The riddle of the model 2: Unnatural variation

This is just a quick follow-up to yesterday's post about trends in residential and nursing care for adults with learning disabilities, prompted by Steve Broach's typically perspicacious question about variations in residential and nursing care across different local authorities in England. So, here are a few bits and pieces from another look at the NHS Digital data.

First, the two big graphs below (they wouldn't fit into one, I discovered) are simply the numbers of adults with learning disabilities aged 18-64 and 65+ in residential and nursing care homes during 2015/16, by individual local authority. It's important to realise that these are numbers of people (rounded to the nearest 5), not rates of people in residential or nursing care per 100,000 population. Obviously, local authorities with bigger populations might be expected to report larger numbers of people in all types of service. [I've tried making the graphs extra large in Blogger in the home that people can expand them on their devices - if not, do shout, as they're pretty weeny].

Second, the next graph shows the same information (so on numbers of people rather than rates) broken down by region rather than local authority.

Last, the two snippets below come from the @ihal_talk publication People with Learning Disabilities in England 2015, and relates to 2013/14 information rather than 2015/16 information [conflict of interest note - I was responsible for putting this report together], although the variation would be very similar for 2015/16. This analysis of variation in usage of residential care is based on rates of placement of people in residential care homes rather than raw numbers, and shows quite extreme variation across local authorities in how much they use residential care.

Be careful what you wish for Steve!