So it’s been a while since I’ve actually made just a ridiculous number of graphs, fortunately this drought is at an end. Thanks to Open Prescribing we have access to an easily searchable repository of GP prescription data. I’ve used it before to show trends in statin prescription, how for example Atorvastatin is taking Simvastatin’s lunch money.

Comparison of Statin prescriptions over time
Comparison of Statin prescriptions over time

While this is all very well and good, I was wondering what we could combine with this information to draw some more interesting conclusions. An answer came from the MHRA (Medicines and Healthcare products Regulatory Agency), in the form of drug safety updates. Over at www.gov.uk/drug-safety-update is a curated list of alerts issued by the MHRA that go out to healthcare workers whenever new information about a medicine is discovered.

I wrote a scraper that searched for the names of drugs appearing in this list, then plotted the dates these alerts appear over the Open Prescribing data. What you get is something like this:

Calcitonin was found to increase cancer risk in August 2012
Calcitonin was found to increase cancer risk in August 2012
Hydroxyzine was found to increase risk of arrhythmias in April 2015
Hydroxyzine was found to increase risk of arrhythmias in April 2015
Domperidone was found to cause cardiac side effects in 2014
Domperidone was found to cause cardiac side effects in 2014
In 2012 new contraindications were put in place for Aliskiren following new information on cardiovascular and renal adverse reactions
In 2012 new contraindications were put in place for Aliskiren following new information on cardiovascular and renal adverse reactions
The use of Strontium Ranelate became more restricted in 2014 following the discovery of great cardiovascular risk
The use of Strontium Ranelate became more restricted in 2014 following the discovery of great cardiovascular risk

It should be noted at this point that these instances don’t mean that medications were suddenly found to be killing large swathes of people. In most cases when adverse information is found out, requirements become stricter or patients are more closely monitored. Just because a drug is found to be slightly more dangerous doesn’t mean it isn’t still the best choice in a particular scenario.

If anyone wants to mess around with the code or verify my findings all the code can be found at glfharris/MHRA-Updates. Quick disclaimer, most of the hits weren’t as interesting as the ones I’ve shown. Drugs with common names/uses can be mentioned in alerts not about themselves, and not all alerts are that significant or wide ranging. Here for example was the plot for Cetirizine.

Cetirizine prescriptions over time.
Cetirizine prescriptions over time.