Secondary Logo

Journal Logo

Physician–Patient Relationship

Life on the Receiving End: A Qualitative Analysis of Health Providers’ Illness Narratives

Kempainen, Robert R. MD; Bartels, Dianne M. RN, MA, PhD; Veach, Patricia McCarthy PhD

Author Information
doi: 10.1097/ACM.0b013e31802d9513

Abstract

Patients desire care that is both technically sound and empathic, and clinical empathy is integral to forging therapeutic patient–physician relationships.1,2 Patients’ and family members’ illness narratives can be a rich source of insight into what it feels like to be a patient. Although such narratives routinely appear in medical journals and the popular media,3 they typically have not been subjected to formal qualitative analysis. Qualitative research methods are well suited to explore complex interactions among patients, family members, and the health care system, as they are designed to interpret phenomena in light of the meanings people bring to them.4

We performed a qualitative analysis of essays from the Annals of Internal Medicine’s “On Being a Patient” series to develop an explicit framework for understanding “what it feels like” to be a health care consumer and to explore how provider behaviors influence that experience. Essays in that series provide concise, yet powerful, descriptions of pivotal moments in patients’ and family members’ medical encounters. The majority of the essayists are health care professionals whose experiences afford them unique insight into patient, family member, and caregiver dynamics.

Method

Data collection.

We reviewed 31 consecutive essays published in the “On Being a Patient” series between January 1, 1999 and December 31, 2003 in the Annals of Internal Medicine.5–35 Seven were excluded: two poems,12,13 two essays without direct reference to medical care,23,29 and three essays not written from a patient’s or family member’s perspective.24,25,31

Data analysis.

We analyzed the remaining 24 essays using a modified form of consensual qualitative research,36 a methodology based on principles of grounded theory.37 Researchers use grounded theory to derive a conceptual framework to explain a phenomenon from continuous, iterative analyses of data. Core ideas, expressed in essayists’ words, were the data that we classified into domains and categories through a coding process. Domains are inductively derived topics used to group, or cluster, data. Categories are common themes reflected in the core ideas within domains across cases (or, in this study, across essays). Two of us (RRK, an attending physician at a teaching hospital; and DMB, a nurse with expertise in bioethics) first independently, and then by consensus, assigned content from five randomly selected essays into domains. We sequentially coded groups of five essays in similar fashion, continuously comparing essays and modifying derived domains. We then abstracted major categories from the domains. No new domains or major categories were identified in the latter third of the analysis. The third member of our team (PMV, a licensed psychologist and professor of educational psychology) reviewed the coding, domains, and categories. Her audit resulted in modification of terminology but yielded no additional domains or categories.

Trustworthiness.

Trustworthiness of the analysis was demonstrated via testimonial validity.38 Specifically, we contacted each essayist, who reviewed and indicated agreement with the analysis of his or her essay. Of 25 authors, three were deceased and one was anonymous. Sixteen of the remaining essayists provided feedback; their comments resulted in minor revisions of the analysis.

Results

The essayists were 15 physicians, one psychologist, two nurses, two medicine trainees, and five individuals without declared medical background. Sixteen essays were written from the patient’s perspective; 14 focused on encounters in the outpatient setting, typically in the context of a chronic disease. Family members wrote the remaining eight essays; six involved a parent’s death. On a five-point scale (1 = strongly disagree, 5 = strongly agree), the mean rating of agreement with the analysis among essayists was 4.88 (range 4.0–5.0).

The analysis yielded a total of 10 domains, each of which encompassed between two and six categories. We considered only those categories that were identified in three or more essays. Table 1 presents the definitions of the domains and their associated categories.

Table 1
Table 1:
Domains and Major Categories Associated with Each, Found in 24 Illness Narratives

Patients’ and family members’ psychosocial challenges

All but three essays referred to emotional challenges precipitated by illness. Fear, stress, and helplessness were especially prevalent among both patients and family members. Reports of sadness and emotional pain were common among family members; embarrassment and isolation were more prominent among patients. Both groups experienced disruption of family life and careers, struggled with increased uncertainty regarding their futures, and endured many hours waiting to see providers, undergo tests, and receive results. In the following excerpt, a medical student who develops a severe asthma exacerbation while on call experiences a host of negative emotions during a prolonged wait for assistance from her fellow housestaff:

