- State how frequently physicians care for family or relatives.
- Describe the risks associated with caring for family members or relatives.
- Appreciate some of the steps hospitals take to prevent identification errors.
(Submitted by a respected senior physician at an academic medical center.)
My healthy and active 13-year-old daughter had a syncopal attack—she just passed out. This was the first time she had ever been sick. Her blood pressure and pulse were normal right after the event and she recovered quickly. I was a bit shaken up and because I was not sure of the cause, I took her to see her pediatrician. The pediatrician felt it was probably dehydration (my daughter is an athlete), but wanted an electrocardiogram (ECG). The ECG was performed, and as we were leaving the office I asked the front desk clerk for a copy of the ECG. I think the clerk recognized me as a physician on staff and handed it to me. As I walked away, my heart nearly stopped. The ECG was not normal. My daughter's ECG was not normal. The rate was 44 beats per minute, and the tracing met criteria for left ventricular hypertrophy.
My mind raced, filled with the worst diseases I could imagine. Syncope combined with an abnormal ECG is never a good combination. I immediately paged a pediatric cardiologist colleague. He responded by phone that the heart rate was too low even for an athletic child and she should get an echocardiogram and Holter monitoring. I panicked even further and couldn't get the terrifying vision of my daughter frail, sick, and dying in a hospital bed out of my head.
It is probably safe to say that every physician in the world has, at some point, provided a close relative with medical care, be it an off-the-cuff opinion or a long-term clinical commitment. That assumption makes this practice the single most common ethical violation committed in medicine, for most advisory bodies specifically warn against it.
The ban on caring for family members is longstanding. At the turn of the 19th century, in what is considered the first modern code of medical ethics, British physician Sir Thomas Percival wrote that doctors should depend on their colleagues to care for sick relatives, for "solicitude obscures the judgment."(1) In 1847, the founders of the newborn American Medical Association (AMA) elaborated: "... the natural anxiety and solicitude which [a physician] experiences at the sickness of a wife, a child, or any one who by the ties of consanguinity is rendered peculiarly dear to him, tend to obscure his judgment, and produce timidity and irresolution in his practice."(2)
The sentiment endures in the AMA code's most recent iteration ("Physicians generally should not treat themselves or members of their immediate families. Professional objectivity may be compromised....").(3) Medicare will not pay for services physicians provide to immediate relatives (presumably this decision reflects a fear of fraudulent claims as much as a purely ethical deliberation).(4)
Nonetheless, the practice is almost universal. A 1991 survey of doctors staffing a Midwestern community hospital found that 97% of respondents had provided a medical service for a family member, ranging from free drug samples (72%) to written prescriptions (80%) to surgery both elective (9%) and emergent (4%).(5) At another midwestern hospital 74% of physicians reported treating their own children for minor acute illness.(6) If the child had a fever, most called the child's actual pediatrician, but parents who were pediatricians were likely to treat both afebrile and febrile children themselves. The care these doctor/parents provided included otoscopy, chest auscultation, and antibiotic prescribing. Most cited "convenience" as the major reason for their care, as well as confidence in their own diagnostic skills; few were motivated by cost savings or privacy issues.
The majority of these in-family medical encounters probably end uneventfully. There is no data to characterize the ones that do not, but anecdotes abound. More than a third of surveyed physicians stated they had observed a colleague become "inappropriately involved" in a family member's care by vetoing indicated procedures or demanding unnecessary or contraindicated ones.(5) Some were guilty of the opposite, providing care that was cursory or inadequate. In a small survey of pediatricians in a single Vermont community, most respondents agreed that doctors' children came to the office less often for acute care than other children did, and often later than was optimal.(7) They noted that sometimes the children of physicians had only cursory attention paid to social issues or behavioral problems.
The list of other potential risks from keeping medical care in the family is a long one.(8) Some are primarily medical issues: for instance, doctors are unlikely to perform a full, intimate physical exam on a relative. One physician recalls performing a "half-hearted" abdominal exam on his wife when she complained of pain shortly after suffering an apparent miscarriage, coming up with no diagnosis, and feeling "overwhelmed and resentful." She turned out to have a large tubal pregnancy.(8) Similarly, doctors are unlikely to take a full social history from a relative, particularly from their own offspring, and pertinent but difficult questions about sexual activities or substance use may go unasked. They may try to spare a relative from undergoing painful procedures, even if indicated. They may leap to diagnoses, either in their own field of expertise or far afield. Once involved in a case, they may have difficult time backing off and admitting it is time for a consultant to step in.
