- Joined
- Nov 13, 2006
- Messages
- 47
- Reaction score
- 0
From: General Surgery News
ISSUE: OCTOBER, 2006 | VOLUME: 33:10
The Ups and Downs of Pay for Performance
Ready or Not, Here It Comes
Ted Bosworth
Los AngelesIn a recent symposium held during Digestive Disease Week, there was no shortage of disagreement on the issue of pay for performance. And while potential problems were envisioned for essentially every method proposed for using financial incentives to induce physicians to improve quality of care, participants also acknowledged that avoidable complications are expensive and occur often.
There is an important study just published that suggests that in all areas of medicine, we get it right only about 55% of the time. It is amazing how consistent this was across the different conditions that were evaluated, observed Lee N. Newcomer, MD, chief medical officer, United Healthcare, in Chicago. Referring to a study conducted by the Rand Corporation that looked at quality of care for 30 chronic and acute diseases in 12 communities in the United States (Asch SM et al. N Engl J Med 2006;354:1147-1156), Dr. Newcomer suggested we clearly have a lot of room for improvement.
One of the problems of pay for performance, which describes a system in which physicians are compensated for following specific protocols that are based on best practice, is that the proportion of medical conditions for which there are well-established and accepted guidelines remains small. According to George Diamond, MD, Cedars-Sinai Medical Center in Los Angeles, only about 15% of clinical interventions fall into areas of practice where such guidelines might be appropriate, and these represent an even smaller amount of total revenues.
Pay for performance only applies to what can be measured, and a lot of what we do cannot be measured. It also only tends to deal with one disease at a time when patients often have multiple medical problems, Dr. Diamond said. Moreover, he expressed concern that pay for performance places too great a focus on processes and not enough focus on direct evidence that outcome has been improved. In a complicated patient who does not fit the guidelines, pay for performance could be a potential negative incentive for optimal management.
Measuring quality is the stumbling block for any methodology to change physician behavior, including nonfinancial methods that involve peer review. At the University of Utah, David J. Bjorkman, MD, dean of the School of Medicine, said the effort to enjoin physicians to adhere to protocols is much like herding cats. He reported that the first step is to enlist physicians to buy into the legitimacy of the measures of quality of care. Physicians are not likely to be impressed with imposed processes if they are not convinced that they yield better outcomes. However, the University of Utah is developing a large data pool which it hopes will provide definitive evidence for the approaches that yield the best outcome. The data should help, but physicians will need to believe them.
We need a culture change, said Dr. Bjorkman, a past president of the American Society for Gastrointestinal Endoscopy. Referring to the willingness of physicians to rapidly accept and adopt specific protocols when proven effective, he cautioned that although we want innovation, we can no longer reward independence. He warned that bureaucrats who monitor healthcare reimbursement have a strong interest in establishing cost-effective care and emphasized that if we do not set the standards, then someone else will.
At United Healthcare, a variety of initiatives are being pursued to encourage physicians participating in their system to deliver better care. One example provided by Dr. Newcomer was in colorectal cancer resections. Despite well-publicized standards that at least 12 lymph nodes should be excised for pathologic evaluation in these procedures, the U.S. average has remained unchanged over several years at a substandard nine lymph nodes. Citing 2004 data, Dr. Newcomer reported that 54% of colorectal cancer resections are incorrect on this basis. At United Healthcare, physicians who want to be on a referral list must now present five case reports in which they demonstrate that the procedure was done correctly with 12 lymph nodes excised.
We have received very strong support from surgeons, reported Dr. Newcomer, who noted that the American College of Surgeons has endorsed removal of 12 lymph nodes as a standard of care in colorectal cancer resections. He said that gastroenterologists in their system are now provided with a $75 incentive to call United Healthcare to report a new diagnosis of colorectal cancer so that a nurse can contact the patient, describe the importance of standard-of-care therapy, and encourage them to use a surgeon on their referral list.
Differentiating this approach from pay for performance, which he believes has limited applications, Dr. Newcomer says that more fees will be awarded for specific quality actions. Of the $75 fee paid to gastroenterologists who make a call to United Healthcare to report a new colorectal cancer diagnosis, he said if you have to make an extra effort to help us solve a problem with quality of care, we will pay for that. It is only fair. He also said there are many efforts to make life easier for those physicians who are trying to provide both high quality and cost-effective care. This not only includes placing them first on referral lists and providing higher financial rewards but looking for other forms of preferential treatment. He acknowledged that we are at absolute infancy in learning how to do this well, but we would rather start with something than not move in this direction.
The major problem with fee for service, Dr. Diamond said, is that paying more for more services encourages overutilization. The problem with managed care is that it can provide incentives to do less, particularly when coding is insufficient to compensate for a complex clinical case. A potential problem for pay for performance is that rewards are often provided so long after the activity that they benefit those already delivering good care but do little to change physician behavior. One novel approach first proposed by Dr. Diamond in 1993 (Diamond G et al. J Am Coll Cardiol 1993;22:343-352) is a fee-for-benefit program in which larger fees are paid for procedures that can be expected to yield larger benefits. This encourages physicians to perform those services most likely to improve outcome and avoid procedures with little likely benefit.
