Resident Training:  The Bad and the Ugly

N A R R A T I V E  M A T T E R S

September/October 2002

Dolores



A young medical resident takes her own life, shining a harsh spotlight on the pressures of the current residency education system.



by Daniel J. Derksen



Dolores used her purple Robbins pathology textbook to weigh down the gas pedal to commit suicide in her garage. It was fourteen years ago, but I vividly remember the phone conversation with the resident on call, the bearer of the very bad news. Dolores had died of carbon monoxide poisoning, he told me. Her death came as a shock to her husband, mother, friends, colleagues, and patients. She was a compassionate, caring person—shy but well liked by patients and staff. Dolores (not her real name) left behind no suicide note but many unanswered questions.



Dolores had been barely six months into her family practice residency training. As a new family practice faculty member, I wondered how we missed the warning signs. Suicide risk factors were certainly present—new employment, isolation from family, a stressful work environment, chronic sleep deprivation, considerable student loan debt and service obligation, and a spouse also in residency training. But thousands of residents year in and year out endure the same training and pressures without tragic outcomes. At the time—the situation has since eased somewhat—the stress and long hours of residency were a rite of passage, a boot camp that once completed assured privileged membership in the medical profession.



Wondering whether other residents in our institution’s training programs shared Dolores’s risk profile, I conducted a survey of all first-year residents. A high response rate (85 percent) showed that residents were keenly interested in the subject. The survey revealed that trainees working more than eighty hours a week were much more likely to have suicidal thoughts (22 percent) and be depressed (47 percent) than those working fewer than eighty hours (9 percent and 27 percent, respectively).



Hoping to prevent another tragedy, a few months later I presented the data at a hospital executive committee meeting. Several department chairs asked who would pick up the slack of patient duty if residents’ work hours were reduced. The hospital CEO questioned the data’s validity and commissioned the Arthur Andersen consulting firm to do another study—one that deleted annoying-to-management questions about depression and suicidal tendencies. Naïvely I had believed that Dolores’s death and the compelling data would catalyze change. I had much to learn about hospital inertia and economic barriers to reform.



Overwhelming Strains

My interest in changing the
physician training system was spawned by my own and my wife Krista’s prolonged
internship and residency experiences, Dolores’s death, and my institution’s
resistance to reform. I didn’t know Dolores well but shared some of the same
stresses of residency training. For example, when she was a fourth-year medical
student, Dolores had not “matched” in family practice—meaning that through a
computerized lottery process, an available residency position was not found for
her specialty. She therefore followed her spouse to New Mexico and hoped for the best. In our
fourth year of medical school in 1984, Krista and I unsuccessfully competed in
the couples’ match program. Unmatched, we scrambled to find programs that
hadn’t yet filled all of their training slots. Rather than residing in
different states to pursue our preferred specialty residencies (hers in OB/GYN,
mine in family practice), we chose to accept internal medicine internships at
the same hospital.



To train in our desired specialties, in 1985 Krista and I went outside the
computerized lottery system and obtained positions on our own. This move
required a second internship. At that time, interns’ work hours exceeded ninety
per week; we were granted one day off a month and two weeks of vacation a year.
Like Dolores and many other two-physician couples, we had large debts. We’d
delayed purchase of a home and car and had to start paying off loans during
residency. It was exhausting to work long hours and a struggle to scrape by,
paycheck to paycheck. We didn’t have a day off together for the first six
months.



On a rare free evening together during our second internship, Krista shared
some shocking news. On her way home from the hospital that day she had
considered slamming the car at high speed into a tree. She was exhausted and
depressed and felt that she was a bad mother to our then one-year-old daughter,
Shannon.



We discussed the “microsleeps” that we experienced
when driving home after thirty-six-hour shifts without sleep. We talked about
how much longer it took us to read electrocardiograms and interpret fetal
monitor strips when we were so sleep-deprived. It was like swimming in
molasses.



One night on call, a nurse paged me to the ICU to replace a partially
obstructed endotracheal tube in a patient on a
ventilator. It was 2 a.m., and I had been soundly sleeping in the call room.
Dutifully I tried to replace the old tube with a new one while the nurse
manually oxygenated the patient and talked me through the procedure. But in my
muddled state I could not thread a new tube over the metal guide. Fortunately,
a second nurse’s frantic page brought help from a senior resident, saving the
patient an emergency tracheotomy.



