The Health Secretary Jeremy Hunt has been going round saying* “At the moment, for example, if you have a stroke at the weekend you are 20% more likely to die. That cannot be right, and that is something every doctor wants to sort out as well.”. This is a rather alarming way of saying that 11.1% of those admitted on a weekday because they have had a stroke are dead within 30 days, whereas 12.9% of those admitted on a weekend die within 30 days. This is from work of Roberts and coworkers.
Roberts and coworkers go on to note that fewer people are admitted to hospital due to a stroke on a weekend day than on a week day, so there looks to be some difference between how strokes occur and are dealt with between weekdays and weekends, that occurs before they turn up at a hospital. This could be skewing the statistics, indeed it almost certainly is to some extent, but we don’t know by how much.
Mr Hunt is using a statistic to push the idea of a “seven-day” service by the NHS. In this context, I found the paper by Meacock, Doran and Sutton, interesting. The paper is in the journal Health Economics, and looks at a seven-day NHS from the point of economics. They estimate running the NHS like a weekday for all seven days would cost an extra roughly £1 billion/year (for context: this is a bit less than 1% of the annual cost of the NHS to the taxpayer). They then say, what if Jeremy is right and the excess deaths (for all causes not just stroke) for weekend admissions is due to hospitals, and so can be eliminated by spending this £1 billion per year. Note that this is dodgy assumption, but ok let’s go with it for now.
Now, these excess deaths are around 4,000/year, so spending £1 billion to eliminate them works out at £250,000 per life saved. This is too much, in the sense that a drug that cost that much, may well be rejected by the NHS as too expensive. The NHS has a limited budget so in effect there is rationing, apparently at a level that would reject spending £1 billion/year on a seven-day NHS.
Simply speaking, roughly 1 million people die in the UK every year, but being admitted at a weekend and then dying with 30 days is actually a relatively rare way to go**, so there may well be better ways of spending limited NHS resources than on employing more doctors, nurses, etc. on a weekend. Better here being in the sense of preventing more deaths.
*That link is to a, I think very good, article in Buzzfeed. Go them for that, and they have lists of cute kitten pictures too. Good numerate articles on stats, and pictures of cute kittens. I don’t think I need any other website, apart from Wikipedia of course.
** Sorry this not a very cheery blog post, but maybe if in the unfortunate case that you are admitted to hospital on a weekend, you can reassure yourself with that observation.
Chance & Necessity 