On computer simulation, the costs go down, the quality of care goes up, and physician reimbursement goes up, Dr. Diamond said. He said that the benefit from such an approach accrues to the payer, the provider and the patient. Although this approach has not come much closer to application than when it was proposed more than 10 years ago, he maintained that the principles remain valid.
One approach to improving quality of care favored by some third party payers, such as the U.S. Centers for Medicare & Medicaid Services (CMS), is simply to make data on quality of care available so that consumers can make their own choices. However, there are a number of problems with this approach. One is that there is no evidence that these data influence patient choices. Another is that many patients may have trouble interpreting what the data signify. However, the greatest concern is that physicians will feel compelled to avoid sick patients to reduce the risk of bad outcomes.
There is always the danger of gaming the system, warned Dr. Bjorkman. He questioned whether it can be concluded that a hospital discharging 91% of postmyocardial infarction patients on an ACE inhibitor is really doing a better job than one with 85% on this drug at discharge. We have seen misrepresentation of data in our area in advertising. There are a lot of problems with this approach.
There is agreement that data are the key to improving healthcare no matter what approach is imposed. In his review of initiatives to improve quality of care at the second-largest independent gastrointestinal practice in the United States (50 gastroenterologists and 450 employees), John Allen, MD, reported that one of the first steps was to enlist all the physicians to agree to be transparent in their practices and assist in collecting data to document relevant aspects of patient care. Providing an example with colonoscopy, Dr. Allen, in private practice with Minnesota Gastroenterology in Minneapolis, reported not only have they accumulated data on outcome and patient satisfaction but such details as what proportion of procedures started on schedule. These data have revealed a variety of ways to streamline procedures, improve quality of care and reduce costs.
As a practice, you need to decide what to measure and agree on how to measure it, but you will be amazed at the ways in which these data can be used to improve care, Dr. Allen reported.
Evidence-based medicine has been a catch phrase to identify delivery of care that is based on objective data. The recent Rand study indicates, however, that it takes more than objective data and clear endorsements by organized medicine to alter physician behavior. Whether a carrot, through such approaches as pay for performance, or a stick, through such negative incentives as a block on referrals, are used to change physician behavior, there seems to be agreement that one or both are coming.
ISSUE: OCTOBER, 2006 | VOLUME: 33:10
The Ups and Downs of Pay for Performance
Ready or Not, Here It Comes
Ted Bosworth
Los AngelesIn a recent symposium held during Digestive Disease Week, there was no shortage of disagreement on the issue of pay for performance. And while potential problems were envisioned for essentially every method proposed for using financial incentives to induce physicians to improve quality of care, participants also acknowledged that avoidable complications are expensive and occur often.
There is an important study just published that suggests that in all areas of medicine, we get it right only about 55% of the time. It is amazing how consistent this was across the different conditions that were evaluated, observed Lee N. Newcomer, MD, chief medical officer, United Healthcare, in Chicago. Referring to a study conducted by the Rand Corporation that looked at quality of care for 30 chronic and acute diseases in 12 communities in the United States (Asch SM et al. N Engl J Med 2006;354:1147-1156), Dr. Newcomer suggested we clearly have a lot of room for improvement.
One of the problems of pay for performance, which describes a system in which physicians are compensated for following specific protocols that are based on best practice, is that the proportion of medical conditions for which there are well-established and accepted guidelines remains small. According to George Diamond, MD, Cedars-Sinai Medical Center in Los Angeles, only about 15% of clinical interventions fall into areas of practice where such guidelines might be appropriate, and these represent an even smaller amount of total revenues.
Pay for performance only applies to what can be measured, and a lot of what we do cannot be measured. It also only tends to deal with one disease at a time when patients often have multiple medical problems, Dr. Diamond said. Moreover, he expressed concern that pay for performance places too great a focus on processes and not enough focus on direct evidence that outcome has been improved. In a complicated patient who does not fit the guidelines, pay for performance could be a potential negative incentive for optimal management.
Measuring quality is the stumbling block for any methodology to change physician behavior, including nonfinancial methods that involve peer review. At the University of Utah, David J. Bjorkman, MD, dean of the School of Medicine, said the effort to enjoin physicians to adhere to protocols is much like herding cats. He reported that the first step is to enlist physicians to buy into the legitimacy of the measures of quality of care. Physicians are not likely to be impressed with imposed processes if they are not convinced that they yield better outcomes. However, the University of Utah is developing a large data pool which it hopes will provide definitive evidence for the approaches that yield the best outcome. The data should help, but physicians will need to believe them.
