Having recently moved to London, I decided to consult the results of the GP Patient Survey commissioned by NHS England to help me find a new practice. I was surprised to realise that while there were a number of practices in my vicinity, the percentage of people describing their overall experience with the surgery as good was consistently lower than the national average. Had I been unlucky in my choice of neighbourhood, or was this part of a wider trend in London, or in cities in general? Thankfully, the GP Patient Survey publishes its results for anyone to download, so I took to the data to find out.

To get an idea of what I was looking at, I started by calculating the overall percentage of respondents rating their overall experience as good for each of NHS England’s regions in January 2016 to produce the table below:

Figure 1: Percentage of residents rating overall experience as good or very good per NHS Region, January 2016

Already I felt I might be onto something – at the bottom were places I knew to be densely populated, such as the Central Midlands, and especially, London, while at the top were places I knew to be much less densely populated such as the South West and Cumbria. My hypothesis: that there is an inverse relationship between population density and average patient satisfaction with their local GP.

Using open source software QGIS, I constructed a visual to show the results at a more granular level, that is, at the level of CCGs ( Clinical Commissioning Groups), the 209 NHS organisations responsible for health of local areas. Using more publicly available data to get the population of the CCGs, and harnessing QGIS to calculate the areas of the CCGs, I was also able to make a corresponding visual representing population density. Placing these side by side, the relationship becomes more apparent:

Figure 2: Maps showing: the proportion of respondents describing experience as good or very good per CCG; the number of residents per square kilometre per CCG

While the correlation is by no means perfect, it is certainly apparent to the eye, both on a wider scale (with a high concentration of red/orange around London and Birmingham in both maps) and in a number of particular CCGs (Thanet at the eastern tip of Kent, or Kingston upon Hull on the Humber Estuary, for example, standing out from their neighbours).

The correlation is also apparent when these CCGs are plotted on a scatter graph:

Figure 3: Proportion of respondents rating overall experience as good or very good against the number of residents per square kilometre (log scale)

Indeed, the Pearson product-moment correlation coefficient, a common measure of the extent to which two variables are related, is -0.51: a discernible negative correlation. As anyone who has spent 5 minutes in a statistics class will testify, this does not necessarily imply a causal relationship, but it is fair to say from this analysis that low population density and high patient satisfaction are often found together.

Returning to our original question, where exactly in England are you most likely to be satisfied with your GP service? The table and map below show the highest proportion of happy patients in the very rural Hambleton, Richmondshire and Whitby CCG in North Yorkshire, the most sparsely populated of all the CCGs, though the residents of Surrey Heath are not far behind, with a few more neighbours to share their positive experiences with. At the other end of the table, London boroughs take 3 of the bottom 5 spots, with the urban areas of Bradford City and Slough home to the least satisfied patients.

Figure 4: Top 5 and bottom 5 CCGs for patient satisfaction

So why are patients more satisfied with their GP care in less densely populated areas? Do more densely populated CCGs face additional challenges, or is the CCG funding formula at play? Or are we city dwellers just more difficult to please? This blog post leaves these questions for another time, but for now we can conclude that if you’re deciding where in the country to live, relative disposable income isn’t the only measure where London doesn’t come out on top.