Just the other night, I did not assess my second patient until 11:00. Then, I did not realize the importance of timeliness; now, I wait for my physicians with great anticipation. I need them, emotionally and physically. I feel vulnerable and alone. I am uncomfortable and distraught that I cannot breathe. When the team arrives, it is not the fear or discomfort that they address. Instead, they assault me endlessly with questions that I am too breathless to answer.10

The following excerpt illustrates how family members’ emotional reactions are similar to those of patients. A couple has just learned that their developmentally delayed 18-month-old daughter has extensive white-matter damage:

We had lived in a world of security and certainty, and somehow a horde of viral thieves crept into our life and plundered our baby’s neurons, robbing her future. My wife experienced anxiety, loss, mourning, dread, and a feeling of stepping into a total unknown.28

Patients’ and family members’ coping mechanisms

Patients’ responses to the challenges of illness varied, but courage, hope, denial of the severity of illness, resignation, and placing trust in one’s providers were the most frequently mentioned coping mechanisms. The following excerpt demonstrates coping through courage, hope, and denial, as well as the burden of fear, uncertainty, and waiting shared by patients and families. The essayist’s father is receiving hospice care for advanced cancer:

I took him to his appointments and learned what it is like to spend countless hours in waiting rooms, and how fear of the unknown can cast a shadow over all things. All in all, he was extremely brave, never complaining, although I came to realize how deeply disappointed and afraid he was … Even in his final days, it was inconceivable to my father that he wasn’t going to make it. He had bounced back from every previous challenge: the 106th Infantry in World War II, two divorces, and open heart surgery, among others.33

Essays by family members contained fewer descriptions of coping mechanisms but included hope, patient advocacy, and a focus on the logistics of care rather than the overall prognosis.

Patients’ and family members’ needs

Improved communication with providers, and between providers, was the most prevalent need of patients and family members (18 essays). Essayists noted deficits in both the quantity and quality of providers’ communication, including their ability to communicate coherently, humanely, and completely with patients, family members, and one another. Patients and families also shared a need for emotional comfort and reassurance. Control of physical symptoms was a prevalent patient need (12 essays), whereas patients’ comfort was frequently mentioned in family members’ essays. A family’s need for a patient’s comfort, as well as improved communication with and between providers, is apparent in the following example. The patient has suffered consecutive cardiac arrests, and his children are seeking a dignified transition to palliative care.

Another resuscitation would be futile, we reasoned. But where do we turn after the attending, in his only conversation with us, explains that he did not believe in “do not resuscitate” orders? Nurses interpret DNR orders as signs of hopelessness, and care will suffer, he asserted. He assured us that his nurses would “know what to do.” Why then, did the nurses ask us what they should do if he arrests?30

In another essay, a coherent explanation by a surgeon, coupled with his verbal and nonverbal reassurance, convinced the patient (the essayist) to proceed with repair of his mitral valve prolapse:

In clear, short, logical, precise terms, he argued that rejecting elective valve repair now meant urgent valve replacement in the near future, with the next chordal rupture only waiting to happen … For a period at the end of my interview, nobody in the room spoke. During this sudden and unexpected silence, Dr. Carpentier scrutinized me as if searching for a concluding statement that would help him dismiss me with grace. Then, instead of concluding with a ready-made formula, he reached across the table, took my hands in both of his, looked at me intently, and said, “I think I made it clear why the operation is necessary. Let me tell you now that if you decide to have it done, I will be there to do it for you as best I can.” The austere judge had turned friend, if only for a minute, but it was this gesture, more than anything else, that made me ready for my trial.16

Essayists also indicated a need for greater opportunity to inform providers of their symptoms and emotional state. One author argues that “‘no news’ should not automatically be viewed as ‘good news’” and calls for practitioners to actively elicit patients’ concerns.8 This type of proactive communication is exemplified in the following excerpt in which a perceptive physician discerns that the essayist’s clinic visit was prompted by more than a desire to be screened for head lice:

… she crossed the small room and sat informally on the examination table. (Now let’s talk like colleagues. You really are OK, this said to me.) And suddenly it was clear to me that I was not OK. Slowly, I began to tell the real story, not the pretext or the first level of concern. I was sad, very sad. I was struggling at work. I was alone and heartbroken, having just ended an important relationship. There was no one to help me. My children and my patients needed me to be full and overflowing. Instead, I was a well gone dry.9