We need a culture change, said Dr. Bjorkman, a past president of the American Society for Gastrointestinal Endoscopy. Referring to the willingness of physicians to rapidly accept and adopt specific protocols when proven effective, he cautioned that although we want innovation, we can no longer reward independence. He warned that bureaucrats who monitor healthcare reimbursement have a strong interest in establishing cost-effective care and emphasized that if we do not set the standards, then someone else will.
At United Healthcare, a variety of initiatives are being pursued to encourage physicians participating in their system to deliver better care. One example provided by Dr. Newcomer was in colorectal cancer resections. Despite well-publicized standards that at least 12 lymph nodes should be excised for pathologic evaluation in these procedures, the U.S. average has remained unchanged over several years at a substandard nine lymph nodes. Citing 2004 data, Dr. Newcomer reported that 54% of colorectal cancer resections are incorrect on this basis. At United Healthcare, physicians who want to be on a referral list must now present five case reports in which they demonstrate that the procedure was done correctly with 12 lymph nodes excised.
We have received very strong support from surgeons, reported Dr. Newcomer, who noted that the American College of Surgeons has endorsed removal of 12 lymph nodes as a standard of care in colorectal cancer resections. He said that gastroenterologists in their system are now provided with a $75 incentive to call United Healthcare to report a new diagnosis of colorectal cancer so that a nurse can contact the patient, describe the importance of standard-of-care therapy, and encourage them to use a surgeon on their referral list.
Differentiating this approach from pay for performance, which he believes has limited applications, Dr. Newcomer says that more fees will be awarded for specific quality actions. Of the $75 fee paid to gastroenterologists who make a call to United Healthcare to report a new colorectal cancer diagnosis, he said if you have to make an extra effort to help us solve a problem with quality of care, we will pay for that. It is only fair. He also said there are many efforts to make life easier for those physicians who are trying to provide both high quality and cost-effective care. This not only includes placing them first on referral lists and providing higher financial rewards but looking for other forms of preferential treatment. He acknowledged that we are at absolute infancy in learning how to do this well, but we would rather start with something than not move in this direction.
The major problem with fee for service, Dr. Diamond said, is that paying more for more services encourages overutilization. The problem with managed care is that it can provide incentives to do less, particularly when coding is insufficient to compensate for a complex clinical case. A potential problem for pay for performance is that rewards are often provided so long after the activity that they benefit those already delivering good care but do little to change physician behavior. One novel approach first proposed by Dr. Diamond in 1993 (Diamond G et al. J Am Coll Cardiol 1993;22:343-352) is a fee-for-benefit program in which larger fees are paid for procedures that can be expected to yield larger benefits. This encourages physicians to perform those services most likely to improve outcome and avoid procedures with little likely benefit.
On computer simulation, the costs go down, the quality of care goes up, and physician reimbursement goes up, Dr. Diamond said. He said that the benefit from such an approach accrues to the payer, the provider and the patient. Although this approach has not come much closer to application than when it was proposed more than 10 years ago, he maintained that the principles remain valid.
One approach to improving quality of care favored by some third party payers, such as the U.S. Centers for Medicare & Medicaid Services (CMS), is simply to make data on quality of care available so that consumers can make their own choices. However, there are a number of problems with this approach. One is that there is no evidence that these data influence patient choices. Another is that many patients may have trouble interpreting what the data signify. However, the greatest concern is that physicians will feel compelled to avoid sick patients to reduce the risk of bad outcomes.
There is always the danger of gaming the system, warned Dr. Bjorkman. He questioned whether it can be concluded that a hospital discharging 91% of postmyocardial infarction patients on an ACE inhibitor is really doing a better job than one with 85% on this drug at discharge. We have seen misrepresentation of data in our area in advertising. There are a lot of problems with this approach.
There is agreement that data are the key to improving healthcare no matter what approach is imposed. In his review of initiatives to improve quality of care at the second-largest independent gastrointestinal practice in the United States (50 gastroenterologists and 450 employees), John Allen, MD, reported that one of the first steps was to enlist all the physicians to agree to be transparent in their practices and assist in collecting data to document relevant aspects of patient care. Providing an example with colonoscopy, Dr. Allen, in private practice with Minnesota Gastroenterology in Minneapolis, reported not only have they accumulated data on outcome and patient satisfaction but such details as what proportion of procedures started on schedule. These data have revealed a variety of ways to streamline procedures, improve quality of care and reduce costs.
As a practice, you need to decide what to measure and agree on how to measure it, but you will be amazed at the ways in which these data can be used to improve care, Dr. Allen reported.
Evidence-based medicine has been a catch phrase to identify delivery of care that is based on objective data. The recent Rand study indicates, however, that it takes more than objective data and clear endorsements by organized medicine to alter physician behavior. Whether a carrot, through such approaches as pay for performance, or a stick, through such negative incentives as a block on referrals, are used to change physician behavior, there seems to be agreement that one or both are coming.
