How to design quality incentives Quality measures are hospital medicine’s “new business case”—and a growing part of physician compensation by Phyllis Maguire
Published in the January 2008 issue of Today's Hospitalist
It used to be that rewarding hospitalists for their performance on quality measures was the mark of a progressive program. But with many hospital medicine groups employing some sort of quality incentive as part of their compensation plan—41%, according to recent data—quality measures are now commonly being used to justify the existence of hospitalist programs.
According to Winthrop F. Whitcomb, MD, director of performance improvement at Mercy Medical Center in Springfield, Mass., hospitals and hospitalist programs increasingly view quality incentives as “the new business case” for hospitalists. As a result, there is more
“Feedback may be more powerful than the check at the end of the year.”
–Winthrop F. Whitcomb, MD Mercy Medical Center
pressure than ever for hospitalist groups to implement some kind of quality measurement and incentive program.
When it comes to creating incentives, however, groups need to use some basic rules of thumb to make sure that they are improving not only physician compensation, but patient care. Here are some tips that Dr. Whitcomb offered at the University of California, San Francisco’s hospital medicine conference held last fall.
All or nothing?
One principle to follow when choosing quality incentives: Use something that is already being measured. “That way, you don’t have to go back and do chart review,” Dr. Whitcomb pointed out.
You also have to choose a particular philosophical approach: Should the group incentivize excellence, rewarding the best possible performance? Or should you reward improvement made, even if that falls short of superlative? Or a combination of the two?
According to Dr. Whitcomb, rewarding a combination is the best way to go. You should incentivize excellence if, for instance, you want the group to continue to maintain a 100% level of performance on a measure like prescribing aspirin at discharge for heart attack patients.
But you should also incentivize improvement on measures where physicians currently perform poorly. How much of an improvement should physicians make? Dr. Whitcomb suggested picking targets based on prior performance at your hospital, or checking Hospital Compare data to find benchmark targets for a measure for your state.
Once you set benchmark targets for improvement, groups have another decision to make: Should your program take an all-or-nothing approach to paying out on quality incentives?
“If you miss the aspirin on discharge in MI target by one percentage point, should you miss the entire payout?” he asked. That type of hard-line stance, he said, isn’t really helpful.
Instead, he said, quality incentive plans should recognize gradations of performance. He pointed out that he was in the process of “not making his numbers” on a quality incentive for the second time in his career, and “it doesn’t feel good when you have done good work in improvement and don’t realize some kind of reward.”
He prefers incentive programs that set intermediate goals where physicians would get, for instance, half of the payout.
Individual vs. group payouts
In considering payouts, another key question is whether to pay the group as a whole or pay individual physicians. For many measures, Dr.Whitcomb said, it makes sense to split payouts equally among group members.
But if you can track performance by individual physicians, he added, do so. “You might be able to do that for something like medication reconciliation,” Dr. Whitcomb pointed out.
Because money is important, bonuses must be paid out both fairly and frequently, and incentives need to be paid out at regular intervals to affect behavior. “We do it every six months and make it large enough to be meaningful,” Dr. Whitcomb said.
Another consideration that’s almost as important as the money: feedback. When physicians get only a check without being told how they did, Dr. Whitcomb said, they never marry the reward to the desired behavior change.
“Data is an incentive,” said Dr. Whitcomb. “Feedback may be more powerful than the check at the end of the year.”
Evolution of an incentive program
To show how well quality incentives can work, Dr. Whitcomb explained how his hospital went about creating quality incentives in early 2004.
The hospital appropriated $100,000 to pay out to the 10-physician group, with bonuses to be paid equally to all every six months. That first bonus pool was pegged to three measures and targets: a 45% target for pneumovax screening and/or administration in pneumonia patients; an 85% target for prescribing an ACE inhibitor to heart failure patients; and an 85% target for left ventricular function assessment, also in heart failure patients.
