Using financial incentives to improve health system performance Anthony Scott Melbourne Institute of Applied Economic and Social Research The University of Melbourne Funding is acknowledged from an ARC Future Fellowship
“That any sane nation, having observed that you could provide for the supply of bread by giving bakers a pecuniary interest in baking for you, should go on to give a surgeon a pecuniary interest in cutting off your leg, is enough to make one despair of political humanity.” (George Bernard Shaw, The Doctor’s Dilemma, 1911)
Hammurabi, King of Babylon, 2,300BC Babylonian Medical Fee Schedule Operations that saved a life10 shekels of silver (nobleman) 5 shekels of silver (poor man) 2 shekels of silver (slave) Setting a fractureFreeman – 5 shekels Son of a noble – 3 shekels Slave – 2 shekels Loss of a slave’s eyePay half the slave’s value Kill a slave during a major operationProvide a new slave Kill a nobleman during a major operation Hands cut off
Background National Health Reform –Increased emphasis on performance measurement and public reporting of performance –Changes to payment systems (ABF and Co-ordinated Care for Diabetes Pilot) Increasing use of pay for performance in other countries –Quality and Outcomes Framework (UK) –CMS Premier Quality Initiative (US) –Primary Care Medical Home (US) Need to re-aligning funding arrangements to meet health system objectives
Background Changing the level and method by which health care providers are paid has the potential to address: –health workforce shortages –the mal-distribution of health professionals across specialties, sectors, and geographic areas –improve the quality and costs of health care provided.
Salaries, sessional payments Inputs (time) Devolved budgets, hard/soft budget constraints, capitation, ABF Costs Fee-for-service (incl. ABF), capitation Volume (number of services or patients) Pay for performance (P4P) Health outcomes / clinical quality indicators PerformanceType of payment
Does the amount/level of pay matter? Evidence of effects on: –hours worked for doctors and nurses –specialty choice –workforce participation Backward bending labour supply How are pay levels set? –Flexibility of pay and EBAs Incentives in salary scales, career structures and subjective performance evaluation
Do different methods of payment matter? FFS or salary or capitation? Pay for performance –Primary care Scott et al, 2011 Eccles et al, 2011 –Hospitals Scott and Ouakrim, 2011 Eccles et al, 2011 Quality of evidence
Unresolved questions Doubts about the use of financial incentives to change health care providers’ behaviour –Quality of the evidence –Poorly designed incentive schemes Political Assumes providers are largely motivated by money Potential unintended and undesirable consequences (‘gaming’, ‘multi-tasking’)
What’s happening in Australia?
Activity-based funding Australian Government State and Territory Health Departments Local Hospital Networks National Funding Pool Independent Hospital Pricing Authority National Health Performance Authority ABF: Fixed price per DRG Service Agreement Non-ABF payments (block funding) Efficient price
Incentives in activity-based funding
Will it work? Depends on: –What happens now in each State/Territory –Hard or soft budgets? –Level of fixed payment from funding pool and % of hospitals with costs above or below the price –Performance assessment framework – where are the teeth? –Percentage of hospitals which continue to be ‘block funded’ (eg in rural areas) –Special pleading (IHPA takes submissions - lobbying) –All of the above will vary across States/Territories and so behavioural effects will be different across States/Territories Role for P4P?
Co-ordinated care for diabetes 2 year pilot starts in 2012 Three elements –Voluntary patient enrollment –Flexible payment per patient –Pay for performance Design of scheme (including level and type of payments) not pre-specified
New payments in CCDP
Factors influencing success To encourage practices to participate and enroll patients, changes in expected revenue must be greater than expected costs Three key elements of payment design –Paying for improvements in quality –Avoiding cream skimming – risk adjustment –Avoiding exception reporting
2. Avoiding cream skimming
3. Avoiding exception reporting QOF, for diabetes, –a median of 5.4% (0-40%) of practices exception reported patients –median gains of between £1,700 and £15,000 per practice Solution – only pay for the numerator –Payment for each patient who achieves target
Other issues Who receives the payment, and what is it used for? A single funder strengthens the effect of incentives A stable/enrolled population strengthens the effect of incentives
Summary No magic bullets –complex interventions recognising that money isn’t the only, or main, motivation Both the quality of evidence and design of schemes are poor Effects of ABF depend on States Effects of CCDP depend on payment design
Issues for research Policy design (as well as policy evaluation) Careful thought about behavioural effects Qualitative as well as quantitative