Humanity

Essayists frequently discussed their own, or their family member’s, “nonmedical history,” describing how major life events, personality, important relationships, and values influenced patients’ behavior and the bereavement of families. One essayist shares the story underlying her father’s “lifelong hatred of hospitals,”18 whereas another opts for a high-risk bone marrow transplant in hopes of “living long enough to play an important role in my grandchildren’s lives.”11 In the following excerpt, the deterioration of the personality—the loss of humanity—of a man with Parkinson disease has a profound impact on the caregiver role and the grieving process of his son:

The father I knew as a child was loving, intelligent, patient, talented and humorous. He was skilled with his hands, a dental technician who took pride in his handiwork. His dexterity was manifest more sublimely in his violin playing, a lifetime love at which he was better than I knew … but in his last years it was hard to feel love for him. His increasing confusion often veered into anger, delusion, paranoia, irrational demands, and unrealistic assessments of his own capabilities … Most difficult of all was a gradual but progressive loss of the person we had known—the pain of gazing into those gray, once bright, intelligent eyes and seeing no spark at all, just a lost wandering look … Toward the end, and after his death, I struggled to regain the father who once taught me chess, baseball, and politics.15

Several essays describe instances when providers failed to recognize, or acknowledge, their patients’ humanity. This is readily apparent in an essay by a physician stricken with amyotrophic lateral sclerosis undergoing electromyography:

The effect of their [students’] presence on Dr. L.’s demeanor was immediate. Suddenly, I was the vehicle for teaching these students about motor-neuron pathology. He addressed all comments to them. I might not have been present except for the obvious need for my muscles. His enthusiasm about the array of abnormal findings was clear as he lectured the students on the “classic findings in ALS.” He positively bubbled with joy at finding even more evidence of the sickened state of my motor neurons. If I moved or said that the needle he placed in a muscle was uncomfortable, I was regarded with polite but cool irritation, as though an experimental preparation were not behaving in the expected manner.19

Perceived positive and negative characteristics of providers

The most prevalent positive qualities of providers were compassion, family/patient advocacy, and competence. Patients and families often expressed gratitude for the extra time providers spent with them despite their busy schedules. The mother of a toddler with a poorly defined connective tissue disorder thanks the pediatrician “… who, in this HMO world, gave more of her time to us than is imaginable.”17 In the following excerpt, a physician’s compassion, advocacy, and competence are apparent to the children of a dying patient:

Two days later, it was clear that our father would not wake up. His mind, the basis of his brilliance, creativity, and curiosity, had been silenced … But how do we negotiate a dignified withdrawal of care with an attending who could not communicate with us? We were helped by a physician who cared. Once our professor, now our friend, he was never our father’s doctor, yet he became his strongest advocate. With gentle expertise, this kind healer guided us, especially our mother, through this most difficult process.30

The negative characteristics of providers were generally the antithesis of the aforementioned positives, namely, incompetence and lack of compassion, availability, and advocacy. One essayist recalls multiple occasions when his father’s physicians provided technically sound care, yet failed to convey empathy:

When I was seven, I recall being terrified when he [the essayist’s father] fell out of a tree and fractured his spine trying to put a tire swing up for me. I didn’t realize that a simple compression fracture was not life-threatening. All I knew was that I had never seen my father in so much pain. I was scared and asked the doctor in the emergency room if my dad was going to live. After hesitating for what seemed like hours, he responded with an unreassuring “probably,” then quickly left the room.21

In many instances, a perceived lack of compassion is linked to a lack of access to providers. One essayist, who has dealt with decades of severe arthritis, comments on

the physicians for whom time is such a precious commodity that they start looking at their wrist watches just minutes into the history-taking to signal their impatience.20

Although essayists primarily describe care that is neither grossly negligent nor extraordinarily heroic, they frequently make reference to issues of both competence and incompetence. Many note providers’ efficiency and skill and express gratitude for the care they have received (nine essays). However, others note examples of poor care and express frustration when providers cannot agree on a diagnosis or treatment plan.

