From the 1970s to the present, a steady stream of literature has insisted on the importance of ethics education and the need to establish an adequate ethics curriculum within medical schools.1–6 A variety of arguments for ethics education have been offered, however, Thomasma and Pellegrino summarize well the main point: the practice of medicine is an intrinsically ethical enterprise because patients are suffering and vulnerable, and medical treatments are not merely technical, they often invade patients' bodies and engage their consciences.7
Several efforts have been made to develop an “ideal” medical ethics curriculum. Elements of ideal curricula typically include approaches to ethical problems (e.g., principlism or casuistry), issues related to professional ethos (e.g., codes and duties to treat), multidisciplinary issues (e.g., ethics committees and impaired colleagues), patients' autonomy and clinical dilemmas (e.g., consent, privacy, use of life support), student—physician issues (e.g., disclosure of students' status in the clinical setting), issues in academic medicine (e.g., authorship and research ethics), and social issues (e.g., resource allocation and preventive medicine).8–12 Addressing each of these ethical topics arguably contributes to the ultimate aim of enhancing the quality of patient care in ways that respect other basic human needs in society.
Arguments for ethics education have not fallen on deaf ears. The consensus that ethics education is an important element in the formation of new physicians is reflected in the American Association of Medical Colleges' (AAMC's) Curriculum Directory, where all medical colleges in the United States claim to require ethics education.13
Nevertheless, a medical college's claim to require ethics education tells us nothing about the nature of the requirement. For example, most schools (58%) state that they teach ethics only as one component of a larger required course.13 Ethics might receive 20 or more formal classroom hours or no formal time (instead, all faculty may be encouraged to engage ethical issues in their courses and on clinical rounds). Moreover, little is known about the contents of required formal ethics components. While studies have been done on what ethics education ought to be, no study to date has determined what actually are the objectives, teaching methods, course contents, and assessment methods used in ethics education in U.S. medical schools.
Unfortunately, present curriculum directors are unable to evaluate the strengths and weaknesses of their ethics offerings relative to those of other schools. Moreover, medical ethicists are unable to determine how actual ethics courses compare with the several “ideal curricula” that have been developed by committees of experts. Finally, knowledge of the actual breadth of ethics education is currently limited to anecdotes and to reviews of the literature, thus limiting creative efforts to rework syllabi. These three gaps in our knowledge and abilities motivated this study.
This study's primary aim was to determine the scope and contents of required, formal ethics components in the curriculum of U.S. medical schools. We analyzed ethics course syllabi to identify and rank order the courses' objectives, teaching methods, content areas, and methods of assessing students' performances.
The AAMC provided mailing labels for all curriculum directors of four-year medical colleges in the U.S. (n = 121). A one-page questionnaire was sent to all curriculum directors asking whether ethics at their schools was taught as a formal, required component, as an elective, or not at all. We also asked for the year or years in which ethics was taught to students. Finally, we requested course syllabi for all required, formal ethics components in the four-year medical curriculum used during the 1999–00 school year.
We read all syllabi using an open-coding method to produce a comprehensive list of all elements found in the syllabi that fell into one of four generic categories: course objectives, teaching methods, course contents, and methods of assessing students. All other statements (e.g., those pertaining to class times, locations, and instructors) were ignored. We then analyzed the open-coding elements to subsume them under natural and, where possible, non-over-lapping categories. The categories were used to create a database in a standard software program. The syllabi were then read individually and their coded data entered into the database. Schools, rather than syllabi, constituted cases in our database; if a school had more than one required ethics component, data from all required courses' syllabi were entered into that case. Data from ten syllabi (17%) were entered by two researchers to establish interrater reliability. The data from remaining syllabi were then entered by one researcher.
We used t-tests for independent samples to compare participating schools with non-participating schools with regard to enrollment and tuition cost to estimate the generalizability of our findings. Our statistical analyses of syllabi data primarily aimed to provide descriptive data on the frequencies of various course objectives, teaching methods, course contents, and methods of assessing students. We performed two-sided t-tests to compare the mean numbers of course objectives, teaching methods, course content areas, and assessment methods between schools that are religiously affiliated and those not so affiliated, and between schools that do and do not grade their ethics courses. We calculated correlations between mean numbers of course objectives, teaching methods, course content areas, and methods of assessing students.
Finally, we compiled and rank ordered readings that were recommended or required in the syllabi.
Questionnaires were returned by 87 representatives of the schools (72%). Of these, 69 (79%) claimed to require a formal ethics course, and of this subset, 58 (84%) provided their ethics courses' syllabi. Participating schools did not differ significantly from non-participating schools with regard to enrollment or tuition costs.
