MIPS Part Two: How "Quality" will slay real quality in geriatric health care
May 21, 2016
How can quality be measured in health care? Certainly Medicare feels that they have the answer. Half of primary care doctors’ MIPS grades will be based on how we prove that our care is “quality” care. And since our grade will determine our salary, this stuff is really important. You may think that if we are taking good care of our patients based on their individual wants and needs, if we take time to talk to them about their medical problems, if we avoid overtreatment and drugs with excessive side effects, and if we use the best data available to work with our patients to make shared decisions, then we are practicing quality care. But in fact, this approach is the very antithesis of Medicare’s definition of quality. It will give us an automatic F.
What Medicare considers quality is protocol based medicine. We have discussed protocols, or quality indicators, many times in these pages already. Essentially a medical protocol is a binary means of assessing the efficacy of an intervention: either the doctor/np has ordered the test or has not ordered it, given the medicine or has not given it; there is a right answer and a wrong one. There is no nuance, no individual distinction. You are either right or wrong. For instance, in our medical practice we have been discussing with some of our very elderly patients and their families the risks and benefits of statin cholesterol medicine. There is minimal proof of meaningful clinical benefit in the elderly, and ample evidence that these drugs can cause fatigue and muscle weakness. Most elderly patients and their families are happy to shed some medicines, especially when quality of life takes priority and when the benefits are so small. But now, in our ACO, we are being graded by how many of our patients are taking statins. It doesn’t matter how old they are, what their own goals of treatment are, what their co-existing conditions are, if statins are impacting them in a negative way. They are either on the statin (we pass), or they are not on the statin (we fail). That is protocol based medicine. That is Medicare’s definition of quality. Doctors get paid for performance in a binary way.
My friend Steve Martin shared with me a recent article in BMJ that outlines the problems with a protocol-based approach to quality assessment in health care (CLICK HERE for a copy of the article). The article reiterated much of what I and many others have concluded in regards to grading quality this way. State the authors: “Pay for performance systems are part of the problem. They do not encourage shared decision making because doctors are penalized if they do not comply with [the protocols]. Such schemes should take into account what choices are acceptable to patients, what trade-offs are acceptable, and what factors such as more time or better information would facilitate more options.” In other words, “one size fits all” protocols that will determine a doctor’s pay discourage shared decision making and individualized care. They, in fact, push us to make very poor decisions that may be harmful to the patient sitting in front of us and not consistent with that patient’s wishes, and they discourage us from discussing the decision with the patient, because any course we take that deviates from the “right” answer in the yes/no protocol system will cause us financial harm.
The protocol system as a measure of quality is as Orwellian as it comes. The more we follow the “quality care” script, the worse will be the actual quality of our care. We will outline why this is so below. An earlier article in BMJ provides a check list of how quality should be measured based on evidence and common sense (CLICK HERE for the article). My book, Curing Medicare, spends most of its pages similarly demonstrating how Medicare’s quality indicators instigate poor medical care by deviating from what a large body of evidence has shown us about geriatric health, and how we can alter such protocols to derive much more meaningful measures of quality. The problems with Medicare’s indicators are beyond the scope of this blog, but a few issues are:
Use of Faulty Data: The reality is that most protocols are not based on good data, and there is no good evidence that they work. The BMJ estimates that 62% of research used to create primary care guidelines is of nebulous relevance to primary care patients. In fact, only 11% of interventions advocated by protocols have been clearly shown to be beneficial to the patient; 50% have completely unknown impact. And virtually no patient fits the “ideal” patients who are enrolled in clinical studies. Most of my patients are older, are on more medicines, have more co-existing problems, and are frailer than any study subjects. Therefore , the data when applied to them is of specious benefit; my patients may be as easily harmed as helped by interventions that protocols consider standard of care. For instance, a Medicare protocol that implores us to lower blood sugars below a certain threshold (A1C of 9) is arbitrary and is not based on evidence of improved outcome; it can in fact lead to low sugars which we know cause harm and even death. And the benefit of many interventions advocated by protocols, even those that may have some merit, are grossly exaggerated. As we have shown, the use of relative benefits to demonstrate the value of a medical intervention is often misleading. Thus most of the protocols Medicare is forcing us to follow—from using Coumadin in afib, to lowering blood pressure below 120, to using beta blockers and statins in heart disease, to ordering mammograms on most women—have typically very tiny absolute benefit (a few out of 1000 benefit) and possible similar or larger risk. The science behind protocols does not support clear utility for most individuals. Care that Matters (www.carethatmatters.org), a group out of Boston with which I am affiliated and which is doing great work to reverse the trend of protocol-based treatment, graded several of the Blue Cross quality indicators based on their proven efficacy. CLICK HERE to see what they determined; it may well surprise you. You can also look back at some of my blogs on the subject, or buy my book, which spends most of its time debunking the merit of clinical protocols.
