Most clinicians hate targets and view them as something that obstructs clinical care. However, it seems they have worked.

For instance, hospital episode statistics data for 2001 reveals that the total number of cardiac bypass grafts was 23,181, with a median wait of 180 days. Angioplasties numbered 24,773 with a wait of 58 days.

Following a targeted drive, there were 21,596 cardiac bypass grafts in 2006 with a median wait of 68 days, while angioplasties had soared to 59,796 with a wait of 50 days. The median age of patients had also increased, so that treatment was being offered quicker, in greater numbers, to an older population. This is obviously a good thing, but it might have happened anyway given the amount of funding pushed in the direction of cardiac care.

The criticisms of targets include that they encourage "hitting the target but missing the point" and in some cases lead to fabrication. However, the methodology underlying some of the current targets suggest that we will soon be in the novel position of hitting the point but missing the target. This particularly applies to percentage change targets, specifically for infection control.

C difficile targets

Between 2004 and 2006, our hospital's annual C difficile rates quadrupled from 50 to 194. We introduced a restrictive antibiotic policy in October 2006 which was widely disseminated and rigorously enforced. This policy led to an almost immediate reduction in C difficile rates back to the 2004 baseline, running between 50 and 60 cases per year. This amounted to a 70 per cent reduction in 2007 compared with 2006.

We had barely stopped congratulating ourselves when a new target, to reduce C difficile rates by 10-30 per cent per hospital each year, was introduced. It is important to understand that there is a carriage rate of C difficile of 3-5 per cent in the normal population, and that inevitably some patients treated in hospital will develop C difficile-associated diarrhoea. A more important issue is whether infection control policies prevent spread. A hospital such as ours would expect to have roughly five to 10 cases per month. There is thus a strong possibility that we will miss the target (ours is 60), despite achieving a strikingly good result for C difficile reduction the year before the target came in.

Back to the drawing board

Looking at this sort of target in a different way might help. Our rates were 0.78 cases per 1,000 bed days in 2004, 2.39 in 2005 and 3.09 in 2006. Presentation of targets in this way is helpful for two reasons. First, it allows a benchmark to be set for acceptable performance, which could be standardised across the NHS. On the evidence I have seen, it might be that a rate of less than one C difficile case per 1,000 bed days is acceptable, and that organisations with incidence lower than this should be exempt from targets. There could then be a graded percentage reduction target depending on the severity of the problem. Up to a point, this has happened. But wherever the starting point, there comes a time at which no further percentage reduction is possible. On a percentage improvement target, Stoke City is better than Manchester United.

Second, targets with absolute numbers cannot take into account changing patient populations or activity, which is likely to challenge a number of organisations as service reconfiguration accelerates. We therefore have to move to a more sophisticated method of measuring these vital healthcare-associated infection targets. We need to move away from a situation in which organisations that have independently set and met challenging targets seem to fail.

From a target perspective, we would have been better to let patients fester in their own faeces until after the target had been set. I am glad that we did not do this, and I will be delighted to "hit the point and miss the target". I recognise, however, that I do not speak for Monitor, the Department of Health or the media, who will be calling us to account for failure.