Improving productivity in the NHS

Since 2008, economic prospects for virtually all countries have worsened. In England, the economic crisis came after a decade of investment in the NHS that nearly doubled public expenditure for improving infrastructure and capacity to deliver services. That spending trajectory is on a collision course with the new pressures on public spending; estimates of the shortfall in NHS funding by 2014 approach £20bn (€24bn, $30bn), roughly 20% of the current budget.1

In a report published today by the King’s Fund, Appleby and colleagues refine the deficit projections and note some mitigating factors, such as a commitment from the new coalition government to real term increases in NHS funding.2 Nonetheless, they estimate that productivity will need to improve by 3-4% each year until at least 2014 to avoid reductions in service that could threaten the quality of care. The report concludes that productivity could be improved by focusing on clinical decision making in order to reduce variation in clinical practice.

Appleby and colleagues are not alone in targeting such variation to achieve better delivery of health services. During the heated American debate on healthcare reform, President Barack Obama cited an evocative story about the ethical and economic implications of practice variation and declared, “This is the problem we have to fix”.3 Per person healthcare costs vary by as much as 300% across regions in the United States. Rates of major surgeries vary by up to fivefold or more and consultations, diagnostic testing, and use of hospital services are more influenced by local capacity and habit than clinical need. The weight of evidence indicates that care is no better (and may be worse) in the high cost regions, suggesting that 30% of the US expenditure on healthcare is wasted.4 5

Clinical practice variation has been recognised since a tenfold variation in tonsillectomy rates was reported in England and Wales in 1938.6 Decades later regional variations in rates of surgical and medical interventions were described in many countries, despite their different approaches to the organisation and funding of health care. Responses have often been linked to concerns about rising costs. Guidelines have been developed by professional societies and government funded organisations. Efforts have been made to engage patients with shared decision making and information resources such as NHS Choices. And there have been attempts to combine fiscal and clinical accountability through commissioning and GP fundholding, as emphasised in health secretary Andrew Lansley’s recently announced plans.1

All such measures to tackle practice variation have proved inadequate. The lessons from these failures relate to the complexity of clinical decision making and the need to improve the quality of decisions. They are also about the power relationships among clinicians, patients, and the public, and those who pay for health services, including government.7

Reducing variation can be the key to doing things consistently right. But deciding what is the right thing to do is more complex because of uncertainty about the health outcomes that will follow an intervention and the variable assessments that different patients make of the same outcomes. Evidence and professional knowledge about the likelihood of different benefits and harms can contribute to better decisions but it is not enough. Patients’ concerns about what matters to them are also relevant when there are trade-offs to be made.

If all variation were bad, solutions would be easy. The difficulty is in reducing the bad variation, which reflects the limits of professional knowledge and failures in its application, while preserving the good variation that makes care patient centred. When we fail, we provide services to patients who don’t need or wouldn’t choose them while we withhold the same services from people who do or would, generally making far more costly errors of overuse than of underuse.8

It is right to tackle variation to improve productivity with better clinical decision making, but it will take political skill and resolve to do so. A review 20 years ago recognised the failures that underlie practice variation and offered an incisive political analysis for the lack of an effective response.9 The review noted that clinicians cite the scientific basis of medicine as their source of authority in clinical decision making, just as they cite the art of medicine in defending their autonomy. They believe that clinical freedom is crucial to quality healthcare. Patients generally agree with clinicians; they want to believe that science provides the answer to their problems and want to trust their doctors. And they are too ready to defer and avoid responsibility for decisions that can lead to bad outcomes. In deference to that professional-patient alliance, policy makers who are responsible for budget allocations stand down from directly constraining clinical decisions and instead resort to negotiating fees, controlling prices, and, when possible, setting global budgets. The coalition government has promised even more clinical freedom in a liberated NHS even as it demands more accountability. Clinicians, policy makers, and patients all have their reasons to ignore practice variation.

Appleby and colleagues appreciate the difficulty of improving productivity, and note that success will require coordinated activity at every level from government to clinical microsystems. The economic crisis with its projected effect on the NHS may be what is needed to motivate effective solutions to the inefficiency of practice variation. But questions persist. Will policy makers risk controversy by educating the public about excesses of clinical freedom and the collective responsibility to tackle the ethical and economic implications of practice variation? Will health professionals be willing to temper their defence of autonomy and accept accountability for better clinical decision making? And will the public and patients engage more fully in healthcare decisions for the sake of better health at lower cost? With the stakes so high, it would be best if these stakeholders united to confront practice variation.