We live in an era of quality monitoring—from the quality of products, like water or refrigerators, to the quality of customer service received from repair technicians, food delivery services, and even customer service representatives. This drive for quality measurement has naturally spread into the world of medicine.
Since 1990, the medical care delivery model has changed. Instead of receiving care from private practitioners in solo or small group, physician-owned and led practices, patients increasingly receive team-based care delivered through corporate or hospital-employed physicians.
Accompanying this change in care delivery models has been a change in the vocabulary of medical care; instead of “patients and physicians,” we refer to “clients” (or “customers”) and “providers.”
This has led to a consumerization of medicine. It is not surprising that doctors are now rated for customer satisfaction factors such as accessibility, affability, and ability to communicate. Outcome and satisfaction-based quality scores are operationalized as contributors to the base salary and bonus pay of physicians.
Both factors can contribute to burnout and the potential role of burnout in adverse medical outcomes, which has been previously explored. 1
In this piece, I propose the converse, i.e. that the pursuit of unachievable or unjust quality measures is contributing to the epidemic of physician burnout.
As a provider and the physician member of the hospital quality committee, I felt growing frustration with provider quality scores that reflected measures outside of a physician’s control. Feeling one’s effort is futile and that one has no control over the outcomes are two common causes of burnout.1
I reached out to colleagues nationally to find it was a common sentiment, and then did a literature search and found that the observed link was novel. This is the first publication suggesting misguided quality measures may contribute to physician burnout.
Measuring the quality of medical care in hopes of improving outcomes is a noble goal, and certain outcomes such as in hospital infection and wrong side surgery have improved with these efforts. However, improvements have not been demonstrated in most quality measures.2 Furthermore, many factors conspire to make this measurement difficult.
Current efforts to improve the quality and decrease the cost of cancer care, led by CMS and certain certifying professional societies, have led oncologists to feel that they are jumping through an endless series of hoops to reach an elusive target.
These factors include: a) A reimbursement system which favors doing more but not necessarily providing better care, b) The team-based nature of medical care where many providers, physician and nonphysician, contribute to outcomes, and c) Systemic barriers that exist to reporting bad outcomes, safety concerns, or impaired health care professionals.
Another detractor from the pursuit of quality is the selection of measures that are easiest to extract from the EMR but of unproven benefit. Current efforts to improve the quality and decrease the cost of cancer care, led by CMS and certain certifying professional societies, have led oncologists to feel that they are jumping through an endless series of hoops to reach an elusive target.
Overwork, futile work, a sense of being held accountable for things outside one’s control, and moral injury are well recognized triggers of burnout.1,3,4 To the extent that quality measures may elicit these sensations, they will contribute to early physician burnout.
Let’s look at a few examples of oncology quality measures through this lens.
CMS adopted “chemotherapy within 14 days of death” as an indicator of poor care. This was based on the assumption that when this happened, it meant doctors were remiss in recognizing a grave prognosis and transitioning from active treatment to hospice.
This rationale discounts the role of patients who may wish to continue therapy despite thorough discussion of the futility and potential harm of continued treatment.
It discounts published findings that delivery of bad news, such as a poor prognosis, is associated with lowered patient satisfaction scores.5,6,7 Patient satisfaction scores, long a component of hospital reimbursement thru HCAPS, are increasingly being used to calculate provider bonus or salary.8
We must recognize there is a tension between these two performance measures—treatment within 14 days of death and patient satisfaction. We must also recognize that the decision to stop therapy is not entirely in the oncologists’ hands.
Nothing makes one feel as powerless as having one’s quality score depend on someone else’s decision or performance. An improvement in this quality measure would be to require documentation of a discussion of goals of care at critical transition points (for example at initial diagnosis, at first detection of recurrence, and at failure of second line therapy), and then to hold clinicians responsible for providing goal concordant care.
In the late 1990s, addressing pain became the fifth vital sign. In keeping with this, the American Society for Clinical Oncology QOPI determined that pain should be assessed and addressed at each visit.
