Visualising Brighton GP patient numbers
10th June 2018
With increasing frequency in the local Brighton news, I read about GP surgeries closing. This became very apparent to me in January 2017 when I was reading about another closure to suddenly realise that it was my own doctor. On top of that I was unable to register at the nearest surgery to my home, only three-hundred meters away because they were not accepting any new NHS patients. With the number of surgeries closing, a total of nine since 2015, I was wondering whether any new surgeries have opened to counter this, at which surgeries the displaced patients have registered and whether it has caused any overload. In this blog post I will visualise some geospatial data and explore these questions.
The NHS has an open data service, NHS digital which makes it very convenient to find the relevant data. Although I did find it limited in terms of historic data, only dating back to 2013 for the information I was looking for, but that gave me enough to look at the change in Brighton GP surgeries up until October 2017.
To begin with I looked at both the number of GP surgeries, and the total number of patients registered at each within the city of Brighton and Hove. I have used both size and colour to show the number of patients registered at each surgery, with the colour ranging from black for the smallest to red for the largest. We can see by looking through the years that the total number of surgeries decreases as well as the total patients of some increasing, this is particularly noticeable when quickly switching between 2013 and 2017. As I used the postcodes to map the latitude and longitude it can noticed that there are sometimes multiple circles in the same location, this is where there was more than one surgery within the same postcode.
- April 2013
- January 2014
- January 2015
- January 2016
- January 2017
- October 2017
- In 2014 a surgery in Woodingdean seems to have moved temporarily before moving back again, either that or there is an anomaly in the data. I haven’t been able to find any other record of it moving for a single year so I have left it with a footnote as I have no evidence either way.
Plotting a box and whisker plot shows the overall increase in patient numbers, particularly between 2016 and 2017 where the median has increased as well as the third interquartile range. It is also noticeable that an very large outlier surgery first appeared in 2016.
As interesting as it is to see the median number of patients per surgery, the number of patients alone is not a great indicator because a single surgery can house multiple doctors. To account for this I downloaded some data containing the number of doctors present at each surgery. I wasn’t able to find data ranging back to 2013, but I did find some covering a twenty-one month period between January 2016 and October 2017, and using this I was able to calculate the change in doctor to patient ratio. In these plots, I have changed the aesthetics slightly, using size for total number of patients, and colour for the doctor to patient ratio. This gives a much more nuanced picture than the patient numbers alone as we can see the ratio both increasing and decreasing in different surgeries across the city.
- January 2016
- October 2017
Looking at a box and whisker plot, the median ratio has risen but also the third interquartile range has shrunk along with the second interquartile range growing. This confirms what the map visualisation shows us, that patients seem to be spreading out between the surgeries, but that there is an overall trend of the doctor to patient ratio increasing.
By using both transparency and colour to highlight the difference between the two years, giving an increased definition to those that have increased their doctor to patient ratio, we can see that it’s the smaller out of town surgeries that have had the highest change, with three in particular standing out. Around the center of Brighton there seems to have been a mixture of both reductions and increases.
Overall it is a more complex picture, and not simply one of all surgeries increasing their list size. There has been an overall increase in the median doctor to patient ratio, and we can also see a reduction in the total number of GP surgeries with a move towards larger multi-GP surgeries. It would be beyond the scope of this blog post to make any comment on the whether this changes the quality of care as the point here is just to present the data. As I mentioned in the introductory paragraph, the surgery nearest to where I live is the one with the greatest change in ratio so it’s not surprising that they were not accepting new patients. This was not a problem for me because I can easily travel, but for the elderly, the disabled, and those with mobility problems a reduction in the number and spread of surgeries is likely to be more of an issue.
To visualise this data I used R with the Tidyverse and ggmap packages. All of the data with the source links, and commented code is available to download from my GitHub acccount.