Purchasing clinically effective care

Research findings are often poorly translated into clinical practice. One example is the management of acute myocardial infarction, where the evidence of the effectiveness of aspirin and early thrombolysis is overwhelming.1,2 Despite this the proportion of patients receiving the treatment may be low.3,4 Ensuring that patients receive the best possible care should be important for all doctors.

Should purchasers care as well? The NHS Executive thinks so and believes that the issue should be addressed through contracting. Last December all fundholding general practitioners, trusts, and health authorities received a letter from the executive urging them to take clinical effectiveness and clinical guidelines into account in contracting.5 Seven guidelines were attached for consideration, with the hope that purchasers would include at least one of them in their contracts.

The NHS Executive clearly believes that clinical effectiveness should form part of the NHS's medium term objectives. Planning guidance already issued for 1995-6 has included the objective that the NHS should “invest an increasing proportion of resources in interventions known to be effective and where outcomes can be systematically monitored, and [that it should] reduce investment in interventions shown to be less effective.”6 Purchasing authorities will be expected to increase investment in at least two interventions known to be effective, to reduce investment in at least two interventions that evidence has identified as likely to be ineffective, and to increase the use of clinical outcomes and audits in contracts.

Now a further letter from the executive, issued last week, shows some softening of approach.7 The complexity of the task is acknowledged, as is the length of time needed to adapt suitable evidence based clinical guidelines for local use. This shift of emphasis is welcome because evidence of the effectiveness of clinical guidelines themselves shows that a top down approach is less likely to change behaviour than the development of guidelines by those who are to use them.8 Another new approach is the suggested involvement of primary care; family health services authorities are asked to work with medical audit advisory groups, general practice postgraduate tutors, and local practitioners in the development of local documents. Great benefits could accrue from doctors in primary and secondary care working together on clinical policy; it would be wrong to restrict all initiatives regarding clinical effectiveness to hospital providers. Lastly, the letter suggests that patients should be involved in developing guidelines.

Whether any of these initiatives will change doctors' practice - for example, increasing the chances of patients with an acute myocardial infarction receiving aspirin and thrombolysis - is unknown. Haines and Jones have advocated an approach to implementing research findings in clinical practice that incorporates work with opinion leaders, purchasers, and professional organisations; programmes of education and clinical audit; and the use of “patient specific reminders” to support clinical decision making.9 Most of these approaches have been shown to affect clinical practice, although mostly outside Britain. As systematic reviews of research evidence begin to emerge from the Cochrane Collaboration10 and the NHS Centre for Reviews and Dissemination we need to establish which methods of implementation work best in the NHS and to create an infrastructure which could enable the new material to be put to best use.

No one doubts the critical importance of clinical effectiveness, and the NHS Executive is right to make it the concern of both purchasers and providers, but the use of contracting to change clinical practice will need evaluation (just as any other intervention requires evaluation). Stipulating that purchasing authorities should divert investment towards effective interventions and away from ineffective ones has a mechanistic feel to it. Purchasers need to have a more interactive role than this: they need to establish dialogue with local hospital doctors, general practitioners, and patients. In addition, hospital doctors need to talk to each other about policy and practice, and purchasers should insist that they do so.

At this stage the role of the contracting mechanism should perhaps be to tie providers to this dialogue; to ensure that clinicians address issues of clinical policy and practice with their colleagues (including the local adaptation of evidence based clinical guidelines); and to enable local users of health services to have an informed voice. That would give all parties sufficient freedom for local collaboration, while ensuring the commitment of providers.