A Different Approach
Needed

I have witnessed
disciplinary actions levied by residency programs against intoxicated or
substance-abusing residents, presumably to prevent medical errors. Yet
hospitals permit, even encourage, sleep deprivation
and long shifts for residents, despite growing evidence that these also are
harmful to both residents and patients. When investigating errors, hospitals do
not usually collect data on physicians’ or nurses’ work hours and shift
lengths. Yet it seems logical that overwork and sleep deprivation can easily
contribute to medical error.



Recently, the Occupational Safety and Health Administration (OSHA), the
American Medical Student Association, and the Committee of Interns and
Residents advocated federal regulation of residents’ work hours. The
regulations limit these hours to eighty per week, provide at least one
twenty-four-hour
off-duty period each week, and curtail shifts to a maximum of twenty-four
hours. (The Accreditation Council for Graduate Medical
Education, or ACGME, instituted voluntary “work duty” guidelines in 1988, the
year Dolores died.) In June 2002, Senator Jon Corzine
(D-NJ) introduced legislation to make regulation of residents’ work hours a
condition of a hospital’s Medicare participation. The legislation coincided
with ACGME’s report that recommended new limits on
residents’ duty hours—limits that were approved by ACGME’s
board and will take effect in July 2003.



These calls for change are needed. But despite my residency experience, I am
ambivalent about some aspects of reform. The rigors of residency training build
a solid foundation by combining critical medical knowledge with practical
clinical experience. Following a seriously ill patient through a twenty-four-
or thirty-six-hour course is a time-honored educational opportunity that would
be lost with the proposed regulations. The intense training experience tests a
resident’s mettle, resolve, and intellect and teaches self-reliance and
independent problem solving.



I remember at the beginning of a thirty-six-hour ICU shift admitting a
sixty-year-old woman with congestive heart failure due to severe mitral valve disease. Her condition quickly deteriorated,
and she was rushed to surgery, where her mitral valve
was successfully replaced. To admit the patient, follow her through surgery,
and leave the next evening knowing that we had saved her life was intensely
educational and gratifying.



Another factor in my ambivalence about completely changing the status quo is
residents’ substantial contribution to physicians’ workforce capacity.
Decreasing residents’ work hours could reduce access and services for poor and
vulnerable populations served by this sector of the safety net.



But there is a price to pay for taking no action. When a resident commits
suicide or when disturbing trends of increased substance abuse, suicide,
marital discord, and depression are identified in the medical profession,
reform is imperative—for the safety of both residents and patients.



A career in medicine is like a long-distance run—the pace must be measured and
planned over a thirty- to fifty-year haul. Recent data suggest that during
residency many physicians learn work habits that injure and impair them, result
in premature death or retirement, or destroy marriages and other relationships
at alarmingly high rates. Too much is crammed into residency training, while
too little is invested in lifelong learning. Residency is not like staying up
all night studying for a medical school exam. Doing that may have been enough
to pass the exam, but the information is quickly forgotten. It may once have
been possible to learn everything during residency training. But the
exponential growth of medical knowledge and technology requires a radically
different approach to physician education.



For example, physicians could supplement traditional lecture-based continuing
medical education (CME) requirements with practical “mini-residencies” to
enhance clinical skills and use new technologies. At our institution, field
faculty precept medical students and residents, who, in turn, keep these rural
educators informed about what is available online through the center’s library.
Some of these preceptors have learned clinical skills such as colposcopy through these mini-residencies, or attended
through the department’s inpatient teaching services to keep their hospital
skills sharp.



After surviving my own protracted residency training and that of my wife, and
after training many residents and witnessing trainees’ impairment and death, I
still believe that there are ways to reform residency training without
eliminating important teaching moments or reducing safety-net capacity.

What Works

As educators, we should
better identify residents who are struggling and provide effective support,
intervention, counseling, and mentoring. There are signs to watch for: poor
evaluations, scarce attendance at resident meetings, disorganized medical
records or oral presentations of patient cases, failure to answer pages,
lengthy disappearances during the day, and concerns expressed by other
residents. I remember one resident who exhibited several of these signs; it
turned out that fellow residents suspected possible drug abuse. An intervention
was arranged, and the resident received appropriate treatment.



