Humanism, a critical component of caring for patients, must be a key outcome of medical education. Defined as “a way of being,” humanism manifests through altruism, integrity, respect for others, and compassion.1 It comprises a “physician's attitudes and actions that demonstrate interest in and respect for the patient and that address the patient's concerns and values.”2 Humanism has been considered an important component of professionalism since the groundbreaking work leading to the Charter on Medical Professionalism, originally published by the American Board of Internal Medicine in the mid-1990s.3–5
Since then, many medical schools have developed programs to promote humanism in learners. These programs typically create experiences that promote perspective taking, offer opportunities for reflection, and provide high-quality mentoring.6 The timing of such programs is important, however, as interventions during the preclinical years are often overshadowed by the more powerful experiences of the clinical years.7 One study showed that students in their later years of medical school reported attitudes that were significantly more doctor-centered or paternalistic than those of students in their earlier years of training.7 That study notwithstanding, anecdotal reports supply much of the data on the impact and student perceptions of programs focused on the development of humanism.8 It is also likely that factors outside traditional medical education may play a significant role in students' development and practice of humanism.
We sought to learn more about medical students' beliefs and attitudes regarding the development and practice of humanism through their experiences both inside and outside medical school. The study reported here reflects a multi-institution, multi-stage project with two objectives: (1) to develop a conceptual framework to enhance medical educators' understanding of the factors that are the most influential in medical students' development of humanism and (2) to explore fourth-year medical students' thoughts and attitudes regarding the most critical factors that encourage or inhibit the development and practice of humanism in medicine. We hope our framework and findings will inform decision making concerning the admissions process as well as curricular and extracurricular offerings during medical training.
Our research took place in four stages: First, we conducted focus groups of medical students. Second, we developed a conceptual framework to guide further inquiry. Third, we developed a student survey reflecting the components of our conceptual framework. Fourth, we implemented the survey and analyzed the results.
Between August 2006 and February 2007, our research team conducted 16 focus groups at four U.S. medical schools (University of Michigan Medical School, University of Missouri–Kansas City School of Medicine, Vanderbilt University School of Medicine, and Stanford University School of Medicine). Researchers at three of these medical schools used identical sets of questions and probes for 12 of the focus groups, all of which took place in 2006. At the fourth university, researchers added a few specific items reflecting the knowledge gained during the first 12 focus groups to confirm or refute themes that had begun to emerge. They conducted 4 focus groups in early 2007. IRB approval for the focus groups was obtained at each school.
We recruited focus-group participants via broadcast e-mails sent to fourth-year medical students in 2006 for the first 12 groups and in early 2007 for the remaining 4 groups. Participants were offered dinner and a $20 incentive. They were told that the researchers were interested in their thoughts about how medical students develop humanism and that their participation would be anonymous and limited to a single, one- to two-hour session. Participants were told that, for the purpose of the focus-group discussion, humanism should be thought of as encompassing respect, compassion, empathy, and integrity.
We asked participants to refrain from using their names and to maintain confidentiality of all responses given during the course of the focus group. We assigned each participant an ID number for the duration of the session, and each was referred to in the discourse by his or her number (e.g., “Respondent 1 said …”).
We recorded all focus groups and transcribed all recordings verbatim. Researchers at all four schools reviewed and coded all transcripts independently. We convened for an all-day working session in January 2007 to discuss themes that emerged and to use those themes to generate a conceptual framework to address the promotion and inhibition of humanism during medical training. Researchers at three of the four schools had completed their focus groups before the working session. Researchers from the fourth school participated in the working session and adjusted their focus-group script to elaborate on, clarify, and test some of the issues raised during the working session. We then analyzed their focus-group transcripts with an eye toward confirming or disputing what we had learned from the focus groups at the other three schools.
To develop our conceptual framework, we first created a list of the major themes that had emerged from more than 250 pages of transcribed focus-group data, and we grouped the themes into categories. We then reviewed existing conceptual frameworks for behavior development, including those focused on environmental influences, the “three Es” (expectation, experiences, and evaluation),9 the PRECEDE/PROCEED framework,10 and the common sociologic framework depicting the confluence of intrapersonal, interpersonal, and community-related factors. These conceptual frameworks provided a starting point around which we discussed the themes emerging from the focus-group data. We endeavored to build a model that took into account students' beliefs about how humanism is learned and their ideas about learning humanism over time.
We used our conceptual framework to generate a 102-item, standardized, self-administered survey that addressed various aspects of the development and practice of humanism. Items in the survey reflected the types, characteristics, and processing of experiences described in the conceptual framework. Each factor illustrated in the conceptual framework was incorporated into at least one item in the survey, although some items reflected more than one factor simultaneously. The survey included demographic questions (respondent's age, gender, marital status, number of children, educational debt, specialty choice) as well as Likert-scaled questions assessing which factors outside medical school, which experiences during medical school, and which types of situations promoted or inhibited humanism. Questions also assessed which people during medical school were most pivotal in the respondent's development or inhibition of humanism, and which types of experiences respondents thought should be built into the medical school curriculum.
