The Maynard Doctrine: the trouble with incentives

Alan Maynard is professor of health economics, University of York

Monday 21 July 2008

The trouble with incentives is that they work, but they may produce changes in behaviour that are at once welcome and perverse.

Take, for instance, the fines announced in December 2007 for NHS trusts that fail to hit their C.Difficile targets. These fines are potentially large, up to £3.5 million for unsuccessful trusts. Faced by such fines, managers are striving very hard to improve the performance of their hospitals - just as Ministers hoped they would.

However deviance in the use of antibiotics and other contributory factors can make managers realise that their heads may be on the block with for failing to hit targets. At this stage the gaming or “strategic behaviour” begins.

Patients who present with C. Diff early in their hospital stay are more likely to be victims of poor primary care than ill due to hospital failures. Consequently, patients presenting with the condition in the first 48 hours of their hospital stay may not be counted in the targets.

Then there is the nice issue of defining the effects of C.Diff, particularly in terms of the consistency of the stool of the patients! A liquid stool is an obvious symptom of the problem but there are degrees of interpretation, known in the trade as “when is a stool not a stool”?

Policing costs of transactional rewards
With large sums of money at stake, the incentive to ‘game’ increases and the policing costs of the incentive system increase considerably. This will obviously be a problem with some of the Darzi proposals. For instance what will be the gaming effects of setting payment by results (PbR) tariffs more aggressively and equal to the costs of the best quartile of providers? What will be the effects of incentivising PROMs?

The measurement of patient outcomes is long overdue. Bruce Keogh is determined to publish the mortality rates of clinical teams. His initial focus is surgery, which is easier to performance-manage than medicine. But he intends to also get better performance data for medicine by, for instance, using the renal register for nephrology performance measurement.

A stroll along the PROM?
The Government is introducing patient-reported outcome measurement (PROMs) from April 2009. They plan to incentivise implementation in 2009-10 by putting at risk the part of the tariff uplift for those trusts that fail implement the system for hip and knee replacements, hernia repairs and varicose veins.

This is an ambitious agenda. The use of PROMs by BUPA and in the NHS-ISTC pilot focused on a “creamed” population of largely articulate and motivated patients. Using PROMs on an average NHS population will raise nice management issues, such as how to get elderly patients with complex co-morbidities to complete the quality of life returns; and how to interpret the data when it is produced -  when patients are shown to be worse off after a procedure, how will this be managed within the hospital? And how will the PCT use the information to performance-manage the trust?

In addition, there are the problems of driving the response rates up to over 70% to get meaningful data and the issue of whether to sample to reduce collection costs, rather than assess the whole population. Of course all these problems are going to be “solved” in imminent Departmental guidance to the NHS and by tendering for the supply of PROMs from companies who have little experience of using such instruments on general NHS populations!

The Government’s response to these cautions is to say, 'we should have done this 10 years ago'. Amen! But given that you didn’t, should you not be careful now to ensure that you do not throw the baby out with the bathwater?

'Whilst Government is not seeking structural “redisorganisation”, its incentives policies are likely to cause even more profound change.'

The case for piloting the incentivisation of PROMs and other incentives that are now flavour of the month in Whitehall is obvious. However it is being ignored by a Government frustrated by the failure of its policies to reduce clinical practice variations and get public attention focused on outcomes - i.e. making patients better.

Whilst Government is not seeking structural “redisorganisation”, its incentives policies are likely to cause even more profound change. Let’s hope these changes benefit the patient and the taxpayer rather than produce short-term and optimistic newspaper headlines!