Feature

From here to perversity?

The new NHS is founded on the government’s belief in the efficiency of the market. Steve Harrison explains why too much ‘economic rationality’ may be bad for your health

Reorganisations of the NHS and local government nearly always take place on 1 April and so it is a cliché to ask whether they are a joke. But april fools’ Day 1991 saw the introduction of so-called internal markets in health services and with them the possibility of an Alice in Wonderland world of perverse incentives. A perverse incentive occurs where the design of a system or organisation incorporates incentives for actors to behave in ways which the system’s designers do not want: a sort of shooting oneself in the organisational foot.

It is quite likely that all organisations have some such incentives. The pre-Working for Patients NHS, for instance, makes it rational to discharge hospital patients with minimal supplies of medication. This is because the hospital’s budget is cash limited, while family practitioner committee reimbursements to high street pharmacists are not. Yet dispensing costs in the community are higher than in hospitals, hence the label ‘perverse’.

Such incentives occur in the social care sector too; it is not rational for a commercial residential home to try to rehabilitate a resident and thereby lose a customer.

But even if all organisations are likely to offer some perverse incentives, some are more likely to offer more. Since such incentives largely spring from the erection of organisational barriers and the use of quantitative data to govern transactions between the resulting sections and to measure their performance, perverse incentives are likely to multiply along with such practices.

The post-Working for Patients NHS will operate along these lines, and a whole new range of perverse incentives will come into force. Lack of space prevents a full account being given here, but it can be helpful to use the following three categories when thinking about them.

First, shunting is the dumping of workloand and/or expenditure on another department, institution or agency. An obvious example would be the possibility of premature discharge from hospital in order to minimise costs, thuse ‘shunting’ the care and expense of the patient on to the separately funded community service budget.

A fundholding GP would find it more rational to suggest that a patient attended the accident and emergency department, which would not be a charge on the GP funds, than to refer the patient to the outpatient department, which would be.

Second, gaming is the deliberate manipulation of information, and the concepts upon which the information is based, in the interests of securing better financial or other quantitative performance. Thus is a hospital charges a purchasing authority per ‘case’, and a case is defined as at present as a death or discharge, premature discharge and a subsequent readmission will double the income from a patient. Even when more sophisticated casemix measures such as diagnosis related groups are employed, ‘diagnostic inflation’ (the ‘discovery’ of complications, for instance) can occur. In the US this is known as the ‘DRG creep’.

Third, creaming is the selection of work which is financially advantageous to the provider. This may take the form of a GP seeking healthy patients for whom a capitation fee can be obtained in return for little work. Its corollary, ‘adverse selection’ is that the most needy patients may have the greatest difficulty in obtaining service. Services may also be created, leaving the ones that do not pay unprovided; a GP with little chance of reaching vaccination targets would be irrational to try.

I have, of course, used a very narrow idea of ‘rationality’ (to mean economic rationality) in this article. Fortunately, people are not like this. To take some of the examples used above, residential homes may try to rehabilitate patients, GPs may accept all comers, and so on. But what is sauce for the goose is sauce for the argumentative gander; if people are not economically rational why should the competative paraphernalia of Working for Patients have its supposed effects?