The impact case studies below detail how we are:

• Improving outcomes through early sepsis alerting
• Using machine learning to analyse patient feedback and make changes to care
• Monitoring the COVID-19 vaccination uptake across NWL to support the vaccination roll out
• Using the results of one of our large observational studies to improve the future care of patients with heart problems
• Understanding the number of patients who are hospitalised with COVID-19 using linked primary and secondary care data

Impact Case Study #1 - Embedding patient experience feedback for quality improvement in healthcare

Funding awarded from the Healthcare Foundation’s Innovating for Improvement Scheme was used to develop and implement a machine learning algorithm analysing free-text responses from the NHS Family and Friends Test, enabling Imperial College Healthcare NHS Trust, and its frontline staff, to act on patient experience feedback for quality improvement (QI), in near-real time. This innovation was the first of its kind in the English NHS, and was recognised by winning the 2019 British Medical Journal’s Digital Innovation Team of the Year Award.

This successful project provided evidence to support the NHS Insight and Feedback Team (NHSE/NHSI) Friends and Family Test policy redevelopment which was launched in September 2019. Through a national selection process, a further four NHS Trusts have been selected to implement this digital platform to create an IT Toolkit for NHS England. Through a further three year programme of work (Health Foundation Spreading Improvement Scheme) Imperial College Healthcare NHS Trust, is partnering with 25-30 NHS organisations to evaluate the sustainability of NHS Trusts using free-text patient feedback, in near real-time, as part of a culture of patient-centred working, and identifying and implementing improvement opportunities from that feedback.

Impact Case Study #2 - Improving outcomes for patients with sepsis

NIHR Sepsis DiAlS Programme is a large study run by Imperial College and Trust, collaborating with Oxford University NHS Trust, Chelsea and Westminster NHS Foundation Trust, University College London, Reading and Berkshire NHS Trust, and the University of Cardiff to understand and improve early warnings for sepsis in NHS Patients.

The study builds on earlier research based on Imperial’s sepsis algorithm (and the 2020 paper published in the Journal of the American Medical Informatics Association) providing robust evidence that the real time digital alert system in the Electronic Patient Record was providing direct benefit to patients. Analysis and findings from the study were directly incorporated into the quality improvement process and helped support changes in the design of the digital alert system to improve the response rate to the alert and acceptance amongst clinicians using the digital alert system.

We are now expanding the project further to include new sites through a new NIHR Health Informatics Collaborative Theme.

Impact Case Study #3 - COVID-19 Vaccine uptake and effectiveness

The North West London (NWL) integrated dataset links vaccination data with NWL patient population and hospitalisation data enabling agile real-world analysis to be undertaken. This has provided further evidence that a single dose of either the Pfizer/BioNTech vaccine or the Oxford/Astrazeneca vaccine was effective at reducing the risk of testing positive for COVID-19 (up to 60 days) across all adult age groups, ethnic groups, and risk categories in a diverse urban UK population, when compared with not being vaccinated. Furthermore, the risk of contracting COVID-19 and/or becoming hospitalised after vaccination has been demonstrated to be very low in the vaccinated population, highlighting an important vaccine safety question that may contribute to vaccine hesitancy. There was no definite evidence to suggest COVID-19 was transmitted because of vaccination hubs during the vaccination roll-out in NWL, although a small increase in the rate of testing positive for COVID-19 in days 7-13 after vaccination does suggest some potential for transmission at or after the time of transmission. This reinforces the need for clear messaging of the importance of continuing to follow social distancing restrictions after vaccination and for up to 21 days.

The rates of people declining a vaccine varied considerably across all ethnic groups and when stratified (grouped) by age. In addition, there was a strong correlation (connection) between socio-economic deprivation and vaccine uptake, with those living in the most deprived areas more likely to decline a vaccination Using the unique opportunity of the NWL WSIC data, geo-spatial (data which includes a geographical aspect) representations of lower super output areas identified communities of between 1000 and 1500 citizens, where rates of vaccine uptake were lowest. This enabled an important focus for quality improvement, public and targeted societal engagement and outreach initiatives to improve vaccine uptake across all population groups.

Impact Case Study #4 - RECAP (REmote COVID-19 Assessment in Primary Care): A risk score to predict hospital admission using a learning system approach

Patients with COVID-19 have a variety of symptoms and it is hard to predict who will become more seriously ill with the diagnostic and assessment tools currently available for primary care clinicians using remote consultations. The RECAP risk prediction tool is benefitting patients by providing clinicians with a systematic way for assessing patients and guiding decisions about a patient’s treatment and management of their care according to their likelihood of severity of their infection, contributing to a decrease in severe disease and death.

This project is now validating, (using data-driven evaluation), a primary care early warning score (RECAP – Remote COVID-19 Assessment in Primary Care) that is specific to COVID-19 and based on data that can be reliably collected during a remote GP consultation. Patients with COVID 19 (which has been clinically diagnosed) have been recruited (recruitment is now closed) after attending their GP. In addition we have gathered data from the NHS 111 COVID Clinical Assessment Service, to enable validation of the early warning score in the NHS 111context as well.

Impact Case Study #5 - Using real-world evidence to improve patient outcomes from myocardial infarction (heart attack)

The senior N-stemi analysis of routinely captured electronic patient record clinical information from five large NHS organisations as part of the NIHR Health Informatics Collaborative Cardiovascular Theme aimed to mirror a clinical trial comparing surgical and medical interventions in patients aged 85 or more. The current standard care pathway for these patients in the NHS uses medication and not surgery. This study provided strong evidence that very elderly patients could have significant benefits in survival as a result of surgery rather than just medication (if they were considered healthy enough to have surgery).

The evidence has been presented to the Imperial College Healthcare NHS Trust cardiology clinical reference group and recommendations to change the local care pathway for these patients is being considered. An evaluation of this change in care would then feed into national changes through NICE.