Looking at service provision against assessed need across London’s Primary Care Trusts (PCTs):
As part of our group’s work at the 20th February Intelligent Impact event we decided to look at whether service provision met need. We found data on the number of people who have low level depression and anxiety disorders in each London PCT area and data on the number of referrals to psychological therapies in each London PCT area from the Improving Access to Psychological Therapies website . The number of people diagnosed in each London PCT area (called KPI1 in the dataset) came from the Psychiatric Morbidity Survey conducted in 2000. The data on referral to psychological therapies was from 2011-12. We also needed to merge and add data from each of the four quarters to get total data for 2011-12, which we did in a Google fusion table.
We also looked at the number of antidepressant prescriptions adjusted for assessed need in each PCT area. These figures were available from the UK Government’s datastore http://data.gov.uk/ but only for the year 2010-11, the year before the data on psychological therapy referral. Nevertheless it again showed huge variations in the number of anti-depressant prescriptions adjusted for need across London PCT areas
When anti-depressant prescriptions and referrals to psychological therapies were plotted according to each London PCT area, there was the suggestion of a trend for the PCT areas with the highest levels of referral to psychological therapies to have lower numbers of anti-depressant prescriptions and vice versa (with two outlying exceptions).
Looking at anti-depressant prescription adjusted for need across London, http://batchgeo.com/map/cf74c597facfc1cbfc4bbeaf6bee574b, it can be seen that anti-depressant prescription against need appears to be much higher in outer than inner London. In contrast relatively low levels of referral to psychological therapies in relation to need seem to take place in outer London, particularly the north west and west http://batchgeo.com/map/f49e1701ecb0272af967c6bb6da1b626
Here is another way of representing these figures on openheatmap
What did we learn?
Open data allowed us to look at the need for treatment for low level depression and anxiety problems across London and then compare the amount of treatment (both anti-depressant prescription and referral to psychological therapies) provided in comparison with this need. It showed significant differences across London PCT areas, suggesting that neither anti-depressant prescription nor referral to psychological therapies reflected assessed need in each PCT area.
There was some suggestion that inner London PCTs were more likely to refer their patients to psychological therapies, while outer London PCTs, particularly in the north west and west of London were more likely to prescribe anti-depressants.
Data was difficult to find – you need to understand NHS jargon and know which websites to look on to do this type of analysis.
The data has several limitations – particularly as it was difficult to find data with matching timescales: the needs assessment came from survey data from the year 2000 while anti-depressant prescription data was from 2010 – 11 and referral to psychological therapies from 2011- 12. The latter was also collected quarterly and so had to be summed to provide a comparison with the annual number of anti-depressant prescriptions.
We couldn’t find any data set that mapped PCT boundaries as shapes, so mapping was limited to a simple latitude and longtitude measure for each London PCT, which made it more difficult to visualise the geographical differences across the capital.
This piece of work currently offers little opportunity for interaction with other open data sets and analysis – we should ensure that the finalised datasets are available on our new Your Voice Your City folder within http:www.datahub.io, and link with other related analysis perhaps through the reference manager http://www.mendeley.com/ or by using more visual analysis through software such as Tableau http://www.tableausoftware.com/.
It is also pretty dry – we could look at using Sparkwise http://sparkwi.se/ to link data visualisation with videos, blogs or with an analysis of how the subject is being discussed on social media such as Twitter and Facebook. Storify http://storify.com/ is another way of linking up stories, videos and social media to engage people in events or campaigns.
But we were pretty pleased with the information that we managed to get together in such a short time.
LVSC will be looking at how we can better link this type of data analysis (‘stats’) with ‘stories’ and display it on our website in a way which engages people and supports our research and campaigns. We need to let you see the raw figures from any data we, and others, collect, analyse it in a way that makes it easy to understand, communicate it better using new visualisation techniques and make sure that this intelligence informs the decisions that are taken in the capital in the future.
Head of Policy and Knowledge