The challenge
The UK faces a sobering reality: we have the worst death rates for lung conditions like asthma and COPD among comparable nations, according to Asthma + Lung UK analysis. Behind these statistics lies a fundamental problem – thousands of patients with respiratory conditions remain unidentified and uncoded in GP systems, missing out on vital reviews and monitoring that could save lives.
Accurate patient coding isn't just administrative housekeeping – it's the gateway to proper care, regular monitoring, and ultimately, preventing avoidable deaths.
Our innovation: smarter detection through machine learning
Working with The Health Foundation, we developed an algorithm support tool that fundamentally changes how practices identify 'missing asthmatics' – patients with undiagnosed or incorrectly coded asthma.
Traditional approaches to improving recorded asthma prevalence demanded significant clinician time, creating an impossible trade-off between thoroughness and efficiency. We knew there had to be a better way.
Our solution? We trained a supervised machine learning model to mimic human clinical judgement, analysing patient records with the precision of a GP but at unprecedented speed and scale.
Transformative results
Practice-level impact: 80% time reduction
Our algorithm delivered immediate, measurable benefits:
- 4.9 hours saved per 100 patients analysed
- Maintained clinical accuracy whilst dramatically accelerating the identification process
- Freed GP time for direct patient care rather than record reviews
National-scale potential: half a million patients found
When we extrapolated our findings from two pilot practices to England's entire population, the implications were staggering:
- 512,000 additional patients could be identified with asthma
- 55,000 GP hours saved – equivalent to 27 full-time GPs working for a year
- £5.4 million in primary care funding released for frontline services
Beyond the numbers
This isn't just about efficiency gains or cost savings. Each identified patient represents someone who can now receive proper asthma management, regular reviews, and preventive care. In a healthcare system where the UK lags behind comparable nations in respiratory outcomes, our tool offers a practical path forward – one that improves patient safety whilst respecting the reality of stretched GP resources.
By combining clinical expertise with machine learning capabilities, we've created a scalable solution that addresses one of primary care's most pressing challenges: finding the patients who need help before they become statistics.
