In 2000, the Institute of Medicine estimated that 98,000 people die annually in U.S. hospitals as a result of preventable medical errors. The release of the seminal report, To Err Is Human, led to a number of publicized efforts among hospitals and regulatory agencies to reduce the number of deaths and adverse events attributed to unintended and preventable medical errors.1–3 Today, mandated reductions in graduate medical education work hours seem to be the only widely implemented response despite calls to expand patient safety into the undergraduate medical school curriculum.
The Association of American Medical Colleges (AAMC) Medical Schools Objectives Project 2001 report, “Contemporary Issues in Medicine: Quality of Care,”4 was one of the first to advocate adding patient safety education to undergraduate curricula. Despite evidence of curricular innovation at single institutions, five years after the AAMC's call for national implementation of instructional programs it is unclear to what extent all medical schools have initiated patient safety training for their undergraduate students.5–7 Instead, much of the effort to introduce patient safety education has been directed toward residents and physicians already in practice.8,9
Today's medical trainees continue to demonstrate poor knowledge of patient safety concepts or are only superficially engaged in improving patient safety across a broad range of training levels, degrees, and specialties.5,10,11 However, a recent qualitative analysis of student reflective write-ups during their medicine clerkships suggests that students frequently are actively considering patient safety and system issues.12,13
The purpose of this study was to describe current patient safety curricula at U.S. and Canadian medical schools by means of an annual survey administered to the Clerkship Directors in Internal Medicine (CDIM) and to identify factors associated with adoption of such curricula.
We used data from the April 2006 CDIM annual survey of its U.S. and Canadian institutional members (each institution is represented by one institutional member in CDIM). Of 142 eligible schools, CDIM has 110 institutional members—the other 32 schools do not have CDIM members. If an institutional member does not know the answer to an annual CDIM survey question, it is CDIM's expectation that he or she will query other individuals at the institution to find the answer. CDIM institutional members span every region in the United States and are employed by a mix of public and private institutions.
The first section of the survey asked for demographic information about the respondent, including age, academic rank, number of years in current job position, and institution's class size; one section addressed patient safety education. Four multipart questions on patient safety were developed on an a priori basis and were dichotomous, multiple-choice, and modified Likert types (see Tables 1–3). Two of the authors (E.A. and S.D.) reviewed the available literature and developed items based on discussions with other clerkship directors. The section was pilot tested with members of the CDIM research committee and by members of the CDIM council before being e-mailed to institutional members as part of the CDIM annual survey in April 2006. Nonresponders were contacted up to three additional times through e-mail, postal mail, and/or telephone contact.
The study protocol was approved by the Uniformed Services University institutional review board.
The results were analyzed in de-identified, confidential format. Institutional members were used as the unit of analysis, as CDIM has only one institutional member per medical school. Unanswered items were not included in the analysis. Descriptive statistics were performed on all responses calculated using a standard statistical software program (SPSS, version 12, Chicago, Illinois).
Additionally, all items were tested to determine whether there were differential responses by school size, faculty age, clerkship director experience, and academic rank using chi-square, Mann-Whitney U, or Spearman rho, as appropriate for the measurement. School size was dichotomized at the median into small (140 or fewer students per year) or large (more than 140 students per year), and clerkship director experience was defined three ways: median age (≤43 versus ≥44), median years in the current role (≤5 years versus ≥ 6 years), and academic rank (assistant professor, associate professor, or professor).
Eighty-three of 110 (76%) institutional members completed the CDIM survey. Five institutional members did not complete the patient safety section and were excluded from our analysis, resulting in a 71% response rate (n = 78 respondents).
Patient safety instruction and instructional methods
Clerkship director responses regarding patient safety instruction, quality, and safe practices are outlined in Tables 1 to 3.
Only 25% (n = 78) of respondents stated their school has an explicit curriculum on patient safety. Forty percent indicated that their school does not have such a curriculum, whereas 29% indicated that they were uncertain whether their schools had one. Seventy-two percent (n = 74) indicated that they believed their school should have such curriculum. Responses to the question about whether a patient safety curriculum was present did not vary by institution size or type (private, public, or federal; P >.05). There were no relationships demonstrated between the presence of a patient safety curriculum and our demographic factors (median age, median years in current role, academic rank, or class size; all P values >.05).
For the 23 institutional respondents who indicated patient safety instruction occurs at their school, instructional methods during all four years of medical school are outlined in Table 2. Lectures and small-group instruction were the most commonly cited methods. Interestingly, it seems that more instruction occurs in the preclinical years (1 and 2) than clinical years (3 and 4).
