The vast majority of entities that have implemented pay-for-performance initiatives have chosen to focus almost exclusively on improving the quality of care rather than on reducing its cost. While it is obviously important to address quality, why should this be at the expense of addressing the equally important issue of cost? Why shouldn't pay-for-performance initiatives do both?
One obvious area where pay-for-performance can be used to contain health care costs involves service or length-of-stay-related increases in hospital costs. Physician practice patterns have a significant effect on such costs.
Payers can establish an effective, efficiency-based pay-for-performance system by using case-mix-specific cost-per-stay benchmarks and then providing incentive payments above and beyond traditional fee-for-service payments to physicians who are able to push costs-per-stay below benchmark levels.
Because pay to physicians is typically only a tenth of the total cost of a hospital stay, reductions in hospital costs can easily fund generous physician incentive payments, while still significantly reducing overall payer costs. This is a true win/win scenario.
|Example of incentive to physicians|
|Average hospital daily payment||$2,000|
|Average physician daily fee-for-service payment||$200|
|Health plan savings associated with one day reduction in length-of-stay||($2,000 + $200) = $2,200|
|Increase in physician payments if half of savings are shared with physician||[($2,000 + $200) / 2] − $200 = $900|
|Savings retained by the health plan||[($2,000 + $200) / 2] = $1,100|
For such a system to work, incentive payments should be based solely on cost savings that are under the control of physicians (e.g., reductions in unnecessary services or lengths of stay) and not on cost savings that are beyond the control of physicians (e.g., changes in hospital prices, changes in patient case mix, or random variations in hospital costs). Proper application of case-mix measurement tools and statistical significance tests are crucial to identifying such cost savings.
Incentive payments also should be based on net payer savings (i.e., cost decreases offset by cost increases) in order to ensure that the pay-for-performance plan at least does no harm to the budget if it is not successful in reducing the payer's overall hospital costs.
It is also critically important that physicians be educated on how to calculate incentive payments and that they be given continuous feedback on progress toward reducing hospital costs and earning incentive payments.
Finally, it is important that the pay-for-performance plan address potentially negative effects on the quality of patient care (e.g., premature patient discharging).
Such an efficiency-based pay-for-performance system can be implemented easily with existing software applied to the insurer's historic and current physician and hospital claims databases. A mere two percent reduction in hospital costs would earn individual physicians thousands of dollars in additional annual income while collectively reducing a payer's annual hospital expenses by millions.