• Sheena Asthana, Alex Gibson, Paul Hewson, Trevor Bailey, and Chris Dibben.
• Faculty of Health, 1School of Computing and Mathematics, University of Plymouth, Plymouth, UK.
• J Health Serv Res Policy. 2011 Apr 1;16(2):95-101.

ObjectivesTo contribute to current policy debates regarding the devolution of commissioning responsibilities to locally-based consortia of general practices in England by assessing the potential magnitude and significance of budgetary risk for commissioning units of different sizes.MethodsPredictive distributions of practice-level mental health care resource needs (used by the Department of Health to set 'fair-share' practice budgets) are aggregated to a range of hypothetical, but spatially-contiguous, consortia serving populations of up to 400,000 patients. The resulting joint distributions describe the extent to which the legitimate mental health needs of consortia populations are likely to vary. Budgetary risk is calculated as the likelihood that a consortia's resource needs will, in any given year, exceed its allocation (taken as the mean of its predictive distribution) by more than 1%, 3%, 5% or 10%. The relationship between population size and budgetary risk is then explored.ResultsIf between 500 and 600 consortia are created in England (serving 87,000 to 104,000 patients) then, in order to meet the legitimate mental health needs of their patients, each year around 15 to 26 consortia will overspend by at least 5%, and one or two by at least 10%. The budgetary risk faced by consortia serving smaller/larger populations can be read off the graphs provided.ConclusionsUnless steps are taken to mitigate budgetary risk, the devolution of decision-making and introduction of fixed budgets is likely to result in significant financial instability. It will be difficult to reconcile the policy objectives of devolved commissioning, best met through relatively small and fully accountable consortia, with the need for financial stability, which is best met by pooling risk across larger populations.

19 keywords...

hide…