The social risks from the arrangement are equally perilous: the dynamics of even the best-adjusted family are apt to be irrevocably changed if a doctor–patient relationship is superimposed for too long on that of husband and wife or parent and child.
But the doctor/father in this case was guilty of none of the above sins. He was not trying to be his child's doctor; he was simply trying to ease their passage through the medical system with a shortcut around that unbearable wait between the performance of a test and the delivery of the result. What could be wrong with that?
I called my daughter's pediatrician to ask for referrals and another appointment to see her later that day, telling the front desk that I was both a physician and the parent. The pediatrician hadn't seen the ECG but she approved the referral, and we were able to get an appointment for the echocardiogram the same day (likely due to my position on staff). After the test, we returned to see the pediatrician. I breathed an incredible sigh of relief as the echocardiogram turned out normal. The pediatrician reexamined my daughter and found her heart, lungs, and complete examination normal. We decided not to proceed with the Holter monitoring, and to allow my daughter to play soccer the next day.
The following week, I stopped by to see the pediatric cardiologist with whom I had spoken and showed him the ECG. He read it and felt that the slow heart rate and the other changes were actually normal for a healthy athletic 13-year-old. I was now feeling much better about everything.
He scanned the tracing into the medical record and e-mailed me a PDF. I was speaking with my wife later in the day and she wanted to see the ECG, as she had not seen it yet (she is not a physician). I e-mailed her a copy and she called me a minute later. "This is not her ECG. It's an ECG from some 24-year-old male." I opened the file and sure enough—it was the ECG I looked at, but it was not my daughter's. It was not her ECG at all. It would have been obvious to me had I looked at the name and noticed it was wrong. All of this was a mistake.
Editorial Note: The physician (case submitter) further reflected on what he learned from this experience. His thoughts:
"Clearly, as a physician, I made an error in not making sure the tracing I was handed was from the patient in focus. Secondly, I used my physician status to force decisions along a path. I didn't give time for the pediatrician to even interpret the real ECG; she took my word that it was abnormal. The cardiologist took my word, to the point of scanning the document, interpreting it, and entering it into the medical record. They believed what I told them. The echo was done prior to the cardiologist looking at the ECG.
My recommendations are as follows: The root cause of this error was getting the wrong ECG in the first place. The front desk worker at the cardiac center simply printed out the wrong tracing and handed it to me. It turns out that our hospital has rules against this. I was thinking at the time the ECG was handed to me that at my hospital a patient cannot request medical information from any other source than the medical records department. I was thinking this because I had been made aware the week prior that it was considered a patient privacy violation to send a copy of a patient evaluation to the patient themselves. This is something I have done for years so the patient can bring my evaluation to their various doctor visits and improve the communication of information. That is no longer allowed at my hospital, and now I understand for good reason. Medical records personnel do consider patient identification very important, and therefore likely do not make this error as often. If this practice had been followed, the clerk at the cardiology desk would have said to me, 'Sir, I cannot do that. But you can go to medical records and get a copy there.' Had I done it the right way, I am sure I would have gotten the correct information."
This case illustrates that even a small step over the line between parent and doctor can have, if not catastrophic, certainly sobering consequences. What exactly happened here? A parent-turned-doctor cast a professional eye over a test result that made little clinical sense. Presumably he, like the rest of us, sees similar reports fairly often in his professional life. But instead of running through the usual possibilities (Did somebody mislabel the blood tubes? Is something screwy going on in the lab? Am I looking at the wrong chart?), his mind began to gallop in circles like a crazed steer. Call a consultant, get a test, double back around to another consultant, cancel another test, and stop, panting, when everything begins to evaporate (as nonexistent health emergencies often do).
Of course, had he been a layperson, like his wife, his eye would have moved immediately from the indecipherable squiggles of the cardiogram to the name on the tracing and the entire episode would have been over before it began.
