Do quality indicators lead to over-intervention?
Some suggest they don't take the needs of elderly patients into account
The use of evidence-based quality indicators to ensure consistent treatment is a great idea because it improves patient care and outcomes, right? While most people would agree that the various organizations that create the measures have the best interests of patients at heart, there are others who think the cookie-cutter approach to medicine causes problems for some specific groups of patients.
In a commentary from the Journal of the American Medical Association (JAMA) last fall, Sei J. Leeand Louise C. Walter argue that for older patients in particular, many of the required reporting elements are inappropriate and can lead to excessive care or care that does nothing to improve the quality of life of the patient.1 In one example, they note that the U.S. Preventive Services Task Force recommends discontinuing colorectal cancer screening for people with a decreased life expectancy. But the Healthcare Effectiveness Data and Information Set (HEDIS) calls for screening for people between 50 and 75 years. Where does that leave a 70-year-old patient with advanced lung disease whose life expectancy would discount the benefits of screening?
"The quality indicators and guidelines are based on research, and that can be driven by the profit motive," says Lee, an assistant professor at UC San Francisco and associate director of the San Francisco VA Quality Scholars Fellowship, Division of Geriatrics. "It is focused on seeing if a treatment or protocol works, rather than if it doesn't work, on doing something rather than not doing something. It is tilted toward intervention. There is not a lot of quality data about over-intervention."
Another issue is that older adults often have so many problems that they are poor candidates to participate in most clinical trials and medical studies. "You have to do a much larger, much more expensive study to develop the same level of information," Lee says. "They have competing risks that younger patients with a single problem don't have. Teasing out what might be because of your intervention and what might be because of some co-morbidity is hard. So we extend findings to older adults and don't force those who create protocols or treatments to take the next step and prove it works for older populations."
The potential problems are not just theoretical but are something that physicians witness regularly, says Julianna Lindsey, MD, MBA, regional medical director of hospital medicine for TeamHealth in Knoxville, TN. "I face this a lot when we have very elderly patients and they have advanced dementia and are bed-bound. You may confirm they have had a myocardial infarction, and core measures say you have to give them a beta-blocker and statins. The same is true with stroke. But that might not be the right thing in this situation. I'm told I have to do this, but it might not improve their quality of life at all. If they could speak for themselves, would they really choose this?"
Other patient populations are affected as well, says Craig Schranz, MD, an emergency department physician in Norfolk, VA. In the emergency department, there are concerns about evaluating CT scan rates for a-traumatic headaches. "The thought is that they may be unnecessary," says Schranz. "But our job in the emergency department is to think the worst first. There are a lot of bad things that can cause a headache that are not trauma-related. If you scan them and are wrong, it hurts you," he says, noting that physicians can lose payment based on quality initiatives that frown on such scans. Worse, though, is if you don't do it and the patient needs it. "The indicators don't ask whether the scan was a good idea or for an explanation," Schranz continues. This is very controversial in the ED world because the National Quality Forum advised against this particular indicator, while the Centers for Medicare & Medicaid Services (CMS) continues to move forward with it.
Another indicator that Schranz says was a problem related to how many hours passed before antibiotics were given to pneumonia patients. "It used to be four hours, then six hours, and now it's finally gone. It was creating a perverse disincentive. There is no science behind that six-hour rule. So if someone was near that time but hadn't had a diagnosis yet, you might treat them anyway, before you even confirm a diagnosis with an X-ray. But that patient may not even have pneumonia." That leaves someone open to potential adverse events from a medication allergy or interaction, he says.
Another issue involves aspirin given to heart attack patients, says Schranz. "Every now and then you find a chart where aspirin wasn't given. Maybe the patient is allergic, or the patient was given it en route to the hospital," he says. "But do they screen the data for things like this?" What about in the cardiac cath lab, and the time to thrombolitics in stroke patients? "There could be a rush to take an action that might not be appropriate if you stop and take the time you need."
Some sense, but not enough
Schranz says he understands the rationale behind indicators, "but these measures take on a larger role with things like Hospital Compare, where the facts and figures are laid bare, without context, for consumers. They don't look at why a standard isn't met in a particular case."
The measures are a good start, but there needs to be leeway, says Schranz. "It takes so much energy to make changes. The antibiotic issue had tons of literature saying the measures weren't good. With the headache thing we have experts — NQF is the expert — saying it's not a good idea and CMS going ahead with it. Our perception is that they don't respond to clinician concerns in a timely manner."
"I agree that there are some valid points [in the JAMA commentary]," says Margaret E. O'Kane, MHS, president of the Washington, DC-based National Committee for Quality Assurance (NCQA). "But HEDIS is a work in progress, and it will get better." She also thinks that some of the concerns raised by providers and in the JAMA article are misplaced. "The theme of life expectancy seems way out there. You can't find that in a chart, and I think you could even put an unattractive spin on it like death panels — that someone is too old or too sick to have a particular treatment."
O'Kane says the NCQA is working to enumerate clinical exclusions for many measures and knows that for the elderly, this is a particularly important issue. "We do generalize treatments to the elderly, and I believe we should be doing more trials with them."
But acknowledging the concerns are real is only half the battle, says O'Kane. "We have to figure out a way to address them, and I think part of that is for the medical profession as a whole to get busy on some evidence-based guidelines for when treatments are not appropriate. We don't develop the guidelines, we follow them."
She also encourages physicians who have problems to send those organizations promoting the indicators letters and comments about exclusions that seem appropriate. "This isn't a war between sides. Point out the problems, but also understand what is possible."
Lee says one relatively easy fix would be to extend rules implemented by the National Institutes of Health to include women and minorities in research to also include older adults. "Make researchers explain how they will include geriatric populations or justify the reasons why they don't," he says. "If nothing else, it will force people to think critically about whether the benefits of limiting research outweigh the potential harm."
Lindsey says physicians need to be able to act in the best interest of their patients and work on a case-by-case basis. "We need to be given the leeway to forgo some of the requirements. We are on the ground doing the work and people aren't giving thought to the potential adverse consequences of these rules."
For more information on this topic, contact:
- Julianna Lindsey, MD MBA, Regional Medical Director, Hospital Medicine, TeamHealth, Knoxville, TN. Email: email@example.com.
- Margaret E. O'Kane, MHS, President, National Committee for Quality Assurance, Washington, DC. Telephone: (202) 955-3500.
- Craig Schranz, MD, Emergency Department Physician, Norfolk, VA. Email: firstname.lastname@example.org.
- Sei J. Lee, MD MAS, Assistant Professor, UC San Francisco and Associate Director, San Francisco VA Quality Scholars Fellowship, Division of Geriatrics, San Francisco, CA. Telephone (415) 221-4810 ext. 4543.
- Lee SJ and Walter LC. Quality indicators for older adults: preventing unintended harms. JAMA. 2011; 306(13):1481-82.