Study finds generation gap in guidelines
From the December ACP Observer, copyright © 2005 by the American College of Physicians.
By Janet Colwell
While internists have approached single-disease practice guidelines with caution, they now have data to back up some of their misgivings. A new study concludes that most guidelines fall short for elderly patients with multiple chronic diseases, a group that makes up the lion's share of internists' practices.
That support comes by way of an article in the Aug. 10, 2005, Journal of the American Medical Association (JAMA). The study looked at the hypothetical case of a 79-year-old woman with chronic obstructive pulmonary disease, type 2 diabetes, osteoporosis, hypertension and osteoarthritis.
When researchers looked at practice guidelines for each illness, they found that if every relevant guideline was adhered to, this hypothetical patient would be prescribed 12 medications requiring her to take 19 doses per day at a monthly cost of more than $400. Treatment would also come with a list of recommended lifestyle modifications—and a host of potential drug-drug and drug-disease interactions.
But the study did more than pinpoint how guidelines can break down when it comes to medically complex patients. It also noted how those guidelines, when put in the context of the growing pay-for-performance movement, may force physicians to choose between good care and financial rewards.
"By rewarding practitioners for providing certain aspects of care based on single-disease guidelines, pay-for-performance initiatives may create incentives for ignoring the complexity of patients with multiple comorbid diseases," said Cynthia M. Boyd, ACP Member, lead author of the JAMA article and assistant professor of medicine at Johns Hopkins University's center on aging and health in Baltimore.
The study highlights a central problem physicians face when caring for these patients. While there are plenty of disease-specific recommendations, there is little to no advice on how to rank those recommendations in terms of clinical value or appropriateness.
National policy-makers and insurers are well aware of that issue, said Kevin B. Weiss, FACP, chair of the performance measurement subcommittee of the Ambulatory Care Quality Alliance (AQA), a national consortium charged with devising a strategy for implementing performance measurement. But because there is so little scientific data on how multiple chronic illnesses intersect, he added, the situation is unlikely to change anytime soon.
'The pay-for-performance movement may be getting ahead of the science right now.'
—Kevin B. Weiss, FACP
"We're at the point where we have to make these decisions on payment policy, and those of us who are thinking about the clinical side are saying 'caution' because we haven't worked out these multiple chronic illnesses issues yet," said Dr. Weiss, who is also director of the Midwest Center Health Services and Policy Research Institute at the Hines VA Hospital in Hines, Ill., and Chair of ACP's Performance Measurement Subcommittee. "The pay-for-performance movement may be getting ahead of the science right now."
According to data released in 1999 by the Centers for Medicare and Medicaid Services (CMS), 48% of Medicare beneficiaries age 65 or older had at least three chronic conditions, while 21% had five or more. Beneficiaries with at least three chronic illnesses accounted for 89% of Medicare's annual budget.
In the short-term, internists may find themselves having to choose between effectively managing these patients and some pay-for-performance incentives.
"At the national level, we're still in the early stages of even looking at single-disease guidelines," said Dr. Weiss. "The intent is to someday have incentives to reward optimal management of multiple chronic illnesses, but right now there is only a hope that we can get there in the next several years."
Guideline generation gap
Experts say another factor tends to make many practice guidelines less than useful when treating elderly patients: Most algorithms use data that don't reflect the experience of seniors.
"Guidelines are developed for people in middle age or late middle age who have one or a few chronic diseases," said Caroline S. Blaum, MD, associate professor of internal medicine and geriatrics at the University of Michigan and a research scientist with the VA Geriatric Research Education and Clinical Center in Ann Arbor, Mich. "You get to a situation where the evidence does not directly apply to older people nor people with multiple diseases."
A good example of the guideline generation gap can be found in diabetes. While diabetes-specific quality measures have been extensively tested, Dr. Blaum said, standard guidelines may not apply to people over age 75. Dr. Blaum and others grappled with that issue when they created diabetes guidelines for older patients in 2003 for the American Geriatrics Society (AGS) and the California Healthcare Foundation, a health care delivery and financing foundation.
Dr. Blaum explained that because most studies looked at people age 65 and younger, most evidence is being extrapolated to older patients. "We have no idea what would actually happen if we did the same studies in 80-year-olds," she said.
In some areas, she noted, there is evidence, but it's not definitive. "There's a lot of evidence that it's good to treat high blood pressure in the very old, for example," she said. "but there's no evidence as to what the level of treatment should be. Many diabetes guidelines now say the lower the better, and that's probably true if you're in your 50s—but what if you're in your 90s?"
While they may not be perfect, the AGS guidelines provide rare guidance for physicians trying to prioritize single-disease treatment recommendations for older patients with multiple diseases. (The JAMA study co-authored by Dr. Boyd also cited ACP's chronic stable angina guideline for including guidance on comorbidities.)
The AGS panel delineated where evidence existed, where it was extrapolated from other groups and where there was some evidence but not at the same level as for younger patients. It also advised physicians to look for common geriatric conditions.