Failure to assess pain lowers an oncologist’s QOPI score. Since cancer patients may have cancer-related pain, it is appropriate for oncologists to address cancer pain. However, many patients with cancer have chronic noncancer pain (e.g. back pain, diabetic neuropathy, fibromyalgia, etc).
Addressing this chronic pain at every single visit does not improve cancer care, distracts from addressing cancer related issues that are more pertinent to the visit and represents busywork. But failure to address it will lower your QOPI score. For oncologists, this measure should be limited to cancer-related pain.
Another CMS quality measure of dubious value is OP35. OP35 counts the number of times patients go to the emergency room or get admitted to the hospital within 30 days of chemotherapy. A higher number is considered an indicator of lower-quality care. But these are sick patients, often near the end of life!
They may develop a serious infection, a perforated viscus, major bleeding, or myriad other unavoidable life-threatening problems that cannot be safely addressed in the outpatient setting. Furthermore, patients who are offered work-in clinic appointments, refuse them, and end up in the ED a few hours later are not currently excluded from the OP35 calculation. These factors lower the physician score but are not in the physician’s control.
When spurious quality measures or team-based factors are used to calculate physician salary or bonus, the effort becomes unfair and will likely backfire, contributing to burnout.
Some larger cancer centers have tried to mitigate the negative effects of this measure by renaming their ERs. Same physicians, same urgent care, but the facility is now called the Acute Cancer Care Clinic to avoid getting demerits! This circumvents a CMS quality measure without actually changing the quality of care. It is not available to smaller, resource-limited cancer centers, and creates an illusory quality difference between large and small centers.
Physicians are held accountable for the OP35 quality measure even though ER visits may be outside their control. This contributes to physician burnout. An improvement in this quality measure would be to count ER visits in the context of whether patients had been offered work-in appts or whether the ER visit was a preventable one.
Promptness of care is psychologically important, especially for cancer patients, and is increasingly incorporated into patient satisfaction surveys and used by organizations to calculate physician compensation. Yet, many insurance companies require preauthorization for the first visit, each scan, and the treatment (sometimes one cycle at a time). Commercial insurers will not guarantee approval for standard therapies in less than 14 business days.
Each preauthorization step is manpower intense and interdependent, so delays are built into the process of being evaluated and treated for cancer. Yet, the patient has no idea the delays are beyond the control of the doctor and will hold the doctor responsible for these subsequent delays. This delay, that is outside of the physician’s control, inserts tension into a relationship that should be built on trust and erodes physician morale.
The promptness of a physician’s response to critical lab values is another quality measure tracked for CMS. Originally, this was intended as a quality measure to ensure that critical labs were promptly reported and acted upon,9 by both the certified lab and the patient’s physician. Hospitals then operationalized this by measuring how often physicians failed to respond promptly to a critical lab.
On the surface, it seems reasonable to expect doctors to answer these calls promptly. But without a denominator, this result becomes meaningless. A physician practicing in a data intense subspecialty, like hematology/oncology, may get 10 such calls a day whereas a family physician may get a total of 10 critical lab calls a month.
The absolute number of times a doctor fails to respond promptly is less meaningful than the percentage of the time that a doctor fails to respond promptly. The denominator is important. This indifference to work intensity creates distress. A simple fix would be to report this as a percentage of labs ordered.
Starting Jan. 1, 2023, all providers will be graded for compliance with CMS appropriate use criteria (AUC) when ordering advanced imaging studies.10 Failure to do so will result in demerits. Oncologists, who frequently order advanced imaging, will feel this burden disproportionately.
Again, on the surface, this performance measure looks reasonable. Only medically appropriate imaging should be ordered. However, it turns out that even if you fulfill AUC (and if NCCN recommends this study), many insurers will deem a study medically Unnecessary and will not pay for it.