Dual role

Sixteen of the 19 essays by health care professionals refer to the unique and often challenging situation of being both a consumer and provider of health care. They variously discuss role reversals (shifting from provider to consumer), role enhancement (one role informs the other), and role conflicts (blurred boundaries between roles). Frequently, the transition to “caregiver as consumer” is accompanied by shock and disbelief, as in the case of this bone marrow transplant recipient:

When I looked up at my intravenous lines—two going constantly, with smaller bags awaiting their turn—I couldn’t believe that I was at the other end. I thought of the sick patients in the intensive care unit I had cared for so often with multiple lines going. Now I was on the receiving end of this life-sustaining device, and it was frightening.11

Most caregivers stated that their experience as a patient or family member resulted in increased emphasis on communication in their practice, newfound recognition of their ability to heal others, and/or greater empathy for patients and their families. One breast cancer survivor states:

But because the struggle with uncertainty after treatment is completed is usually a silent battle waged outside of the physician’s office, most physicians don’t think or talk about it. In my life as a primary care physician before cancer, I certainly did not. Now I believe that we physicians need to talk to our cancer survivors about the unique struggles of survivorship.26

Essayists regarded their medical backgrounds as beneficial to their own care, or that of family members. Benefits included the ability to recognize medical errors, make treatment recommendations, obtain assistance from colleagues, and better understand prognosis and treatment decisions. Many, however, expressed ambivalence over how active a role to assume in medical decision making, and for some, their medical insight created an emotional burden. In the following example, a resident recognizes the benefits of his medical background to his father’s care but acknowledges the potential for role conflicts when practitioners become recipients of health care:

The caretaking family role is stressful for all kinds of reasons, but physicians may face an added concern about crossing the boundary between caregiver and treating physician. Where is that line and how do we know if we crossed it? At one point I wrote a note in my father’s chart, an act that I felt would clarify some issues for members of his care team but that provoked a sharply defensive response from a fellow resident. Perhaps I crossed a boundary at that point, but I am certain of two things: my involvement in his care during the last few months of life was uniquely beneficial to my father, to me, and to his principal doctors, and—there is no chance to “do it over” should the opportunity be missed.33

Discussion

Our analysis of illness narratives provides a framework for understanding the formidable physical and psychosocial challenges faced by patients and families, as well as the role of physicians’ behavior in shaping the illness experience. Physicians’ communication and other interpersonal skills profoundly influenced the essayists’ experiences, which is consistent with previous work identifying such skills as important determinants of patients’ satisfaction.39,40 Our analysis also calls attention to the substantial stress associated with the family member role, an issue that is increasingly recognized in critical care medicine.41

Recognition of each patient’s humanity facilitates individualized, compassionate care by allowing practitioners to see beyond the technical needs of patients and families. The importance of a patient’s nonmedical history in our analysis is particularly germane to the challenge of educating more empathic providers.1,2 The fact that many essayists believed the illness experience had fostered patient-centered care in their own practices lends credence to curricular efforts to provide trainees with insight into the patient experience through the incorporation of literature, reflective writing, and mock admission of trainees to the hospital.42,43

Despite the tremendous number of health care workers in developed nations, relatively little is known about what it’s like to be both a recipient and practitioner of medical care.3 Our analysis suggests the dual-role experience is simultaneously advantageous and challenging. For instance, essayists could turn to colleagues for assistance, but at times perceived their own medical insight as overwhelming. Their experiences may parallel those of patients and families without medical backgrounds, who often feel overwhelmed by receiving medical information and/or partaking in medical decision making. Our findings underscore the importance of empathically communicating medical information tailored to the comfort level of the individual patient or family member.44

As a whole, the illness narratives that were by health care workers seem to reflect experiences that are more similar than dissimilar to the experiences of the general population (but perhaps they may not, because they were written by providers for providers). For instance, vulnerability, uncertainty, hopelessness, and the influence of personal history on illness are prominent themes identified by Kleinman45 among chronically ill people and by Lather and Smithies46 among women with HIV/AIDS. The essays in our analysis also are readily amenable to Frank’s47 classification of illness narratives into three types: restitution, chaos, and quest. To varying degrees, essayists in our study focused on seeking recovery (restitution), encountered feelings of despair and loss of control (chaos), and perceived their experience as a means of personal and professional growth (quest).

