Playing With Department Of Health Statistics

Text - Permalink

Playing With Department Of Health Statistics

Recently, the Department of Health released a leaflet entitled “Working together for a stronger NHS” that emphasised the fact that 95% of people want more choice over their healthcare. Interesting, I thought. The data is from the British Social Attitudes 25th Report published January 2009. Looking at the data from the survey (2007), I found that it contained multiple questions about choice:

How much choice do you think NHS patients should have about which hospital to go to if they need treatment?
And in your area, how much choice do you think NHS patients actually have about which hospital to go to if they need treatment?
How much choice do you think NHS patients should have about the kinds of treatment they receive?
And in your area, how much choice do you think they actually have (about the kind of treatment they receive?)

Not only that, but the rating scale ranged from “a great deal” through “quite a lot”, “a little”, and “none at all”. Looks like the Department of Health added up the first three answers to come up with the ~95% which appears to be averaged between hospital and treatment ratings. But, more interestingly, there was data on how much choice the public actually felt they had. If you take the should have and actually have data you get the following table:

The contingency table shows row percentages, and so for example 6% of people think they should have choice about which hospital and feel they actually have it. The 96% is the (rounded) figure provided by the Department of Health. The table shows that 49% think they should have quite a lot of choice (whatever that means), and 42% think they actually have a little choice (whatever that means). The colours depict where choice is exceeded (green), matched (orange), or not matched (red). For example, if you think we should have quite a lot of choice and actually have a great deal of choice that is assumed to be exceptional. However, there is little evidence to suggest that if you want no choice and get lots of choice that is a good thing. Adding up all of the ~~percentages~~ true values: 5% actually have choice that exceeds what they think they should have; 26% actually have choice that matches what they think they should have; and, 69% actually have choice that does not match what they think they should have.

NB: I think I’ve added up right, but if not…

NB: Percentages updated. Thanks, Chris!