The case highlights some of the safety issues we know well in the modern digitalized hospital, where a single errant stroke on a keyboard can cause chaos. Back in the old days, paper charts often took on oddly human personalities, the thin and crisp folders of the healthy easily distinguishable from the huge, dog-eared volumes of the sick. Identity mixups still occurred, but probably somewhat less frequently than they do now, for in Times Roman Bold on an LCD screen every patient's record looks alike.
Hospitals are deploying computerized tools to cope with this landscape. The scannable patient wristband, for instance, reduces identity errors during the administration of medications.(9) Out in the nurses' station, however, at least in my hospital, it remains quite easy for a harried resident to order Mr. X's methadone for Mr. Y, and Mr. A's Coumadin for Mr. Z. (I was involved in both of these events. The Coumadin error was caught downstream by an alert nurse; the methadone error, unfortunately, was not.) Wristband systems, at least in their present incarnation, do not avert this kind of identity error.
The case highlights privacy issues as well. It is precisely to forestall test results being confused and accidentally revealed to third parties that some hospitals specify that only the medical records department may provide patients with any portion of their record, including individual test results. This policy was actually in effect in the author's own hospital, he writes, "and now I understand for good reason." Had he gone about obtaining the test "the right way," he continues, "I am sure I would have gotten the correct information."
These are the smaller lessons from this case. The big lesson simply reinforces what doctors knew centuries ago: When you love a patient too well you generally cannot take good care of that patient. Interestingly, this piece of wisdom flies directly in the face of one of the other tenets of modern medical practice, the oft-quoted maxim of Harvard's Dr. Francis W. Peabody that "the secret of the care of the patient is in caring for the patient."(10)
How to interpret this apparent contradiction? A philosopher could write volumes on the subject, but the brief version is that either extreme of affection risks destroying the immensely delicate fabric of medical care. Too little love is bad, but too much can be worse. The doctor who manages to balance in between—for a single patient, or for a practice of thousands—has skills worth calibrating and emulating.
- Becoming professionally involved in a family member's medical care is risky.
- Taking short-cuts around a hospital's established policies is risky.
- When confronting bad news, whether for a patient or a relative, first make sure of the facts.
Abigail Zuger, MD
Associate Professor of Clinical Medicine
Columbia University Medical Center
St. Luke's-Roosevelt Hospital Center
Faculty Disclosure: Dr. Zuger has declared that neither she, nor any immediate member of her family, has a financial arrangement or other relationship with the manufacturers of any commercial products discussed in this continuing medical education activity. In addition, the commentary does not include information regarding investigational or off-label use of pharmaceutical products or medical devices.
1. Leake CD. Percival's Code: a Chapter in the Historical Development of Medical Ethics. Chicago, IL: American Medical Association; 1923. [Available at]
2. Proceedings of the National Medical Conventions, held in New York, May 1846 and in Philadelphia, May 1847. Philadelphia, PA: TK and PG Collins; 1847. [Available at]
3. Opinion 8.19: Self-Treatment or Treatment of Immediate Family Members. Chicago, IL: AMA Code of Medical Ethics; 1993. [Available at]
4. US Department of Health and Human Services, Centers for Medicare and Medicaid Services. Code of Federal Regulations. Title 42: Public Health. Washington, DC: Federal Register; 2004. [Available at]
5. La Puma J, Stocking CB, La Voie D, Darling CA. When physicians treat members of their own families: practices in a community hospital. N Engl J Med. 1991;325:1290-1294.
[go to PubMed]
6. Dusdieker LN, Murph JR, Murph WE, Dungy CI. Physicians treating their own children. Am J Dis Child. 1993;147:146-149. [go to PubMed]
7. Wasserman RC, Hassuk BM, Young PC, Land ML. Health care of physicians' children. Pediatrics. 1989;83:319-322. [go to PubMed]
8. Fromme EK, Farber NJ, Babbott SF, Pickett ME, Beasley BW. What do you do when your loved one is ill? The line between physician and family member. Ann Intern Med. 2008;149:825-829. [go to PubMed]
9. Poon EG, Keohane CA, Yoon CS, et al. Effect of bar-code technology on the safety of medication administration. N Engl J Med. 2010;362:1698-1707. [go to PubMed]
10. Peabody FW. The care of the patient. JAMA. 1927;88:877-882. [Available at]