stroke care.
- Integrated ‘machine learning’ algorithms into the national clinical audit for stroke and validated the model with the national stroke quality register in Sweden.
- Ran from March 2018 to June 2022.
Translating the increasingly large and detailed amount of health care data into information that can be used to improve care quality is challenging. Artificial intelligence, and in particular approaches based on ‘machine learning’, has developed rapidly over recent years. Machine learning involves the use of algorithms to detect and learn from patterns in data.
This project led by King’s College London used machine learning to support better use of data in order to improve the quality and consistency of care for stroke patients.
The project team completed training prediction models for 30-day mortality after stroke using machine learning methods, the results of which have been published in BMC Neurology. The training models have been published in a Github repository to make it easy for others to use, build on or implement the work.
Training prediction models have also been completed for stroke-associated pneumonia, a common and serious complication of stroke.
External validation of an adapted version of the mortality model was carried out with the national stroke quality register in Sweden. This showed that the model had excellent accuracy and could be used to make fair comparisons of stroke quality and outcomes across countries.
The model will be used by the SSNAP for benchmarking NHS stroke services from 2023. This will be the first example of machine learning being used at scale by national clinical audits/quality registries to support quality improvement benchmarking, as well as one of the first examples of different health systems aligning to benchmark and compare quality of care across countries.
Various papers have been published on this project, and presentations held, to disseminate the findings widely.
Contact information
For more information about this project, please contact Dr Wenjuan Wang, Research Fellow/Senior Data Scientist, King’s College London.
About this programme
Related links
Work with us
We look for talented and passionate individuals as everyone at the Health Foundation has an important role to play.
View current vacanciesThe Q community
Q is an initiative connecting people with improvement expertise across the UK.
Find out more