In March 2007, we pilot-tested the survey at one of the four medical schools with 19 fourth-year medical students (the target cohort). As a result of the pilot-testing, we made minor revisions to the wording to ensure maximal comprehension and clarity.
We devised a sampling frame based on the Association of American Medical Colleges' 2006 list of medical schools enrolling students in the United States. In the absence of prior survey data with which to calculate the necessary sample size, and cognizant of our need to ensure a nationally representative sample, we chose a sampling strategy to ensure geographic diversity and a balance of public and private institutions and to deliver a 10% sample of the approximately 15,000 annual U.S. medical school graduates. It was our expectation that this sample size would far exceed the risk for type II (beta) sampling error. Assuming that we would achieve a response rate of slightly less than 50%, we arrived at a sample size of 3,000 graduating medical students to enroll 1,500 participants in the study. We then randomly selected medical schools from each region of the United States in turn, until the sample size was achieved. We contacted the deans (or designees) of each selected school in March 2007 and invited them to participate. Of the 25 schools we selected, the deans (or designees) of 20 agreed to allow students to participate.* There were no differences between selected schools that participated and selected schools that did not in terms of presence of a humanism honor society, humanism awards, size of the student body, public/private status, or geographic region. All participating schools' IRBs approved or exempted the research.
Survey data collection
We administered the survey at the 20 schools from April to May 2007 and repeated the same survey protocol at 19 of the schools from February to April 2008. We used participating schools' listservs to solicit fourth-year medical students by e-mail, sending a follow-up reminder if needed. Students who completed the survey were given the opportunity to enter their identifying information in a separate survey file to allow us to send them a $10 gift card as an incentive for participating. We collected data via Survey Monkey and downloaded all data to a Microsoft Excel file. The Excel file was cleaned, edited, and uploaded to SPSS (SPSS Inc., Chicago, Illinois) for analysis and interpretation.
Survey data analysis
We used descriptive statistics to characterize key features of the respondents, and we then calculated frequencies to identify those experiences that respondents characterized as the most and least influential in promoting humanism. We compared the 2007 and 2008 survey data to determine potential sampling differences using Student's t tests and chi-square analysis. We also compared respondents' chosen specialties with 2007 and 2008 National Resident Matching Program (NRMP) data to determine whether our sample reflected national distribution of medical students' specialty choices. We explored possible differences by gender, marital status, debt burden, or primary care specialty choice in the factors mentioned by students as promoting or inhibiting humanism. These comparisons were made using ANOVA, Student's t tests, and chi-square analysis. A P value of .01 was taken to be statistically significant.
Conceptual framework development
The data we collected from the focus groups held at the first three medical schools (12 focus groups; 57 fourth-year students and 7 first-year residents) were corroborated by the data collected from focus groups at the fourth school (4 focus groups; 12 fourth-year students and 4 first-year residents). Thus, despite our explicit attempt to challenge some of our early key findings, our data were similar across all four schools. Overall, focus-group data yielded many themes and sub-themes regarding the promotion or inhibition of humanism, including experiences during childhood, family upbringing, educational environment, personal experiences with illness, role models' effectiveness, seeing people “cut corners” or treat patients rudely, the impact of exhaustion, and being involved in the care of terminally ill patients. We categorized and aggregated similar themes and sub-themes to create a conceptual framework for the development of medical students' humanism (Figure 1). Key factors include the types of experiences before and during medical school, the characteristics of those experiences, and whether the student had an opportunity to process those experiences. All of these factors are situated within the broader environmental context, which includes institutional norms, policies and procedures, and workload.
A total of 1,170 students responded to the survey (592 in 2007 and 578 in 2008; for demographics, see Table 1). Their chosen specialties reflected national averages, within one or two percentage points, in most fields (comparison with NRMP data, not shown). The experiences that students rated mostly highly in terms of promoting humanism were exposure to positive role models, getting to know patients and their families well, and volunteer experiences before and during medical school (Table 2). They ranked formal recognition for humanism among the factors least influential in promoting humanism. Students rated being postcall or tired, on an extremely busy service, or having limited patient contact as the experiences most likely to inhibit the practice of humanism (Table 3). Table 4 reports the levels of importance students placed on various processing experiences as ways to facilitate the development of humanism.
Students felt strongly that some of the activities discussed in the survey should be built into the medical curriculum to encourage humanism. These are listed below.
Early in medical school
- 82.3% (963) felt that an anatomy course cadaver/family memorial ceremony was important.