Pay for Performance Does Not Work: There is no data that paying doctors to perform certain tasks actually results in better quality of outcome. In fact, some data shows just the opposite. What pay for performance does reward is encouraging and rewarding doctors to “play the system.” This is the primary reason that larger organizations will do better with MIPS; they can master the rules needed to win. It is not about taking care of patients or providing good quality care. It is knowing the protocols and how to beat them. It is forcing doctors to persuade patients to do the “right” thing, even if for those patients it may be the wrong thing. When the actual benefits of pay for performance systems are assessed, as they were in a recent study of hospital pay for performance measures (BMJ, 5/10/16), the results are often disappointing. And Medicare’s assumption that pay for performance will save the system money has been discredited by various pilot programs, some of which have been sponsored by the Medicare Innovation Center. In fact, not a single study shows that they save money or improve outcome. Yet, MIPS has vaulted protocol-based assessments to the centerpiece of its quest for quality and value. That would make George Orwell smile.
Protocols Discourage Individualized Care. Recently I talked to a patient in his late 60’s about getting a pneumonia vaccine. I showed him his individual absolute benefit (tiny) and risk (also tiny), and the patient, who had no health problems and was skeptical of interventions not absolutely necessary, deferred the immunization. I told him that Medicare would not be happy with his decision, and, with a serious glance, he said that he would get the immunization if it would help me with Medicare. Is that the actual goal of clinical practice guidelines: to help doctors check off the correct answer on their MIPS test despite what the patient wants and needs? There is no right answer as to whether most interventions are effective; the binary yes/no protocols upon which we will be judged are not applicable to any individual patients. For one patient, lowering blood pressure below 120 may help him avoid a bad outcome; for another it may lead to worse memory, fatigue, falls, and even death. Interventions can be life saving or deadly; they can lead someone to feel better or worse. Some tests advocated by the protocols have no value to certain patients, and may be very useful to others. Interventions must be assessed based on each patient’s medical problems, prior reactions, medications, symptoms, and clinical status; there is no single path that works for everyone. When doctors are paid to follow a protocol, and when that protocol may lead to treatment that is deleterious to a certain patient, then doctors are encouraged to do the wrong thing. To follow protocols blindly can lead a doctor to break the primary rule of medical care: do no harm.
Protocols Are Antithetical to Shared Decision Making. When presented with interventions that have uncertain value and can cause possible harm, doctors and patients should sit down and discuss risks and benefits, and then come to a shared decision. The efficacy of this approach has been widely demonstrated and is generally what patients prefer. Each patient must evaluate an intervention in terms of his/her own specific medical circumstances, his/her own values and perceptions, and his/her own goals of care. When one person sees that using Coumadin in atrial fibrillation can prevent 6 strokes out of 1000 people and cause 12 fatal bleeds out of 1000 people, he may want to take Coumadin for fear of stroke, while another person may see that benefit as negligible when weighed against the risk and inconvenience of the drug. Medicare is telling us there is one right answer: If your patient takes Coumadin you pass, if he does not then you fail. Thus, when we spend time discussing issues with our patients, we are in fact hurting ourselves in the MIPS universe, because we are providing patients with information that may lead them to not follow the protocols. Best that we dictate to our patients what they must do, do not discuss risks and benefits with them, do not ask them what they prefer, and chastise them if they do not follow the course we prescribe. That is how Medicare defines quality. Neither doctors nor patients are involved in decision making. The protocol is the omnipotent decision maker; it is Big Brother.
