Validating pathology data against known shifts in clinical practice
9 September 2019
OpenPathology is our project exploring and feeding back to clinicians on their test requesting rates.
Having established that patient counts are subject to some inaccuracy in existing analyses (see our previous blog), it’s important to sense check the accuracy of our own list size data.
Because two members of our OpenPathology team are also clinicians in North Devon, we have the opportunity to compare our data with known changes in practice.
Since about 2014, there has been a sustained and collaborative approach to improving the use of testing in primary care in North Devon. This has led to overall primary care testing rates in North Devon falling by 14% for full blood count testing and 22% for liver function tests, but without a reduction in the number of tests showing possible significant pathology.
The impact on the volumes of liver profiles requested by primary care can be seen in our OpenPathology data. Cornwall and North Devon had similar requesting rates in 2014, which have since diverged significantly:
This finding supports the notion that different areas are likely to have generally similar requesting rates for many analytes, but that these can be modified by specific interventions, and increases confidence that our list size data is likely to be correct.
It also shows the strength of the Bennett Institute approach: clinicians, academics and software engineers working closely together, across disciplines and different parts of the NHS, to find new approaches and validate existing ones.
We would like to continue to validate this approach against known changes in clinical practice. In particular, we would like to find areas that have deliberately targeted the renal profile for demand management, or perhaps who have removed the potassium from the standard profile (as we know has occurred in some areas due to concerns about specimen stability). When we have data from more laboratories it will be interesting to explore these issues in more detail. If you are in a position to help us get data from more labs, please get in touch.