We share Marshall and Harrison’s caution on over-reliance on financial incentives for improving quality of care manifested in the new UK general practice contract.¹ Financial incentives do not always result in behaviour change as intended for reasons that are not well understood. We argue that “the fascination with financial incentives” is not “based on sound empirical evidence”.

In recent years at least three systematic reviews on the impact of financial incentives on provider behaviour have been published.^2–4 The reviews agree that good quality evidence is lacking and that the available research evidence provides mixed messages.^3,4 For example, a controlled before/after study of 426 Danish GPs assessed the effects of adding fee-for-service to capitation on the use of repeat prescription.⁵ Contrary to expectations, fee-for-service payments were associated with a fall in repeat prescription rates. Some suggested the findings either implied that GPs did not respond to financial incentives or the fee was not sufficient to change behaviour.⁶ Neither could justify the significant reduction that followed the fees.

We subjected the systematic reviews to forward citation searches in the Science and Social Science Citation Index databases (accessed via http://wok.mimas.ac.uk, February 2005). As a result, we identified only two research studies that could have met the Cochrane Effective Practice and Organisation of Care Review Group criteria for acceptable evidence in provider behaviour change.^7,8 The studies did not support the argument that financial incentives necessarily result in intended behaviour changes. Firstly, a US randomised controlled trial assessed the impact of financial incentives on provision of smoking cessation services. It concluded that financial incentives alone did not result in adherence to the performance targets linked to clinical guideline recommendations.⁷ This was in line with a previous uncontrolled UK study that concluded fee-for-service payments were ineffective in increasing provision of smoking cessation advice.⁹ Secondly, a controlled before/after study assessed the effects of changing GP payment from capitation plus fee-for-service to salaried arrangements. No significant change in GP behaviour occurred.⁸ Others also assessed the impacts of salaried arrangements on quality of care in the UK.¹⁰ They found evidence of effect on quality of care, but this was inconsistent and confounded.

Research evidence does not support policymakers’ enthusiasm for using financial incentives for quality improvement while, as Marshall and Harrison point out, there are dangers of unintended negative effects.¹