Dodgy Figures

When the government (Number 10, not the Department of Health) produced their "listening" exercise leaflet Working Together For a Stronger NHS there was a lot of cynical responses, particularly over the "statistics" it contains.

The first response in the Press came from John McTernan in the Telegraph. McTernan said that the leaflet broke the Civil Service Code of Conduct ("you must set out the facts and relevant issues truthfully" and "You must not deceive or knowingly mislead Ministers, Parliament or others"). He also pointed out that if the leaflet was an advertising flyer (it isn't, and so is exempt) then it would break the Advertising Standards Authority code ("must not materially mislead or be likely to do so" and "objective claims must be backed by evidence"). McTernan pointed out that the "statistics" in the leaflet were misleading and were intended to mislead.

More recently, Ben Goldacre wrote a piece about the leaflet in the Guardian. Goldacre rips apart the "facts" and then concludes:

This stuff matters, and the facts in this plainly political pamphlet should be clean, correct, transparent, and justified. As the government defies all reason by claiming that NHS staff support their reforms, I can only fear the results of their listening exercise. 

It appears to me that the article has suffered a little in editing, and the figures need a bit more explanation, which I will do here. Goldacre covers three "statistics" presented in the leaflet:

1. "If the NHS was performing at truly world-class levels we would save an extra 5,000 lives from cancer every year."
2. "Since May 2010 the NHS has gained 2,550 more doctors and has 3,000 fewer managers."
3. "95% want more choice over their healthcare"

The first "statistic" references an academic paper that analyses cancer figures over the period 1985 to 1999 and asks the question "if cancer survival in Britain were the same as in Europe: how many deaths are avoidable?" over that period. (The answer is that if the Thatcher and Major governments had taken action to reduce cancer deaths to European levels there would have been 7000, not 5000, fewer deaths). The paper does not make a comment on cancer survival now, it makes a comment on cancer survival during the last Conservative administration. This figure is clearly irrelevant.

The second statistic uses figures from the NHS Information Centre. However, it presents the information in a misleading fashion. We know that year-on-year the NHS has to do more work, and the number of clinicians will increase accordingly (hopefully). It is the trend that is important: how does the change in the number of doctors compare with previous years? The following graph is data from the NHS Information Centre where I have plotting the numbers of all doctors (not GPs) expressed as Full Time Equivalent from 1995 to 2010:

The graph shows basically two trends, 1995 to 2001 where the count was increasing but shallower than the second trend, 2001 to 2009. These two trends are approximately linear.

If I fit a linear trend line to this graph between 2001 and 2009 it yields a gradient (and therefore, an average increase of) 4,859 doctor per year. The increase of 2,600 quoted in the Cameron leaflet is hardly something to crow over. The plot of FTE shows clearly the drop off in the increase in the numbers of doctors.

What about the number of managers? Well, here is the graph using data from the NHS Information Centre (again, FTE):

Again the leaflet is correct, the figures shows a downturn, but look at the trend over time. The change in the number of managers is quite complex, but it is clear that the figure is not increasing every year. In 2006 and 2007 there were falls in the number of managers, and the drop in the last year is equivalent to the large fall in 2006. If you think that the fall in the number of managers is a good thing (which itself is contentious) you cannot deny that it is a new trend. Why hasn't the government said that the fall in 2010 is equivalent to the 2006 fall?

The last figure the government gives is bizarre. As Goldacre points out, there are no figures about the "choice" that patients want, in fact most evidence shows that patients want to go to their nearest provider (hospital, GP) rather than making any other choice. The government has taken the results from the British Attitudes Survey and Mark Hawker has given a possible explanation of where the government gets a 95% figure from given that they never asked anyone how much choice they wanted. It appears that they have aggregated several responses together and say that those who want choice are not those that do not want choice. This is rather disingenious at best.

The Department of Health's Cooperation [sic] and Competition Panel say that

40% of patients have access to at least three NHS Trusts within less than 20km, and around 80% of patients have access to at least three NHS Trusts within 40km. On average, patients travel around 12km to their chosen provider.

This indicates that most patients choose their local hospital.




The Kings Fund's have also researched this area and have found that of those patients who were given a choice of provider (a right since 2006 that will not change under the Health Bill), 69% chose their local provider: a figure greater than that found by the CCP.

However, it is significant that the CCP point out that when there is bad Press coverage of their local hospital patients choose to go elsewhere (they use the example of adverse publicity over quality standards at Basildon & Thurrock NHS Foundation Trust). This does not mean that people want to go elsewhere, just that they usually choose their local hospital, but only go elsewhere when they see a reason not to choose their local hospital. It's clear that choice is not the important thing, the important issue is ensuring that your local hospital is of a high quality.