Types of learning and practices
All survey respondents were asked to report their agreement about methods for learning about patient safety, quality, or delivering safe care that currently occur or should occur at their institution (Table 3). Not all members completed each item in this section so percent and number responding are shown, with the latter in parentheses. Thirty-two percent (n = 62) indicated that their students currently take part in root cause analysis of medical errors, whereas 66% (n = 66) indicated that they believe that students should be involved in this activity. Sixty-six percent (n = 65) of respondents indicated that their students attend and/or participate in morbidity and mortality review sessions; 81% (n = 62) believe that students should be included in this activity. Only 24% (n = 62) of respondents believe that their students receive training on the Joint Commission National Patient Safety Goals, and only 51% (n = 65) of respondents believe that students should be trained about these. Similarly, 46% (n = 61) perceive that their students receive training on the core measures (myocardial infarction, heart failure, pneumonia), but 74% (n = 66) believe that students should be trained about these quality measures. Seventy percent (n = 63) of respondents indicated that their students already receive training in how to write orders, but 90% (n = 63) felt that students should receive training in this area. Thirty-eight percent (n = 66) of clerkship directors feel that their students have less of an opportunity to play an important role as a patient care provider for their patients than in the past three years, and 33% (n = 66) feel that their students have less opportunity to write patient notes than in the previous three years.
This multicenter, cross-sectional survey of internal medicine faculty serving as clerkship directors found that only 25% of respondents (n = 78) reported that patient safety curricula were in effect at their institutions during 2006.
Consensus-based recommendations emphasize the need for patient safety instruction of medical students, but thus far information regarding instructional methods and curricula has been limited to experiences at single institutions.5–7 As shown in Table 1, although 72% (n = 74) of clerkship directors agree that there should be an explicit curriculum on patient safety instruction during medical school, most schools do not have such a curriculum.
Lectures and small-group discussion are reported as the most commonly employed educational techniques in the schools that do have explicit patient safety curricula. Our study suggests that demographic features of institutions may not affect the implementation of a patient safety curriculum. Several institutions are using innovative techniques such as practice-based learning, standardized patients, simulators for procedures, and Web-based/videotape scenarios. These models may allow students to learn from mistakes in a safe environment rather than in actual clinical practice with the potential for adverse patient outcomes.
Students' participation in activities traditionally employed for learning about patient safety, such as root cause analyses (32%) and morbidity and mortality rounds (66%), is more frequent but not universal. Additionally, responding clerkship directors reported that students are rarely trained about the National Patient Safety Goals, and only 50% (n = 65) think that they should be.
Our data indicate that more training on patient safety seems to occur during the first two years of medical school rather than in years 3 and 4. Qualitative studies suggest that student concern about medical errors begins early in training and continues throughout all four years.11 We believe that patient safety education should be introduced early in medical school, augmented, reinforced, and practiced throughout all four years. What students should learn about patient safety and what methods should be employed to achieve meaningful learning in these areas remain unclear and are important topics for future study.
Our data support that clerkship directors agree that students need to assume an important role in the care of their patients while at the same time receiving appropriate supervision. An important example of this concerns cosigning of medication orders. Medication errors, from ordering to administration to discharge, remain a common cause of medical errors in health care.14–16 Formal instruction on order writing and computerized order entry would likely help to reduce prescribing errors.
Our study has several limitations. Not all of the clerkship directors at the 142 accredited allopathic medical schools in the United States and Canada are represented in CDIM. Likewise, this survey represents the opinions of clerkship directors in internal medicine; other specialties may have markedly different patient safety training for their students. Also, as this was a study of perceptions, actual practice could differ from what we are reporting. We only sampled explicit curriculum. Clerkship directors may not have been aware of patient safety training taking place in their schools, for example. However, medicine clerkship directors are typically in a position to know of most aspects of clinical education. If an internal medicine clerkship director is not aware of this curriculum or whether it is occurring, it is unlikely to be a priority, if it is happening. Nevertheless, 29% of clerkship directors responded that they did not know whether patient safety curricula were in place.
We may not have had the power to detect whether there are actual differences between institution type or size and whether a patient safety curriculum exists, given the relatively small sample size and the fact that the data were opinions of clerkship directors and were not confirmed with the institutions. This warrants further investigation in future studies. Furthermore, although didactic lectures and small-group instruction were the two most common forms of instruction, we are unable to determine the influence of instructional methods on comprehension of the principles of patient safety. Because of its design, our study also does not provide outcome measurements on instruction in patient safety.
Finally, it is notable that these data were gathered in the spring of 2006. Since that time, there has been additional national focus on this issue. There is now a textbook written aimed at teaching patient safety to medical students,17 and the Institute for Healthcare Improvement has become a popular training resource for medical students and other members of the health care community.18 Therefore, this report best represents a snapshot of where medical education was at the time of the survey and may not accurately represent curricula that schools currently offer. Future studies should reassess the state of education in this area on an ongoing basis.