Interrater reliability was .90 on how to categorize a given statement in the syllabi. The total number of relevant statements identified across all syllabi was 1,292. Only 18 of these statements (1.4%) were deemed “uncategorizable” due to ambiguity (e.g., content areas such as “media on the midway” or “current issues”). Codification and analysis of all syllabi identified ten course objectives, eight teaching methods, 39 content areas, and six methods of assessing students. The means for individual schools were three objectives (range = 0–8), four teaching methods (range = 0–6), 13 content areas (range = 3–32), and two methods of assessment (range = 0–5).
Table 1 presents rank orders and descriptions for the ten course objectives. Only two course objectives were found in more than half of the courses: to familiarize students with medical ethical topics (77%) and to develop ethical reasoning/problem-solving skills (64%). Table 1 also presents rank orders and descriptions for the eight teaching methods. Four teaching methods are used by a majority of schools: discussion/debates (84%), readings (83%), writing exercises (64%), and lectures (64%). Finally, Table 1 presents the names and rank orders of the six methods of assessing students. Two methods were used in a majority of required ethics components: class participation (90%) and examinations (64%).
Table 2 presents the names and rank orders of the 39 content areas. Only six content areas were taught in a majority of schools: informed consent (85%), health care delivery (75%), confidentiality and privacy (67%), quality of life/futility/provision of treatment (67%), death and dying (66%), and euthanasia and physician-assisted suicide (60%). Table 2 also shows that 39% of all medical schools addressed research ethics in some manner. Given the recent interest of the Office of Research Integrity and other governmental bodies in research integrity, we have devoted a separate article to analyzing in detail the research ethics course contents of these syllabi.14
Fifty-one percent of all colleges reported teaching ethics in one year only (26.5% in the first year, 14.3% in the second year, 8.2% in the third year, and 2% in the fourth year). Thirty percent reported teaching ethics in two years, 10% in three years, and 8% in four years of medical school. This information was derived from both the syllabi and the questionnaire and does not indicate that all of these year components involved formal education.
Eight colleges of medicine were religiously affiliated, and these schools did not differ significantly from non—religiously-affiliated schools in the numbers of coded elements found in syllabi. A total of 48 schools' syllabi indicated that courses were graded, and the graded courses did not differ significantly from non-graded courses with regard to the numbers of coded elements found in syllabi. Graded courses using a pass/fail method, however, addressed significantly fewer content areas (three fewer on average) than did those using other methods of grading (p < .05, t = −2.01).
The mean numbers of course objectives, teaching methods, course content areas, and methods of assessing students found in syllabi were all significantly correlated with each other at p < .05, with the exception of the number of methods and number of content areas, which showed a strong trend (p < .065).
The 58 syllabi either required or recommended 1,191 distinct readings, only eight of which were used by more than six schools. No reading was required or recommended by more than ten schools. Table 3 lists in rank order those readings used by more than 10% of all schools.
Throughout this discussion, some limitations of this study will emerge that suggest that caution must be used in drawing hard and fast conclusions from our data. Nevertheless, our participation rate of 72% was impressive, especially considering that some schools do not offer formal courses in ethics and that others will not share course syllabi on principle. The fact that participating and non-participating schools did not differ significantly regarding size or tuition cost suggests that our sample was not only large but also representative. Moreover, the 90% agreement in the ways the raters categorized elements of course syllabi suggests that our data are trustworthy.
Overall, the study achieved its aim. Nevertheless, as a descriptive study, our findings leave unresolved the most important questions: Are changes to ethics curricula necessary or desirable? And, if so, what changes should be made?
The question of relevance is best discussed in the context of examining the extent to which this study fulfilled its three main purposes: (1) to enable curriculum directors to evaluate the strengths and weaknesses of their ethics program relative to those of other schools, (2) to enable medical ethicists to determine how “real” ethics education compares with “ideal ethics curricula,” and (3) to enable ethics instructors to see the breadth of methods and topics used in ethics education.
