As the British Medical Association (BMA)[i] warns of the worst
winter on record for the NHS, a pioneering project by More Metrics,
in collaboration with Lancaster University, may provide answers to
the growing crisis in GP practices. With surgeries struggling with
patient backlogs and a severe shortage of doctors, More Metrics’
innovative approach to assessing GP performance and resourcing
could reshape healthcare in the UK and significantly improve patient
outcomes.
The BMA’s “Weathering the Storm” report stresses an urgent need
for action, highlighting rising demand and limited resources facing
GP practices. More Metrics, a specialist in open-source data analysis
on family finances, health, and social issues, has undertaken a project
to address a critical question: How does the level of GP resourcing
impact health outcomes?
Introducing the GP Performance Grid
More Metrics typically works with insurers and charities, to better
understand mortality and disease risks. Recognising the broader
societal implications of their work, the company sought to deepen
its analysis by partnering with Lancaster University to explore GP
performance in more depth.
This led to the creation of the GP Performance Grid, a tool that
evaluates GP practices using data including patient satisfaction
scores, health outcomes, disease diagnosis rates and mortality risk.
More Metrics created a GP Performance Grid, a framework to
establish minimum resourcing standards for GP surgeries. By
addressing disparities in resourcing, this framework could be a game
changer in improving health outcomes for patients. In parallel, two
MSc students at Lancaster University have been investigating specific
aspects of GP performance to provide additional insights.
Data collection and analysis
The analysis covered 5,900 GP practices in England, using publicly
available data from the NHS, the Office for National Statistics (ONS),
and other open sources. Key data points included GP workforce
statistics, disease prevalence, funding calculations, mortality rates,
and patient satisfaction scores. More Metrics also used Care Quality
Commission (CQC) ratings to check the accuracy and relevance of
the data used in their modelling.
This data provided the foundation for exploring the link between GP
resourcing levels and health outcomes. International studies have
shown a strong correlation between the number of doctors and life
expectancy. More Metrics aimed to determine whether a similar
relationship exists at a more localised level, where factors like social
deprivation and the availability of resources could have a significant
impact on health outcomes. The challenge was not straightforward.
The complexity of resourcing and health outcomes
The relationship between GP resourcing and health outcomes is
complex, and almost always confounded by levels of deprivation. To
address this More Metrics used exploratory data analysis to work out
how they might “unpick” these two effects before embarking on more
detailed work. Their approach involved several stages: gathering data,
analysing GP resourcing levels, statistical modelling of excess deaths,
and using the Performance Grid to present findings. This process
enabled the team to propose an appropriate minimum GP-to-patient
resourcing level for UK practices.
The correlation between GP resources and excess deaths
The Performance Grid provided a way to separate ‘resourcing’ from
‘excess deaths,’ allowing a clearer view of how practices compare.
While there is some correlation between resourcing and mortality,
outcomes are also influenced by GP performance.
While two GP surgeries may have identical resourcing levels, one
may have experienced doctors and a very efficient support team and
another less experienced staff and poor processes. It is likely patient
health outcomes will be poorer in the second practice, with shorter
life expectancy and excess deaths.
To isolate factors unrelated to resourcing, More Metrics analysed
disease diagnosis patterns, adjusted for age, sex, and neighbourhood
health, against excess deaths as a proxy for GP efficiency. Using
a grid structure, they demonstrated it is possible to provide a GP
performance assessment that separates out the two key drivers.