Patients will be told their doctor ordered an unnecessary test and will be asked to accept financial responsibility for the test. Many patients may refuse the test. Not only does the effort of fulfilling AUC become futile, but the patients’ trust in their physician is undermined when they are told their provider has ordered a study deemed unnecessary by their insurance company.
This measure should not be rolled out until CMS and insurers agree that fulfilling AUC will lead to payment for the study.
Make no mistake, quality improvement efforts have improved some outcomes and the effort, whether by individuals or teams, is a laudable goal.
However, using team-based care outcomes to calculate individual physician salaries is unjust, particularly when physicians no longer have a say in hiring/firing or even evaluation of team members.
Although one may argue that simply knowing that your work output is being monitored improves the quality of your work, this theory has not been supported by data and it can certainly drive paranoia and workplace dissatisfaction.2
Finally, while the cost of quality improvement efforts has been high, the improvement in quality for most measures has been questionable.2 The fact that CMS suspended quality measurement during the COVID crisis is acknowledgement that quality efforts may divert attention from clinical care.
I have outlined how some quality and satisfaction measures can trigger recognized drivers of burnout.1,3,4 Burnout can lead not only to increased medical errors, but also to early retirement and physician suicide. Existing physician shortages and the cost (in dollars and years of life training) for each physician lost amplify the cost of burnout.
Quality measurement is worthwhile, likely keeps practitioners vigilant, and may improve patient safety. However, when spurious quality measures or team-based factors are used to calculate physician salary or bonus, the effort becomes unfair and will likely backfire, contributing to burnout.
I have been a doctor for more than 30 years and have yet to meet a doctor who wants their patients to get poor quality care. We all want better outcomes for our patients. The pursuit of quality is not the problem.
I have included a number of examples of quality measures that are beyond the control of a physician, at odds with another quality measure, or unproven to be associated with improved quality of care. Each factor can cause moral injury and drive burnout individually,3 but taken together, they give the physician a sense of perpetually striving for unattainable goals.
It is time to take a more rational approach to quality measures.
Include practicing doctors in the selection of quality measures, choose quality measures that doctors are directly responsible for, and if you cannot demonstrate an improvement in quality in a finite period of time, discard those quality measures. Don’t assume the existence of quality measures automatically improves health care delivery or is beneficial.2
This is a call to action to physicians, hospitals, and regulatory agencies.
References:
- Shanafelt TD. Physician well-being 2.0: Where are we and where are we going? Mayo Clin. Proc. 2021; 96 (10) 2682-2691.
- Rosenbaum L. Reassessing quality assessment: The flawed system for fixing a flawed system. NEJM. 2022; 386 (17), 1663-1667.
- Hlubocky FJ, Taylor LP, Marron JM et al. A Call to action: ethics committee roundtable recommendations for addressing burnout and moral distress in oncology. JOP. 2020; 16: 1-10 5
- Saunders EG. 6 Causes of burnout and how to avoid them. HBR, 7/5/2019.
- Weeks JC, Catalano PJ, Cronin A, et al. Patients’ expectations about effects of chemotherapy for advanced cancer. NEJM2012; 367(17):1616-1625.
- Fenton JJ, Jerant AF, BertakisKD, et al. The cost of satisfaction: A national study of patient satisfaction, health care utilization, expenditures, and mortality. Arch Intern Med 2012; 172 (5):405-411
- Tanco K, Rhondali W, Perez-Cruz P, et al. Patient perception of physician compassion after a more optimistic vs a less optimistic message. JAMA Oncology 2015; 1 (2): 176-183.
- Japsen, B. Physicians see more pay tied to patient satisfaction. Forbes, July 8 2019. Https://www.forbes.com › brucejapsen › 2019/07/08.
- The Joint Commission: National Patient Safety Goals Effective July 2020 for the Laboratory Program 94391035-17d0-4b30-91a8-998eadc82ae4
- https://www.eidebailly.com/insights/articles/2019/12/appropriate-use-criteria-program