First, by comparing his or her own course syllabus with the information presented in Tables 1 and 2, a curriculum director can reliably determine how his or her course compares with most other ethics courses taught in medical schools throughout the United States. In contrast to most quantitative survey studies, however, the raw data (the words on the syllabi) and not the survey instrument (e.g., fixed items on a paper-and-pencil questionnaire) generated the categories used. This means that, particularly when comparing course content areas, discretion must be used. Some categories subsume others (e.g., “death and dying” might subsume “euthanasia and physician-assisted suicide,” or “patient—physician relationship” might subsume “truth-telling” and “confidentiality and privacy”), whereas other categories simply overlap in a vague manner (e.g., “professionalism” and “patient—physician relationship”). We left the categories in their present forms because further analytic reduction of the categories would have led to the loss of information and in some cases a possible distortion of course content. This means that a syllabus that lists only very general content areas will be shortchanged in the comparison if the course actually covers several specific topics under these general headings. In turn, discretion must be used in comparing the numbers of content areas found in some syllabi with this study's mean number of content areas. This problem, however, hardly exists with the other generic syllabus categories (course objectives, teaching methods, and methods of assessing students).
Nevertheless, determining the relative strength or weakness tells us very little from a normative point of view, and we hope our data will not encourage schools to be satisfied with being at or even above the national mean. Relating what is real to what is ideal is a complex task. On the one hand, the actual ethics curriculum described in this study has a normative element; it presumably reflects what numerous ethics instructors considered to be the best ethics course they could offer given real time and resource constraints. On the other hand, one cannot simply move from the fact that schools are doing such-and-such to the claim that such-and-such is ideal or even adequate. We found some of our results were clearly inadequate to the needs of future physicians. Less than one third of all schools formally addressed ethical issues surrounding genetic testing and screening. This is worrisome given that genetic testing and screening is already common and will become more common as our knowledge in this arena advances. Further, the results of genetic testing pertain to other important ethical issues such as health insurance coverage, abortion and eugenics, privacy rights, and rights to information about one's health. Perhaps most surprising to us was the fact that only 10% of all schools formally discussed ethics committees and consultations. Students need to be made aware that ethics committees are consultative bodies available to assist staff, families, and patients in making difficult health care decisions. Failure to address the nature of ethics committees may contribute to the perpetuation of misperceptions (e.g., the perception of those who call consults as whistleblowers) and the underutilization of valuable resources.
Other shortcomings of our national curricular performance in ethics are best understood by contrasting the national averages with the results of “ideal” ethics curriculum studies.9,10,12 While none of the ideal-curriculum study groups offers the final word on what an ideal curriculum is, the opinions of these groups of academic deans and medical ethicists are informed and deserve serious attention.
For example, Musick and Jenkins used a Delphi survey to attain consensus on what elements are “essential” or “very important” to teach in medical ethics courses.12 In the first round, 80% of 55 academic deans of U.S. medical schools reached a consensus on 25 topics that should be taught. In round two using the same sample, consensus was reached by 80% of the second-round participants that 19 topics were essential or very important. If the 19 topics are mapped onto the course content areas found in the present study, all 19 “ideal” topics are included among the top 22 course content categories found in this study. This appears to show considerable overlap between the ideal and the real. However, as noted already, only six of these topics are covered by more than half of the medical schools that require formal ethics education. A majority of schools are not covering a majority of topics that the curricular deans of medical schools consider to be essential or very important in the formation of future physicians, which suggests the need to invest significantly more time in the medical school curriculum to ethics education.
Miles et al. discuss the importance of evaluating students not merely to assess their progress toward educational goals, but to communicate that ethics is as rigorous and deserving of attention as other areas of the medical curriculum.9,p.710 Our study determined that the majority of programs requiring formal ethics education do evaluate students, though the most common mode of grading is pass/fail and the most common criterion for assessment is “participation” (which may mean that by merely attending the required sessions one receives the highest grade). The fact that the number of methods of assessing students correlates positively and significantly with the numbers of course objectives, teaching methods, and content areas might reinforce the idea that the rigor and quality of assessment go hand in hand with the overall rigor and quality of a course. In a review article, however, Musick notes:
With the exception of the 1985 accreditation standards, there are no other official requirements that U.S. medical schools provide instruction in subject areas related to ethics, law, humanities, values, and/or professional attitudes. Even the LCME Standards are very broad in nature, leaving ample room for interpretation by each school …. There appears to be little consensus among medical schools as to what a required or “core” curriculum in ethics education for medical students should contain.11p.246
Our study confirms that there is nothing like a common core curriculum in medical ethics at present. (The fact that only eight of 1,191 readings are used by more than six schools—and no reading is used by more than ten schools—illustrates this fact quite powerfully.) Moreover, it confirms that the gap between the real and the ideal is significant even among schools that require formal ethics education.
In the absence of any official requirements of content for ethics education, the role our data can play in aiding individual faculty members in developing their own ethics curricula takes on greater significance.