It’s a slow start to the year policy-wise, which made the Liberal Democrats’ FOI of the figures from the National Patient Safety Agency the major point worthy of comment of the day.

The first thing to say is that the increase over the previous two years’ figures is almost certainly correctly being ascribed to increased reporting of incidents.

The second thing to say is that the stated figure – that hospitals reported 3,645 deaths in 2007-8 from patient safety incidents (including those relating to problems with scans and tests and hospital infections) – still appears very low if we accept the rule of thumb that one in ten people who go into hospital are harmed by the healthcare they receive.

Synthetic outrage
Both opposition parties’ health spokesmen - Lib Dem Norman Lamb and Conservative Andrew Lansley - produced quotes inflated with what they must know is synthetic political outrage.

Lamb told BBC News (www.news.bbc.co.uk/1/hi/health/7813634.stm) "most of these deaths are avoidable and completely unacceptable. Along with very high standards in most hospitals, there are also areas of completely unacceptable practice. At the heart of this problem is weak management and indifferent political leadership. The NHS must get serious about improving patient safety."

Lansley’s take was that “while some of this increase may be down to improved reporting, these numbers are concerning."

This is not only a hospital thing. During a very dull presentation at a conference in 2007, I was shocked into wakefulness by an NPSA official admit that there were PCTs who have never reported an adverse incident or near-miss.

The NPSA’s National Reporting and Learning System (NRLS), which started in November 2003, was a world first in healthcare. By 2007, it has received1,668,437 reports since then – 70,000 to 80,000 per week. The vast majority (72%) of these concern acute care, and a further 10% mental health. Less than 1% come from GP practices.

The vast majority involve no or low harm - hospital slips, trips, falls, bumps and scrapes. Yet 13% are different, involving major harm or killing. Serious near-misses are reported as well as actual events: the ‘accidents waiting to happen’. This enables NPSA to issue speedy advice and early warning to the NHS. Links with such organisations as the MHRA (for devices and medicines) aim to reduce overlap.

NPSA have developed a 1-page web report form, which is logged and acknowledged immediately, and then followed by ongoing dialogue with the organisation in question; review prioritisation; feedback follow-up; and ongoing dialogue.

NPSA links with the DH and SHAs, and is also involved in scoping and data analysis, the international healthcare safety agenda, the NHS Litigation Authority and legal claims.

To forgive incarnadine
“Incarnadine” is a prince of a word. It is used in Macbeth by the tragic hero – its meaning is to dye red with blood. Why use it here? Because if errors are not acknowledged meaningfully and the cause not corrected with proper redesign, they will happen again, and blood will be shed again unnecessarily.

Error is a human thing. However, it is also a bad design thing. Errors happen most frequently where the edges of systems – the meeting or transfer points – are poorly aligned or badly owned by the staff who work across them.

We are not going to eliminate human error from healthcare until we eliminate humans from healthcare. The response to error is crucial.

So how should things be run? Well, in a hospital run by the Daily Mail, there would obviously never be any errors (or if there were, they would be the fault of asylum seekers, illegal immigrants, political correctness and falling house prices). However, once the DM NHS Trust had shot all the scapegoats, they would find that errors still happened. Something Must Be Done in such circumstances, and the whole hospital would have to resign from healthcare.

How, in the rather more messy real world of healthcare, might we proceed against error? Part of the answer could lie in something like the Schwartz rounds covered yesterday (www.healthpolicyinsight.com/?q=node/236).

Getting staff at the most local level – practice, ward, clinic – to own and own up to errors could be very powerful. It would certainly be more powerful than the long and depressing section on complaints at the back of the board meeting papers – although there is much learning on system design to be done in those painful pages.

Driving through my local recycling plant, I noticed a sign saying ‘713 days since last injury caused lost time’. How powerful could something similar, adapted to each setting, be in healthcare? Get a good sense of competition going between wards and teams – so many people in medicine are high-achievers, or want to take personal pride in the work.

Design is one aspect of increasing patient safety. Changing culture is another.