Protocols lead to Overtreatment. While protocols are supposed to enhance quality and cut cost, in fact they do just the opposite. They are kindling for overtreatment. The best way to game the system and get the right answers is to order a plethora of tests, put the patient on a ton of drugs, and don’t worry about the consequences. There is no protocol in MIPS that seeks to curb overtreatment; all protocols are designed to fail doctors who do not test and treat enough. Thus the doctor who orders a stress test on an asymptomatic patient every year, who puts stents in arteries when such a procedure causes more harm than good, who drops someone’s blood pressure so low that she falls and breaks a hip, who aggressively treats diabetes to levels that lead to known severe risk, who uses high-dose statins even if patients are clearly having life-altering side effects and no proven benefit, who treat with Fosamax for over 5 years and cause a patient a spontaneous fracture; these are a few of the limitless examples where overtreatment causes poor outcomes, and in all of these examples the doctors performing the overtreatment will be rewarded with passing grades in the MIPS testing scheme. The ideally treated patient in the MIPS world gets incessant tests that may lead to false positives or nebulous results, is on 20 medicines that interact with each other, and although he may be mentally and physically impaired by such “quality” care, and although he prefers a different approach, his blind compliance with overtreatment will give the doctor a passing grade. Protocols are frankly dangerous and expensive. They encourage doctors to be rewarded by thinking less, ignoring the data, ignoring the patient, and doing what is clearly the wrong thing in excess. And they are exactly what patients do not want and what Medicare does not need.
Protocols Favor Doctors who Care for Healthy Patients. Healthy, financially well-off patients typically adhere more to protocols than those who are poor, who are more ill, who have mental issues, and who are very old. Thus, doctors who care for healthy populations will be rewarded, and those who care for the sickest of our patients will be punished. Ultimately even doctors who help make significant impacts in the health of very ill patients will still fail the MIPS quality test, since the binary Passing grade in MIPS is difficult to achieve in this population. So MIPS will encourage doctors to eschew care of the ill and embrace care of the wealthy and well. Ultimately, due to the severe financial penalties that will be incurred, doctors will simply not be able to afford to care for “less desirable” patients. Is this what the intent is of pay for performance?
Thus the quality report card of MIPS clearly should get an F. Rather than enhancing quality, it pays doctors who do just the opposite. Blindly following protocols is a dangerous path to follow, both for the patient and for the system. And yet it will be a path that Medicare will generously reimburse.
Then how should we measure quality? I wrote an article in Primary Care Progress talking about what I consider the most effective way to provide quality care: talk with your patient and make shared decisions. (CLICK HERE for article) This is not a new concept. Concludes the BMJ article earlier cited: “We believe that GPs should not be paid according to how many patients comply with guideline recommendations. Instead, they should be encouraged to make decisions according to the evidence and patient preference.” In other words, it is not the outcome of the patient’s decision that should be graded, rather it is the process upon which the doctor and patient arrived at that decision. Studies show that this is what doctors and patients prefer, that it leads to better medical outcomes, and that it can save the Medicare system money. A short video (CLICK HERE for video) on Erik Rifkin and my website www.doc-patient-talk.com demonstrates how this may work. Our website and many others, including www.thennt.com and the soon to come www.tandemhealth.org, will provide patients and doctors with easily understood visual tools based on actual medical evidence to enable quality shared decision making. Perhaps Medicare should be rewarding the conversation: doctors and NPs who use quality data to discuss issues with their patients before recklessly plunging into nebulous, dangerous and expensive interventions should be the ones who achieve a Passing grade. Such a simple fix to MIPS will actually improve quality and value, and it is what doctors and patients truly crave. But if MIPS perseveres as it is scripted now and quality is measured by protocol, then the use of actual data, individualized care, and shared decision making will be deleterious to a medical practice’s survival and will thus be discouraged. “Quality” will slay quality. Ultimately we have to make a decision about how to provide the best health care at the lowest price. If we allow MIPS to be implemented as currently designed, that decision will be made for us, and the outcome will be something that none of us want.