Gaming New Zealand's Emergency Department Target: How and Why Did It Vary Over Time and Between Organisations?

BACKGROUND: Gaming is a potentially dysfunctional consequence of performance measurement and management systems in the health sector and more generally. In 2009, the New Zealand government initiated a Shorter Stays in Emergency Department (SSED) target in which 95% of patients would be admitted, discharged or transferred from an emergency department (ED) within 6 hours. The implementation of similar targets in England led to well-documented practices of gaming. Our research into ED target implementation sought to answer how and why gaming varies over time and between organisations. METHODS: We developed a mixed-methods approach. Four organisation case study sites were selected. ED lengths of stay (ED LOS) were collected over a 6-year period (2007-2012) from all sites and indicators of target gaming were developed. Two rounds of surveys with managers and clinicians were conducted. Interviews (n=68) were conducted with clinicians and managers in EDs and the wider hospital in two phases across all sites. The interview data was used to develop explanations of the patterns of variation across time and across sites detected in the ED LOS data. RESULTS: Our research established that gaming behaviour - in the form of 'clock-stopping' and decanting patients to short-stay units (SSUs) or observation beds to avoid target breaches - was common across all 4 case study sites. The opportunity to game was due to the absence of independent verification of ED LOS data. Gaming increased significantly over time (2009-2012) as the means to game became more available, usually through the addition or expansion of short-stay facilities attached to EDs. Gaming varied between sites, but those with the highest levels of gaming differed substantially in terms of organisational dynamics and motives. In each case, however, high levels of gaming could be attributed to the strategies of senior management more than to the individual motivations of frontline staff. CONCLUSION: Gaming of New Zealand's ED target increased after the real benefits (in terms of process improvement) of the target were achieved. Gaming of ED targets could be minimised by eliminating opportunities to game through independent verification, or by monitoring and limiting the means and motivations to game.

Measuring and managing health system performance: An update from New Zealand.

In July 2016, New Zealand introduced a new approach to measuring and monitoring health system performance. This 'Systems Level Measure Framework' (SLMF) has evolved from the Integrated Performance and Incentive Framework (IPIF) previously reported in this journal. The SLMF is designed to stimulate a 'whole of system' approach that requires inter-organisational collaboration. Local 'Alliances' between government and non-government health sector organisations are responsible for planning and achieving improved health system outcomes such as reducing ambulatory sensitive hospitalisation for young children, and reducing acute hospital bed days. It marks a shift from the previous regime of output and process targets, and from a pay-for-performance approach to primary care. Some elements of the earlier IPIF proposal, such as general practice quality measures, and tiered levels of performance, were not included in the SLM framework. The focus on health system outcomes demonstrates policy commitment to effective integration of health services. However, there remain considerable challenges to successful implementation. An outcomes framework makes it challenging to attribute changes in outcomes to organisational and collaborative strategies. At the local level, the strength and functioning of collaborative relationships between organisations vary considerably. The extent and pace of change may also be constrained by existing funding arrangements in the health system.