The data does need to be used with care, and I'm not convinced it is particularly useful. To get a useful picture of CCG (not Councils) practice in relation to CHC you would need to know the numbers of applications for CHC (not just the "newly eligible") both in absolute terms and percentages by age (or age range) and gender and ethnicity, as well as the conversion rate (i.e. % of applications which were successful - the data doesn't show that) broken down by Clinical Code (i.e. the condition in respect of which CHC was paid), and also by gender, age, and ethnicity. I'm not certain all of this data is captured in a usable form, and its release would be legally problematic as the absolute numbers in each category would be small enough to risk identification of individuals. The low numbers in each CCG mean that to get a useful picture data would be needed over several years, and I'd expect things to be more consistent over a longer term and a wider (and more useful) range of metrics. That there are wide variations isn't something I doubt, but some of the data bears the term "Experimental Statistics" - a polite way of the HSCIC saying it isn't entirely sure the product is particularly useful, at least in at its present stage. Achieving a more apparently consistent set of outcomes would require pinpointing not just the CCG (and the teams managing CHC either in CCGs or CSUs), but also issues round the completion of the documentation for the CHC process in respect of different conditions.
W