- 78.6% (920) recommended early preclinical patient experiences (e.g., shadowing).
- 69.2% (810) recommended holding a white coat ceremony (a ceremony at the beginning of the first year, where students are presented with their first white coat and humanism is emphasized).
As part of the curriculum
- 86.6% (1,013) felt community volunteer work should be made part of the curriculum.
- 80.0% (936) thought international clinical experiences should be included in the curriculum.
As part of mentoring
- 85.6% (1,001) would like opportunities to watch and discuss positive role models.
- 78.4% (917) would like the chance to talk with a mentor about their experiences as medical students.
- 75.5% (883) would like opportunities to reflect on their experiences as medical students.
- 76.0% (889) would like feedback on humanistic behaviors.
Table 5 compares participants' responses by gender, marital status, debt level, and primary care as a chosen specialty. Generally, female respondents and students who chose primary care were likely to rank promoting and processing factors as significantly more important than did their male, non-primary-care counterparts. Respondents with lower levels of debt were more likely than those with higher levels of debt to rate exposure to positive role models and getting to know patients and their families well as important factors in promoting humanism. Marital status did not have a significant effect on respondents' ratings.
Table 6 contrasts the characteristics of those respondents who reported that they planned to go into primary care with those who selected all other specialties. Although there were significant differences between the groups in terms of the importance they assigned to promoting and processing factors, the only significant difference between the groups in terms of demographic factors is gender: 67.2% (370/551) of those going into primary care were women, whereas 48.7% (274/563) of those going into non-primary-care specialties were women (P < .001).
This study provides a conceptual framework to enhance medical educators' understanding of how humanism develops during medical training. The framework we present here—supported by survey data from 1,170 fourth-year medical students—suggests that it is not just the types of experiences that are important. Rather, the characteristics of those experiences and students' opportunity and ability to process the experiences are also pivotal in the development and practice of humanism.
Several themes emerged from our findings. Students reported that experiences of greatest intensity had the largest effect on their development of humanism. Of the students who had the following experiences, 95.7% cited getting to know a patient and his or her family well, 89.2% cited being involved in a case where the patient dies, 89.1% listed being exposed to terminally ill patients, and 78.1% listed participating in rotations with intense patient content as promoting humanism. Students also rated their participatory learning experiences, such as volunteer clinics (89.3%) and international clinical rotations (81.8%), as having strong, positive effects on their development and practice of humanism.
Exposure to positive role models (96.7% of those who had that experience) and watching and discussing positive role models (79.3% of all respondents) figured highly in students' ratings of factors that promoted humanism. Good mentors demonstrate for students on a daily basis what it means to be a caring, competent, and humanistic physician.2 Mentoring includes articulating for students what physicians are thinking as they work through various challenges and exactly how they go about building humanism into their daily practice.11,12 Our initial focus-group data and open-ended responses from students (data not shown) suggest that students are looking to their mentors for specific behavior, activities, and wording to help them integrate humanistic concepts into their own practice as physicians.
We were not surprised to find that factors inhibiting humanism included those that were considered stressful. Most of the respondents (81.8%) reported that being postcall or tired somewhat or strongly inhibited the practice of humanism; 78.7% reported the same was true for having a busy workload with a high patient census.
This study also suggests significant differences in the way women and students going into primary care think about the development, facilitation, and inhibition of the practice of humanism. Our findings further suggest that students with lower debt levels rate experiences as more likely to promote humanism than do those with higher debt levels. These observations warrant further research regarding the implications. The potential overlap in these categories should also be noted: Students going into primary care are more likely to be female and to have lower debt burdens. Is one of these factors a stronger predictor than the others in terms of attitudes toward the development of humanism? In addition, as noted above, our results suggest that stressful situations inhibit humanism. Economic stress, such as having a high debt load, may inhibit the practice of humanism as well. In that case, is the significance of debt burden that we observed simply a marker for stress?
Our findings also raise the following questions: Does promoting humanism outside primary care require different strategies? Do male and female physician learners require different approaches? Could policies to reduce debt burden affect attitudes toward humanism? And, perhaps most fundamentally, can humanism be taught effectively to learners who are not “primed” to believe in its importance?
Limitations caution the interpretation of these results. The response rate was lower than we had anticipated. Although 1,170 students completed the survey, we had aimed for 1,500 participants, approximately 10% of the annual 15,000 U.S. medical school graduates. Thus, the survey population may not be representative of all graduates. Of note, we compared our sample with the national sample of graduating medical students entering the National Resident Matching Program in 2007 and 2008 and found no differences with respect to gender, age, and specialty choice (data not shown). In addition, we cannot control for selection bias, given the voluntary nature of participation in the study. Finally, the students who completed our survey may have been those most interested in humanism.
