Statistical niceties

It is right that Dr Foster highlights the differential death rates for hospitals. And it is also right to highlight any disparities between their findings and those of the hospitals inspectorate.

However, it is noticeable that one important fact, buried in the Observer report, has been studiously ignored by the BBC and most follow-up reports:

Overall, the hospital standardised mortality ratio (the actual number of deaths against the expected number) fell by 7% last year. That means 14,500 saved lives.

Of course poorer hospitals should do better for those patients who experience sub-standard safety and cleaning, and on the finding that across the system, to quote the Observer again:

5,024 people died after being admitted for "low-risk" conditions such as asthma or appendicitis, of whom 848 were under 65. A proportion of those deaths will be linked to safety errors.

But is it too much to ask for some context in reporting the whole story?

This post has been picked up by John Rentoul.

Healthy progress?

In what exam is a 93% score regarded as a failure? Yet this is how the BBC reported the news that the government's sensible decision to order a deep clean of our hospitals has led to 93% of hospitals being fully cleaned, with a 35% reduction in MRSA over a year. The point about targets is that they should be challenging: it is far better to get 93% on a tough target than 100% on an easy one. And this is one area where the government was right to respond - despite suggestions from shadow health secretary for life Andrew Lansley that it is no business of government to do so. After all, were not 'cleaner hospitals' one of the main planks on which the Tories fought (and, admittedly, lost) the last election? Turns out that the Tories, whose health policy is now written in BMA headquarters, are no longer thinking what we are thinking about this issue.