After a nurse reviewed all charts for heart failure and pneumonia patients, Dr. Whitcomb said, the program director met with hospitalists who posted “misses” on the quality measures. All the hospitalists received regular reports on their performance.
Dr. Whitcomb described the impact of the initiative as “dramatic.” Pneumovax rates jumped from 54.7% before the incentives to 92.8% afterward. ACE inhibitor prescribing rose from 84.0% to 97.6%, and left ventricular function assessment went from 92.2% to 99.5%.
Looking back, Dr. Whitcomb said that starting with only three indicators was the right number to get hospitalists used to being measured and having feedback. He also said that nearly three years later, the hospital’s quality incentive program has moved on to other measures.
Half of the 2007 quality bonus, for example, will be paid to physicians who achieve medication reconciliation for 90% of their patients. The other half is a composite of six Medicare core measures for pneumonia, heart failure and heart attack, which physicians must meet for 90% of their patients.
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
A formula for productivity-based incentives
Along with quality measures, a growing number of groups are tying some physician income to another type of incentive: productivity.
Physicians may worry that productivity-based incentives will reward workaholic behavior and lead to burnout, an issue that hospitalists have been combating for years. But if these bonuses are designed properly, physicians have relatively little to worry about.
According to Winthrop F. Whitcomb, MD, well-designed productivity incentives not only reward productivity, but do so in a transparent way that everyone in a practice can understand.
A sample plan
In creating an incentive package, Dr. Whitcomb suggested identifying a base salary for a full-time physician. He gave the example of a base salary of $150,000 for a full-time equivalent working 2,000 hours per year.
The group then needs to set a productivity threshold above which a bonus will kick in. If the threshold is set at 3,000 work RVUs and the value for each work RVU above that threshold is $20, a physician who works 4,000 work RVUs per year would receive a $20,000 bonus. That RVU threshold should be adjusted up or down, depending on whether a physician works more or less than a full-time equivalent.
Finally, the bonus amount needs to be large enough to get physicians’ attention--at least 5% of base pay—and paid out regularly to effectively change physician behavior. According to Dr. Whitcomb, a quarterly payment would suffice.
Adjusting for night shifts
How do you handle perennial low producers, like physicians who work nights? If all the physicians in the group take an equal number of night shifts, then low productivity at night among group members becomes a wash.
But if you have group members who prefer doing more than their fair share of nights, Dr. Whitcomb said, you should modify compensation for those doctors, who he said “are throwing themselves on their sword for the group.” The simplest solution, he added, may be to adjust the threshold for payout downward while at the same time paying a higher base rate.
At Mercy Medical Center in Springfield, Mass., where he practices, the payout threshold for nocturnists is 75% of that of the day hospitalists, based on an analysis of prior years’ total work RVU production. (The nocturnists were found to be 75% as productive as the day hospitalists.) The nocturnists also receive a base salary that is 20% higher than their colleagues’.
Physicians worry that because they can’t control patient volume, their productivity bonuses—and their income—are largely beyond their control.
Dr. Whitcomb said that groups should regularly be adjusting physician staffing to control for average patient volume. Even so, he explained, productivity incentives encourage doctors to work a little harder, helping them meet volume peaks and to seek out more consults or admissions when volume is a little slow.
Physicians also worry that productivity bonuses will incentivize longer lengths of stay, a concern that Dr. Whitcomb said is unfounded. “With good case management and a hospitalist ethic, length of stay is not affected,” he said.
And physicians fear that productivity incentives will tie them to a hamster wheel of too much work, leading to less job satisfaction and burnout. Yet studies have shown that linking a portion of physician compensation to incentives does not contribute to burnout, he said. Instead, having a stake in the group’s financial success creates a greater sense of ownership , which in turn leads to greater job satisfaction.
The top quality incentives used by hospitalist programs
According to recent data from the Society of Hospital Medicine, about 41% of hospitalists groups in 2006 participated in quality incentive programs. Here’s a look at the most common quality incentive measures used by those groups: