What a performance: anticipating the impact of new provider league tables in the NHS

The 10-Year Health Plan for England plans to introduce quarterly, ‘easy-to-understand’ league tables, ranking NHS providers on a range of performance indicators. Providers will be categorised into segments from 1 (high performing) to 4 (failing). The stated aim is to enable patients to make sense of NHS performance and quality data, with NHS providers and integrated care boards required to routinely publish information on service quality and access using local authority boundaries. Where local providers perform well, they will be granted greater autonomy and flexibility to develop services free from central control, while struggling organisations may face ‘turnaround’ interventions.

But we have seen this film before, and the sequel “Back to the Future” looks very much like the annual star ratings regime, implemented in the NHS in England from 2000 to 2005. Although star ratings enabled the reduction in hospital waiting times (not least due to additional funding), they were eventually discontinued due to concerns about their effectiveness and limitations. Governance using published performance metrics presumes that they will incentivise individuals and organisations to improve their performance, but that any adverse consequences can be kept to some acceptably low level. However, empirical evaluation of the star rating system revealed significant evidence of damaging as well as beneficial change and identified a myriad of unintended, unwanted and ultimately dysfunctional consequences for organisations, staff and people receiving care. At least 20 dysfunctional consequences of star ratings were identified, including:

• Tunnel vision: organisations focused on achieving high star ratings over other crucial aspects of patient care, leading to the neglect of other important but unmeasured areas and distorting clinical priorities.
• Unprofessional behaviour: pressures to meet the performance targets were seen to create a demanding climate for staff, which could sometimes extend to uncomfortable levels of coercion and even bullying, intimidation, and harassment of staff in the under-performing trusts.
• Ghettoization: because of its impact on trusts’ reputation, it was reported to have had a differential impact on the abilities of trusts to attract and retain high-quality staff. In particular low low-performing trusts reported that a low star rating contributed to their problems as health professionals would be reluctant to join an organisation that had publicly been ‘named and shamed’ as failing organisations.
• Insensitivity and misinterpretation: The star ratings were too blunt a classificatory device to capture the complexity of health care performance. All trusts have uneven pockets of good and poor performance that are not adequately captured by a single measure. There were concerns that some organisations were unfairly labelled as under-performing (Type I error) or, potentially more seriously, that some organisations classified as performing to a high standard were under-performing in areas that were not included in the star rating (Type II error). This situation could be compounded if (actually) underperforming organisations were granted more freedoms from central control under the earned Autonomy policy.

While performance measures can be a valuable tool for improving healthcare, it is crucial to proactively anticipate the potential for dysfunctional consequences and to closely monitor the impact of the new performance measurement regime on all stakeholders, with policies put in place to mitigate them.