Implications

The domains and categories identified in our analysis offer a framework, or heuristic, for understanding the perspectives of patients and families. This framework may be particularly valuable to younger trainees and clinicians with limited experience as recipients of medical care. On the basis of our findings, we created the Appendix as an outline for capturing important psychosocial aspects of the illness experience. Trainees could use it to guide a 45-minute interview with a current patient or family member, qualitatively analyze the responses, and thereby identify supports and barriers to the interviewee’s medical care. The outline could similarly assist students to critically evaluate literary and film portrayals of illness, write their own illness narratives, and/or engage in role-play exercises as part of a course on professionalism.48 Brief narratives comprise a promising educational resource for busy inpatient or outpatient teaching services. In lieu of routine lecture sessions, the health care team could spend 30 minutes reading and discussing an essay selected on the basis of patient–provider dynamics in a recent clinical encounter.

The dual-role experience provides practitioners with considerable insight into how best to foster patient-centered care,49 and, arguably, quality-improvement initiatives should actively solicit input from colleagues who have received care at their institution. Virtually all providers will face challenges similar to those of the essayists during their lifetime, and hence, further efforts to characterize and improve the dual-role experience are merited.

Strengths and limitations

The content of the essays may not reflect the sentiments of the general population, because the majority were written by providers for an audience of providers. Furthermore, the process by which pieces were selected and edited may have altered the original narrative voice of the essayists. Essays were chosen for their ability to convey constructive messages to readers, and some are fictionalized accounts of actual events (personal communication: C. Laine, MD, senior deputy editor, Annals of Internal Medicine, 2004). Only 20% of submissions were published in this essay series during the time period studied, but in the context of qualitative research, this is not necessarily a limitation. Qualitative analyses strive to obtain a credible sample of exemplars of the phenomenon of interest, and the editorial process of essay selection serves as one way of generating a credible sample of essays. Regardless, the essays likely reflect fundamental aspects of patient–physician relationships in general, given that our findings were congruent with those of other researchers. Finally, our analysis is prone to our own individual biases. Feedback provided by essayists, however, indicates that our analysis accurately reflects important themes.

Conclusion

Illness narratives vividly illustrate fundamental aspects of the illness experience and are a potentially rich resource for cultivating more empathic care among health care professionals. Our analysis provides a framework for enhancing trainees’ and practitioners’ ability to understand and address the psychosocial needs of patients and families. Health care professionals face unique challenges and opportunities when they assume the role of health care consumers. Efforts to both foster patient-centered care and improve the dual-role experience are warranted. Further exploration of the illness experience offers the promise of more empathic, individualized care for patients and family members.

Acknowledgments

The authors thank Ilene Harris, PhD, from the University of Illinois, Chicago; Nicole Lurie, MD, MSPH, from the Rand Corporation; and Stephen Cook, MD, of Tacoma, WA, for their substantive reviews of the manuscript.

