Pay-for-performance policies aim to improve population health by incentivising improvements in quality of care.

To assess the relationship between general practice performance on severe mental illness (SMI) and depression indicators under a national incentivisation scheme and suicide risk in England for the period 2006–2014.

Longitudinal spatial analysis for 32 844 small-area geographical units (lower super output areas, LSOAs), using population-structure adjusted numbers of suicide as the outcome variable. Negative binomial models were fitted to investigate the relationship between spatially estimated recorded quality of care and suicide risk at the LSOA level. Incidence rate ratios (IRRs) were adjusted for deprivation, social fragmentation, prevalence of depression and SMI as well as other 2011 Census variables.

No association was found between practice performance on the mental health indicators and suicide incidence in practice localities (IRR=1.000, 95% CI 0.998–1.002). IRRs indicated elevated suicide risks linked with area-level social fragmentation (1.030; 95% CI 1.027–1.034), deprivation (1.013, 95% CI 1.012–1.014) and rurality (1.059, 95% CI 1.027–1.092).

Primary care has an important role to play in suicide prevention, but we did not observe a link between practices' higher reported quality of care on incentivised mental health activities and lower suicide rates in the local population. It is likely that effective suicide prevention needs a more concerted, multiagency approach. Better training in suicide prevention for general practitioners is also essential. These findings pertain to the UK but have relevance to other countries considering similar programmes.

None.

Various initiatives have been tried to improve the quality of primary care in England and Wales in the last fifteen years. Such initiatives can be divided into quality improvement (QI) and quality assurance (QA).

This paper looks at three contrasting models, drawn from data from 48 semistructured interviews with personnel from three primary care organisations (PCOs): two primary care trusts in England and one Local Health Board in Wales.

The first model was collegiate, a voluntary doctor-led initiative begun during the period of GP fundholding. The second is clinical governance, a current government-imposed system administered by PCO officers, which has attracted limited engagement from GPs. The third is the Quality and Outcomes Framework of the new GP contract, which was generally described positively, although the process of administering it was experienced as bureaucratic.

The three models correspond with three organisational types: networks (which use peer relationships to achieve goals), hierarchy (which use ‘top-down’ requirements and monitoring) and market (which use contracts). Although doctors have traditionally preferred network-style arrangements, the success of these arrangements in sustained QA and QI has been questionable. The importance of hierarchical arrangements is inevitable, given the functions and constitution of PCOs, and the risk that GPs will disengage is similarly inevitable. However, it is important that PCO officers find ways to engage GPs as much as possible in quality initiatives if patient services are to improve.