We should consider adapting or codifying proposed regulations of resident work
hours. While the old, voluntary ACGME guidelines were ineffective, the newly
proposed regulations may be too prescriptive. For example, a thirty-six-hour
shift without sleep might be allowed, as long as it is followed by an enforced
rest period. Being in house in a nice call room while answering a few pages and
sleeping for six or seven hours (a relaxed scenario that occurs with some
specialties at certain nonbusy times of the year) is
not the same as working all night delivering babies. Yet the proposed
regulations would simply count duty hours and not allow flexibility across
specialties.



The larger problem, of course, comes down to money. It will be difficult to
wean our hospitals and training programs off a cheap, overworked resident
workforce. Hospitals receive from the federal government’s Centers for Medicare
and Medicaid Services about $70,000 a year per resident. Residents’ stipends
and fringe benefits average $46,000. If residents’ work hours shrink, calls
will likely be made to extend residency training by a year, reduce stipends, or
eliminate moonlighting by residents. But with the average debt among medical
students exceeding $100,000, such measures will surely make things worse for
the potential Doloreses among young residents and
exacerbate the downward trend in medical school applications.



How to finance such reforms in the residency system is considered the biggest
sticking point, but it need not be so. Perhaps Dolores would not have chosen to
kill herself if she had received an average stipend of $37,380 for forty hours
of work a week in a fifty-week year, and if her overtime hours, defined as
forty-one to a maximum of eighty a week, could have occurred either in the
training program or by moonlighting. If her overtime had reflected fair market
value for physician services ($50–$75 per hour), she would have had the
resources to reduce her student loan obligations—easing some of her terrible
stress. Such calibration also would reduce hospitals’ temptation to exploit
residents.



Doing Good For All

Incorporating moonlighting
into a residency education program can be a win for all. As an upper-level
family practice resident, I moonlighted on weekends by providing locum tenens (practice relief) coverage for a doctor in a
community health center in Questa, New Mexico. That work provided a stress test
that taught me what I was well trained for and what I was not. (As a result, I
arranged elective rotations to enhance my electrocardiogram and suturing
skills.) It also permitted me to earn enough to pay off our credit card debt.



In 1993 I helped to create at the University of New Mexico
School of Medicine an academic locum tenens
program emphasizing practice relief in rural and medically underserved primary
care practices. These are mainly located in the twenty-nine New Mexico counties that are federally
designated as having shortages of health professionals. In April 2002 the
program had a record month, providing the equivalent of 300 days of primary
care practice relief. Demand across the state, including in the Albuquerque area, calls for more than 500 days
per month. Many, many community health centers, Indian Health Service clinics,
public and private hospitals, urgent care centers, emergency departments, and
private physicians are willing and able to pay fair market value for residents’
time. Getting to know residents also gives these institutions a crack at
recruiting them to their community after graduation. (Our locum tenens program is the nation’s largest and oldest of
its kind, but a few others exist, such as those at East Carolina University and the University of Kansas.)



Medical students, residents, and physicians in practice too often sacrifice a
healthy balance between personal and professional activities. It is not
altruistic to work more than eighty hours a week for our patients if by doing
so we poorly manage them or sacrifice relationships with our own family and
friends.



Dolores’s journey in the medical profession ended prematurely and tragically. I
hope that the latest calls for reform in the residency education system will
chart a different course that allows residents to learn at a more rational and
measured pace than current long work hours and sleep deprivation permit. The
training regimen should be more balanced, more relevant to our lifelong
vocation, and more respectful of students and resident trainees. Neither
residents’ lives nor those of their patients should be put at risk because
physicians believe that they are immune to the effects of long work hours or
sleep deprivation. They are not.



The stakes are huge. After all, any one of us or someone we love could be on
the receiving end of a medical error committed by an overworked, stressed-out
resident. And I never want to get another phone call about a resident’s
suicide.



Daniel Derksen,
Dderksen@salud.unm.edu, is an
associate professor in the Department of Family and Community Medicine at the
University of New Mexico (UNM) School of Medicine. He is director of the UNM Center for Community Partnerships and
principal investigator for the W.K. Kellogg Foundation's Community Voices inititative to improve health care services and access for
uninsured and underserved populations in eight New Mexico countries.

Hot Spots 2003 Faculty