Our study also has several strengths. First, to our knowledge, this is the largest report of patient safety curricular experiences for medical students enrolled in U.S. and Canadian medical schools. The responses involved in this study originated from 83 different private and public U.S. and Canadian medical schools, lending generalizability to the results. Also, our response rate of 76% makes response bias less likely.
Our data also provide some potentially concerning information about clerkship directors' observations of the clinical education environment. Perhaps as a result of the increasing clinical focus on providing higher-quality, safer care, increasing use of electronic systems for documentation, and greater supervision, many of our respondents reported that students seem less able to be active participants in their patients' care. Others have noted that students are getting less exposure to performing a number of procedures before graduation.18 This phenomenon may delay or otherwise fundamentally alter the clinical education that students can receive in the quality and patient safety era. Training novice learners in medicine while providing the best possible care to patients will continue to remain a major challenge to medical education. We believe that helping students, residents, and faculty to better understand patient safety will be a critical factor in resolving this conflict.
In summary, our study suggests that most schools do not have an explicit patient safety curriculum, yet most clerkship directors recognize the importance of patient safety instruction during the undergraduate curriculum. We believe that more explicit and formalized instruction in patient safety across the four-year undergraduate curriculum is needed, and additional faculty development is likely also needed. Subsequent work should be done to determine how to optimally teach the requisite knowledge and skills, how to provide experiences to students so they can most effectively learn this discipline, and how students can practice medicine most safely. Such a formula likely would include formal didactics and reading on the latest findings in patient safety science, experiences in simulation, systems and error analysis, reflection on patient experiences, and mentorship.
The authors would like to sincerely thank the members of the Clerkship Directors in Internal Medicine Research Committee for their assistance with preparation of the survey, survey data collection, and assistance with manuscript preparation, and the CDIM staff for their assistance with survey distribution/management.
The opinions expressed in this paper are solely those of the authors and do not reflect the official policies of the Department of Defense, the United States Air Force, or other federal agencies.
1 Kohn L, Corrigan J, Donaldson M, eds. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 2000.
3 Jha A, Orav EJ, Li Z, Epstein AM. The inverse relationship between performance in the hospital quality alliance measures and mortality rates. Health Aff (Millwood). 2007;26:1104–1110.
5 Madigosky WS, Headrick LA, Nelson K, et al. Changing and sustaining medical students' knowledge, skills, and attitudes about patient safety and medical fallibility. Acad Med. 2006;81:94–101.
6 Halbach JL, Sullivan LL. Teaching medical students about medical errors and patient safety: Evaluation of a required curriculum. Acad Med. 2005;80:600–606.
7 Moskowitz E, Veloski JJ, Fields SK, Nash DB. Development and evaluation of a 1-day interclerkship program for medical students on medical errors and patient safety. Am J Med Qual. 2007;22:13–17.
8 American Board of Internal Medicine. Practice Improvement Modules. Available at: http://www.abim.org/pims
. Accessed August 10, 2009.
9 Kachalia A, Johnson JK, Miller S, Brennan T. The incorporation of patient safety into board certification examinations. Acad Med. 2006;81:317–325.
10 Kerfoot BP, Conlin PR, Travison T, McMahon GT. Patient safety knowledge and its determinants in medical trainees. J Gen Intern Med. 2007;22:1150–1154.
11 Fischer MA, Mazor KM, Baril J, et al. Learning from mistakes. Factors that influence how students and residents learn from medical errors. J Gen Intern Med. 2006;21:419–423.
12 Fischer M, Harrell H, Haley H, et al. Between two worlds: A multi-institutional qualitative analysis of students' reflections on joining the medical profession. J Gen Intern Med. 2008;23:958–963.
13 Nazem A. Teach us how. JAMA. 2008;300:2463–2464.
14 Bates DW, Cullen DJ, Laird N, et al. Incidence of adverse drug events and potential adverse drug events. Implications for prevention. ADE Prevention Study Group. JAMA. 1995;274;29–34.
15 Winterstein AG, Sauer BC, Hepler CD, et al. Preventable drug-related hospital admissions. Ann Pharmacother. 2002;36:1238–1248.
16 Gandhi TK, Weingart SN, Borus J, et al. Adverse drug events in ambulatory care. N Engl J Med. 2003;348:1556–1564.
17 Wachter R. Understanding Patient Safety. New York, NY: McGraw Hill; 2008.
19 Wu EH, Elnicki DM, Alper EJ, et al. Procedural and interpretive skills of medical students: Experiences and attitudes of third-year students. Acad Med. 2006;81(10 suppl):S48–S51.