References

1 Halpern J. What is clinical empathy? J Gen Intern Med. 2003;18:670–674.
2 Larson EB, Yao X. Clinical empathy as emotional labor in the patient–physician relationship. JAMA. 2005;293:1100–1106.
3 Fromme E, Billings JA. Care of the dying doctor: on the other end of the stethoscope. JAMA. 2003;290:2048–2055.
4 Patton MQ. Qualitative Evaluation and Research Methods. 2nd ed. Newbury Park, Calif: Sage Publications; 1990.
5 Anonymous. A road less traveled. Ann Intern Med. 2003;138:1004–1005.
6 Bauer IS. Star treatment. Ann Intern Med. 1999;130:447–449.
7 Bendor R. Arthritis and I. Ann Intern Med. 1999;131:150–152.
8 Bingham J. A complaint against “complaints.” Ann Intern Med. 2003;138:73–74.
9 Bradshaw DY. A visit to the doctor. Ann Intern Med. 1999;131:627–628.
10 Clay AS. The medical student as patient. Ann Intern Med. 1999;131:225–226.
11 Dohan MC. Reflections on a bone marrow transplant. Ann Intern Med. 2000;132: 589–590.
12 Dewar C. Brain scan. Ann Intern Med. 1999;130:535.
13 Dewar C. On finding a penny. Ann Intern Med. 1999;130:535.
14 Durso SC. Trust. Ann Intern Med 2002;137:1004–1005.
15 Epstein N. On being a patient. To dream. Ann Intern Med. 1999;131:540–541.
16 Fetscher S. An entirely benign disorder. Ann Intern Med. 1999;131:786–788.
17 Ghose R. Complications of a medically complicated child. Ann Intern Med. 1999;131:105–108.
18 Hislop JO. Any oasis will do. Ann Intern Med. 2003;138:848–849.
19 Horn MO. The other side of the bed rail. Ann Intern Med. 1999;130:940–941.
20 Khan MA. Patient–doctor. Ann Intern Med. 2000;133:233–235.
21 Lesho EP. Fathers, doctors, and time. Ann Intern Med. 2001;134:623–624.
22 Lundberg M. On being a patient. On growing old. Ann Intern Med. 2002;136:484–485.
23 Luther K. Lester’s getting a bottle. Ann Intern Med. 1999;130:73–74.
24 McClenahan JL. Drop-in. Ann Intern Med. 2002;137:293–294.
25 McClenahan JL. The greenhouse effect. Ann Intern Med. 2003;138:434.
26 McKinley ED. Under toad days: surviving the uncertainty of cancer recurrence. Ann Intern Med. 2000;133:479–480.
27 Nowak TV. The ritual. Ann Intern Med. 1999;130:1025–1026.
28 Parker RA. The rebirth of Isabelle. Ann Intern Med. 2000;133:917–919.
29 Quinn C. The pirates. Ann Intern Med. 1999;131:472–473.
30 Rao JK, Koppaka VR. Santi. Ann Intern Med. 2002;137:852–854.
31 Schwartz TB. Deep waters run still. Ann Intern Med. 2001;135:378–379.
32 Tipton PJ. Nightmare from the sixties. Ann Intern Med. 2000;132:1002–1003.
33 Tsomides T. Reflections on a death. Ann Intern Med. 2001;134:246–248.
34 Udycz C. A presentation at the weekly meeting of the institutional review board. Ann Intern Med. 2001;135:58–59.
35 Van Peenen HJ. On being a patient. Consumer. Ann Intern Med. 1999;131:976–978.
36 Hill CE, Knox S, Thompson BJ, Williams EN, Hess SA, Ladany N. Consensual qualitative research: an update. J Couns Psychol. 2005;52:196–205.
37 Glaser BG, Straus AL. The Discovery of Grounded Theory: Strategies for Qualitative Research. Chicago, Ill: Aldine Publishing Company; 1967.
38 Stiles WB. Quality control in qualitative research. Clin Psychol Rev. 1993;13: 593–618.
39 Federman AD, Cook EF, Phillips RS, et al. Intention to discontinue care among primary care patients: influence of physician behavior and process of care. J Gen Intern Med. 2001;16:668–674.
40 Keating NL, Green DC, Kao AC, Gazmararian JA, Wu VY, Cleary PD. How are patients’ specific ambulatory care experiences related to trust, satisfaction, and considering changing physicians? J Gen Intern Med. 2002;17:29–39.
41 Azoulay E, Pochard F, Kentish-Barnes N, et al. Risk of post-traumatic stress symptoms in family members of intensive care unit patients. Am J Respir Crit Care Med. 2005;171:987–994.
42 Charon R. Narrative medicine: a model for empathy, reflection, profession, and trust. JAMA. 2001;286:1897–1902.
43 Wilkes M, Milgrom E, Hoffman JR. Toward more empathic medical students: a medical student hospitalization experience. Med Educ. 2002;36:528–533.
44 Keating DT, Nayeem K, Gilmartin JJ, et al. Advance directives for truth disclosure. Chest. 2005;128:1037–1039.
45 Kleinman A. The Illness Narratives: Suffering, Healing and the Human Condition. New York, NY: Basic Books; 1988.
46 Lather P, Smithies C. Troubling the Angels: Woman Living with HIV/AIDS. Boulder, Colo: Westview Press; 1997.
47 Frank AW. The Wounded Storyteller: Body, Illness, and Ethics. Chicago, Ill: University of Chicago Press; 1995.
48 McCarthy Veach P, LeRoy BS, Bartels DM. Facilitating the Genetic Counseling Process: A Practical Manual. New York, NY: Springer-Verlag Publications; 2003.
49 Berwick DM. My right knee. Ann Intern Med. 2005;142:121–125.
Appendix
Appendix:
Outline for Capturing Important Aspects of the Psychological States of Patients and Family Members
Appendix, continued
Appendix, continued:
Appendix, continued
© 2007 Association of American Medical Colleges