On the basis of the themes that emerged from our focus-group and nationwide survey data, we offer several recommendations for enhancing the development of humanism among medical trainees before they enter medical school and after they matriculate. Students in this study considered the richest experiences to be those in which they were heavily involved with patients and their families, active in community service and palliative situations, and exposed to superb role modeling.
Targeting premedical students.
We believe that the results of this study suggest that premedical programs may want to consider integrating into their curricula discussions of issues such as students' upbringing, religion, volunteerism, and personal experiences with illness. Medical school admissions committees may wish to consider asking applicants to discuss those experiences that prospective students view as the most formative.
Targeting matriculated medical students.
Our findings suggest that medical schools should include or enhance early clinical experiences and prioritize close contact with patients and their family members. Student involvement with community, regional, and international volunteer programs should also be strongly encouraged and supported. Medical schools should also explore ways to expose students to the care of terminally ill patients. The growth in specialized palliative care services at hospitals and hospices offers opportunities to formally integrate students into these activities as one way of assisting in their development of humanism.
Medical schools need to identify and alter educational environments that inhibit humanism. They should also explore ways to minimize fatigue and to institute programs that help students alleviate stress. They should give medical students opportunities to receive faculty feedback on humanistic performance and time to reflect on especially challenging cases. Perhaps most important, medical schools need to explore, develop, and support programs that enhance students' exposure to positive faculty role models.
The authors would like to acknowledge Barbara Packer, Sandra Gold, Ann Bruder, Allison Sole, and Emily Sippola for their contributions to this article.
This research was made possible through a grant from the Health Care Foundation of New Jersey and the Arnold P. Gold Foundation for Humanism in Medicine.
All research reported here was reviewed and approved or exempted by the University of Michigan institutional review board. The focus-group research was approved (HUM00006677); the anonymous survey research was exempted from ongoing review (HUM00011008). Each school participating in the focus groups or student survey received from its IRB approval, exemption, or notification of deference to the University of Michigan's umbrella application.
Preliminary results of this research were presented as a poster at the 2008 annual meeting of the Association of American Medical Colleges Annual Meeting in San Antonio, Texas.
1Cohen JJ. Linking professionalism to humanism: What it means, why it matters. Acad Med. 2007;82:1029–1032.
2Branch WT Jr, Kern D, Haidet P, et al. Teaching the human dimensions of care in clinical settings. JAMA. 2001;286:1067–1074.
3American Board of Internal Medicine. Project Professionalism. Philadelphia, Pa: American Board of Internal Medicine; 1994.
4Arnold L, Stern DT. What is medical professionalism? In: Stern DT, ed. Measuring Medical Professionalism. New York, NY: Oxford University Press; 2006:15–37.
5American Board of Internal Medicine. Medical professionalism in the new millennium: A physician charter. Ann Intern Med. 2003;136:243–246.
6Stern DT, Cohen JJ, Bruder A, Packer B, Sole A. Teaching humanism. Perspect Biol Med. 2008;51:495–507.
7Haidet P, Dains JE, Paterniti DA, et al. Medical student attitudes toward the doctor–patient relationship. Med Educ. 2002;36:568–574.
8Brainard AH, Brislen HC. Viewpoint: Learning professionalism: A view from the trenches. Acad Med. 2007;82:1010–1014.
9Stern DT, Papadakis M. The developing physician—Becoming a professional. N Engl J Med. 2006;355:1794–1799.
10Green LW, Kreuter MW. Health Promotion Planning: An Educational and Environmental Approach. 2nd ed. Toronto, Ontario, Canada: Mayfield Publishing Company; 1991.
11Gieselman JA, Stark N, Farruggia MJ. Implications of the situated learning model for teaching and learning nursing research. J Contin Educ Nurs. 2000;31:263–268.
12Cruess RL, Cruess SR. Teaching professionalism: General principles. Med Teach. 2006;28:205–208.
*The 20 participating medical schools were Baylor College of Medicine; Creighton University School of Medicine; Dartmouth Medical School; Georgetown University School of Medicine; Loyola University Chicago Stritch School of Medicine; New York University School of Medicine; Chicago Medical School at Rosalind Franklin University of Medicine and Science; Texas A&M Health Science Center College of Medicine; Tufts University School of Medicine; University at Buffalo, School of Medicine and Biomedical Sciences, State University of New York; University of Arkansas for Medical Sciences College of Medicine; University of Cincinnati College of Medicine; University of Iowa Roy J. and Lucille A. Carver College of Medicine; University of Kentucky College of Medicine; University of Massachusetts Medical School; University of Michigan Medical School; University of Missouri–Kansas City School of Medicine; University of North Carolina at Chapel Hill School of Medicine; University of Tennessee Health Science Center, College of Medicine; and The Warren Alpert Medical School of Brown University.© 2010 Association of American Medical Colleges