"We know that many geriatric conditions that affect older people, like falling, incontinence, depression and cognitive impairment, are very common in older people with diabetes," said Dr. Blaum. "The AGS guidelines put forth that it would be very useful for doctors to look for these conditions in older people with diabetes."
In addition, while traditional diabetic guidelines often focus on glycemic control, the AGS panel acknowledged that it might have a lower priority for older patients with other diseases.
Some experts say that even guidelines that incorporate age-based evidence may still fall short in clinical practice. Samuel C. Durso, ACP Member, clinical director of geriatric medicine at Baltimore's Johns Hopkins University and an advisor for the AGS guidelines, noted that treating older patients with multiple diseases is by necessity individualized because patients cannot be stratified by age alone. For example, two 75-year-old female patients with diabetes—one highly functioning and one poorly functioning—have different predicted longevity and may have very different preferences for treatment.
"The very fit woman has a higher likelihood of benefiting from intensive blood pressure, and possibly blood sugar, management," Dr. Durso explained. "The other extreme is the frail woman with a short life expectancy who has diabetes and osteoporosis with a high risk for falls. She might want to focus more on reducing her medications to lower her risk of falls as well as gait and balance training, reconditioning and starting on an agent to treat osteoporosis. So you've got two women the same age who both have diabetes and your recommendations might be very different."
Fifteen years ago, the debate over how useful practice guidelines are was largely academic. Now, as payers measure physicians' adherence to key parts of guidelines, the issue is taking on more practical overtones.
"If you are going to look at the whole country and pay doctors to do something, the level of evidence has to be very high," said Dr. Blaum. "Most patients would have to be covered by the quality measure."
The lack of clinical evidence on how to treat the older patient with multiple diseases is even more significant when considering performance incentives at the national level. Performance measures for diabetes, for example, are not based on data from patients who are older than age 75, yet 40% of diabetes patients with Medicare are in that group.
"From the standpoint of quality assurance, I would want most of my patients with diabetes to be treated according to the traditional quality measures that are being put forward," said Dr. Blaum. "But if you're going to tell all the doctors in the country to make a person's blood pressure 130/75, does that really apply to a 90-year-old with advanced dementia?"
Dr. Boyd said it's one reason physicians are increasingly calling on pay-for-performance programs to address more complex issues like care coordination, patient and caregiver education, counseling on self-management, and complex decision-making, including taking a patient's circumstances and preferences into account. That's because assessing a physician's ability to make complex treatment decisions is far different from measuring whether or not a patient was given a medication for high cholesterol or an ACE inhibitor for congestive heart failure.
"This is one of the challenges we face in elder care," said Richard D. Della Penna, MD, director of Kaiser Permanente's Aging Network, which is based in Oakland, Calif. "The things that are used to measure improvements in diabetes, heart failure and other chronic diseases can be pulled off of electronic data, but it's hard to do that once you start throwing in things like advanced years, frailty or cognitive impairment. It just becomes more complicated."
Physicians and organizations may respond by removing sicker or more complicated patients from their "denominator" (the patients included in a population for reporting purposes). For example, a standard might apply only to people with a disease who are under age 76, or certain patients deemed very sick or with certain diagnoses might be excluded from a population for reporting purposes.
However, Dr. Boyd pointed out, even though these complex patients would remain in the practice, excluding them from the denominator may still remove incentives to improve their care. "Pay-for-performance measures may focus physicians' attention on single-disease interventions that miss the bigger picture for more complex patients," Dr. Boyd said, "or focus their attention on younger, healthier patients."
By the same token, physicians who continue to provide complex care for older patients with multiple diseases are likely to become frustrated if their efforts aren't being rewarded.
"Physicians are under pressure in our traditional system to take care of these patients," said Dr. Blaum. "If you have an 85-year-old with 15 diseases and 10 medications, it's hard to see them in a 10-minute visit and pay for performance might make that pressure worse. So it's a little bit scary for the frail elderly."
With pay-for-performance moving full speed ahead, it remains to be seen if these issues can be worked out before most physicians are asked to participate in incentive programs. In the long term, the medical community is working to ensure that incentives become increasingly relevant to the way physicians actually practice.
"We're at a very early stage of development of performance measures for individual physicians and an early stage of pay for performance, and up until recently there was no reason why those two had to go together," said Dr. Weiss. "It's our role as physicians to try to make sure that our incentives are not going to be in conflict with providing good clinical care."
The information included herein should never be used as a substitute for clinical judgment and does not represent an official position of ACP.
Internist Archives Quick Links
Have questions about the new ABIM MOC Program?
One Click to Confidence - Free to members
ACP Smart Medicine is a new, online clinical decision support tool specifically for internal medicine. Get rapid point-of-care access to evidence-based clinical recommendations and guidelines. Plus, users can easily earn CME credit. Learn more