rising five times faster than the inflation rate.
Overlooking the Biggest Determinants of Health
The U.S. health care system is primarily devoted to the treatment of diseases after they occur. That has led to the creation of a health care delivery system that often provides leading‐edge treatments for a range of conditions and diseases. But medical care directed toward treating disease plays only a small role in determining our health outcomes. Much more important are behavioral, environmental, and social factors. Those factors are largely overlooked within the U.S. health care system—something I explore in detail in the next chapter.
Physician Burnout
There’s one other key contributor to dysfunction in the delivery of health care: physicians. More specifically, physicians who are burned out. That was the conclusion of a 2018 paper by Tait Shanafelt, director of the WellMD Center at Stanford, and Daniel Tawfik, an instructor in pediatric critical care medicine at Stanford. As part of the study, they sent surveys to physicians throughout the United States. Nearly 6,700 responded—and 10 percent declared that they had been responsible for at least one major medical error in just the previous three months. The survey also revealed that physicians who reported burnout were more than twice as likely to commit a medical error than physicians who were not burned out [29]. Those errors included mistakes in medical judgment, mistakes in diagnosing illnesses, and technical mistakes during procedures.
The frequency of mistakes is even more worrisome given the prevalence of physician burnout. The same study revealed that 55 percent of physicians were experiencing symptoms of burnout, which can stem from many different factors. While we need to be focused on the well‐being of all health care professionals, a 2017 article coauthored by Bryan Bohman, a Stanford professor and founder of our physician wellness committee, highlighted a few reasons to be focused on physicians:
First, physicians have been hard‐hit by the organizational transformation of the health care system, resulting in an epidemic of burnout and declining professional fulfillment. They have suffered a reduction in their sense of professional autonomy, have experienced a significant increase in clerical duties, and are beholden to a growing array of imperfect and inconsistent quality and productivity metrics. Second, medical training has historically acculturated physicians to deny their own self‐care in the service of others [30].
Another contributor to burnout is the documentation demands associated with electronic health records. Studies show that physicians devote 34 to 55 percent of their workday to EHR‐related tasks. Steven Lin, the medical director at Stanford Family Medicine, points out that while some of this time bolsters ongoing care, “much of it instead serves billing documentation, defense against litigation risk, and regulatory compliance.”
I don’t think many people become doctors because they love doing documentation. While documentation is part of the medical school curriculum at Stanford, for many aspiring physicians it’s something they must figure out as they go through their clerkships and residency training. Then, when they start to practice, the burdens are likely to grow: EHR record‐keeping, inbox management, digital messages from patients, managing providers, etc. “It’s a huge shock to a lot of the younger trainees, as well as older, experienced physicians,” says Lin. “Across the age and experience spectrum, many doctors are just deciding not to go to EHRs and risk the penalties surrounding meaningful use, or to leave medicine altogether. It’s a big problem.” Indeed, a 2017 study commissioned by the Association of American Medical Colleges predicted that by 2030, the United States will face a shortage of between 40,800 and 104,900 physicians [31].
Studies show that physicians who use medical assistants to act as digital scribes and record the content of the patient‐doctor interaction show far more satisfaction and lower rates of burnout. University of Colorado Health experimented with increasing the ratio of medical assistants (MAs) to physicians, from 0.4 MA per physician to nearly two. Before the physician enters the room, an MA has spent 20 minutes talking with the patient, updating the medical records, and handling minor medical issues, such as vaccines and screenings. When the physician walks in, the MA stays in the room, acting as a scribe during the exam. They found that over the course of a year, this approach went a long way toward relieving physician burnout: the metric they use to measure physician burnout declined from 55 percent to 14 percent. Assigning two people for each physician to act as scribe may not be a cost‐effective solution, however.
AI researchers are working on automating the job of the scribe. Google and Stanford Medicine have been collaborating on a digital scribe project that would listen to the dialogue in a patient visit and take notes. The idea is not merely to take a transcription, but rather to knit the dialogue into a narrative. In the study, each doctor wears a microphone to capture conversations with patients, which are used to train machine‐learning algorithms in getting the gist of a doctor‐patient interaction. The goal is to train the algorithm to generate a pithy progress note. Google researchers say that its scribe can capture complex conversations typical of a patient‐doctor conference even when family members and other practitioners are present in a noisy environment.
Scribes—whether live or automated—can play a valuable role and hopefully help remedy doctor burnout. In the meantime, there are significant expenses for health systems when physicians leave medicine. The process of identifying and recruiting replacements can involve costs from $268,000 to $957,000, according to a study published in 2017 by Stanford’s Maryam Hamidi [32].
I see the manifestations and the consequences of physician burnout. From the colleagues who leave the medical profession at relatively early stages of their career to colleagues whose personal and professional lives are disrupted by stress and anger, the manifestations of burnout are hauntingly tragic. Shortly after I arrived at Stanford in the fall of 2012, a group of clinical faculty in the School of Medicine came to see me to discuss the problem of physician burnout within our own academic medical center and nationally. These faculty wanted to do something about burnout, and they wanted Stanford Medicine to take the lead in proactively and constructively addressing the problem.
To help address these issues, in 2017 Stanford’s School of Medicine hired a chief physician wellness officer, Tait Shanafelt, making us one of the first academic medical centers in the United States to hire someone for such a position. As mentioned above, he is the director of the Stanford Medicine WellMD Center, which is focused on improving the health and professional fulfillment of physicians.
Shanafelt’s interest in the issue stems from observations during his residency at the University of Washington, when he saw the extraordinary demands being placed on interns he was supervising. He subsequently helped design and lead a study to examine the nexus between burnout and quality of care. The study, which was published in the Annals of Internal Medicine in 2002, revealed a close response relationship between burnout and suboptimal patient care: the higher the burnout score, the greater the frequency of residents reporting errors or providing suboptimal care to patients. Among residents with a high burnout score, 53 percent indicated that they had provided sub‐optimal care at least once in the previous month, a rate markedly higher than those without high burnout scores [33]. Shanafelt also spent over a decade at the highly respected Mayo Clinic in Minnesota, where he conducted studies of physician well‐being.
Today, as the head of the WellMD Center, Shanafelt is focusing on several issues. One is building safety nets for physicians who are in distress. Another is helping physicians and leaders create a culture where physicians support each other and create a practice environment that makes it easy for physicians to provide patients the care they need. He is also creating operational metrics that define whether the practice environment is a source of frustration or facilitates efficient delivery of the care patients need and promotes professional fulfillment. As Shanafelt has remarked,
It’s amazing to me that administrative leaders can create spreadsheets with detail to three decimal places for metrics such as how many patients are seen per room per day, how often a clinic runs late, and other operational aspects of the practice environment. But they rarely have categories for things like the consistency of an operating room team working together or how many minutes per night primary care doctors