How Deep the BiasHow Deep the BiasHow Deep the Bias

During a recent radiation oncology consultation, I saw a patient who had just undergone lumpectomy for breast cancer. The patient had traveled to have her surgery at the academic medical center at which I am a faculty member, but she wished to have her adjuvant radiotherapy, which would entail several weeks of daily treatment, closer to her home in another state. She asked me to recommend someone, so I turned to my professional society's membership directory to determine whether I knew anyone in her area. I found the names of two radiation oncologists, both unfamiliar to me.

I then did something that I, in retrospect, simply cannot excuse. I noticed that one physician listed his training in the 1970s in India; the other failed to list where he had trained. I shared this information with the patient, noting that I would generally favor a physician trained in the United States. This was a preference quietly inculcated by many of the respected faculty who trained me—faculty who did not seem racist in any way and who, among other things, were highly supportive of me, a US-trained physician of Indian descent. The concerns they expressed when faced with similar situations really did seem to relate to genuine doubts regarding the quality of training in some other countries and the quality of the care provided by physicians so trained. However, as soon as the words left my lips, I regretted them.

Before I could take them back, the patient thanked me and said that she “would have chosen the ‘American’ doctor as well,” noting that “Foreign doctors often have trouble communicating.” I backpedaled as well as I could, noting that the directory indicated that the Indian-trained physician had been practicing in the United States for several decades and that it provided no information regarding where the other physician had trained, so it was probably an even bet either way. But the damage was done. In my attempts to do my best for the patient in front of me, I had engaged in and endorsed a form of profiling, even though I have long believed profiling to be a morally unacceptable means, regardless of the ends it is intended to serve.

I wonder how often this very sort of profiling leads patients to request an appointment with one of my partners rather than with me. Although I trained in this country, my name belies my Asian ancestry, and I suspect that race and ethnic origin may actually be even more important to patients than where a physician trained. Patients with serious illness may naturally seek health care professionals with whom they share certain commonalities of experience. Indeed, studies have shown that minority patients may prefer to have physicians of their same race or ethnicity. For similar reasons, one might reason that some members of the majority group may also prefer physicians of the same racial or ethnic background. Of course, one can make the argument that those of the minority group have faced unique issues and situations related to their membership in that group and that their special vulnerability justifies their preferences. Yet when choosing a health care professional, where all patients have a certain level of vulnerability, I am not convinced that this argument fails to apply to all patients. And yet if it does, it is disquieting.

When placed in a situation of uncertainty, I suspect that a nontrivial proportion of patients consider the few factors that are easily discerned from a clinician's name alone: gender and ethnicity. The ways in which these factors influence patients' choices of physicians are generally opaque to the physicians themselves. In rare cases, however, one may catch a glimpse of how deep such biases may run.

When I was a resident, the attending physician with whom I was paired was scheduled to see an elderly woman with metastatic breast cancer in reconsultation regarding a new, painful bone metastasis. As was our clinic's procedure, I was to see the patient first. I smiled and entered the room with my standard greeting, “Hi, my name is Reshma Jagsi. I’m the resident physician working with Dr P, and he has asked me to see you first to get the details of your history before he comes in to talk with you.” She quickly interrupted: “Oh, no, dear. I can't be seen by you. I can't understand people with thick accents.” I smiled to hide my surprise and apologized, noting in a louder and clearer (and still entirely accent-free) voice that while I had thought I had lost my Texan accent many years ago, I do have a terrible tendency to speak too quickly and sometimes too quietly. She then noted, “Oh, no, dear. You do have a very thick accent. I could tell the moment I saw you.”

When I left to explain the situation to Dr P, he could do nothing but laugh: Dr P has the thickest British accent I have ever encountered. He noted that he had cared for the patient in the past, and she had never complained about his accent. Furthermore, as Dr P flipped through his old dictations, he remembered that she had simply adored the tall, blond resident with whom he had treated her for her initial radiation course. That resident was born and raised in Germany, attended medical school there, and only relocated recently to the United States for his residency. He had an even thicker accent than Dr P. Dr P sent me on another follow-up case instead and finished with the elderly woman on his own. Two hours later, the white South African and Australian technicians (immigrants to the States) manning the radiation simulator, after hearing the anecdote, noted the irony of the fact that the patient had voiced no objections when they brought her in for her treatment planning scans, given their own heavy accents. Indeed, the only health professional lacking a “foreign” accent whom the patient had seen all day was the one she had turned away.

This sort of deep-seated bias rarely reaches physicians like me who practice in a largely outpatient setting; most patients have the opportunity to self-select to see those with whom they feel comfortable. Nevertheless, race has been a prominent factor in a fair number of the encounters I have had with new patients. With almost surprising frequency, patients make comments that reveal their failure to recognize that one can be an American even when one is of non-European ancestry and the misguided perception that one may offer a compliment by making a racial generalization: “You know, your people have really found a great niche in medicine” and “You're lucky you're not American. You’ll probably never show your age.”

Sadly, I have come to expect comments like these in my daily practice. I treat patients with cancer, often only recently diagnosed, and it is understandable that they do not always manage to censor themselves as well as they might under other circumstances. I appreciate that it is only natural for these patients—who face a serious diagnosis and need to consult with multiple new clinicians, often in a new and unfamiliar hospital setting—to yearn to forge a personal connection. But I do find it troubling that something about my name and skin color seems to make some patients feel they must focus on my ethnicity in order to forge that bond. I was raised in the United States and am likely to have learned in the same school systems, enjoyed the same music, and cooked the same foods as my patients. Comments that in other situations would be nothing less than offensive take on an air of sadness: sadness because some people still cannot see past race as a defining characteristic, sadness because these comments are themselves signs that a patient must be yearning to bond with her physician, sadness because the patient-physician relationship must face obstacles over which I have little control.

As a new faculty member in a department with several residents from ethnic minority groups, I am now compelled not only to serve my patients but also to be an appropriate role model to my trainees. Recently, I heard one of our Asian American residents struggling to answer a patient's question: “So where are you from?” (His answer—Troy, a town just outside Ann Arbor, Michigan, failed to satisfy, and he was prompted, as I knew he would be, with the vaguely inappropriate follow-up question, “No, but where are you really from?” that I have come to expect if I answer that same question by saying that I grew up in Texas or that I recently relocated from Boston.) And I was left speechless as a mixed-race mentee noted that she often faces the question, “So what are you, anyway?” and must fight the urge to respond, “A human being.” As I watch my residents struggling with framing responses appropriate to these questions and comments that are a quiet thorn in my own side, I have really begun to wonder what the best approach is.

For my own part, I will do my best never again to refer a patient to another physician based on race, ethnicity, or country of training. This sort of profiling is a morally unacceptable means to any end because it fails to respect the fundamental human dignity that we must accord every individual. While I may have learned that profiling was acceptable by the example of certain persons with whom I trained, I am now committed not to pass that attitude on to my own residents.

I will try gently to correct obviously incorrect statements made by my patients, such as ones mistaking my citizenship, as my trainees look on with their keen eyes. Yet in most cases of inappropriate but benignly intended comments, I will continue simply to move on and later explain to my residents that I truly believe that these are signs of a patient's desire to connect, rather than comments worthy of indignation.

One day, I hope our society can move beyond seeing race and ethnicity as such defining characteristics and embrace the commonality of human experience that binds us together. In any case, I will attempt to be the best role model I can be for my trainees and can only hope that others will do the same.