Evidence-based medicine in internal medicine clerkships: A national survey.

ABSTRACT

Background. Although evidence-based medicine (EBM) has become widely accepted, the extent of its implementation during clinical clerkships is not well described. This study was done to characterize the implementation of formal EBM curricula in internal medicine clerkships.

Methods. In 1999, the Clerkship Directors in Internal Medicine surveyed its membership from 123 medical schools. The EBM section of the four-part survey addressed implementation, teaching and evaluation techniques, and barriers to implementing EBM.

Results. Survey response was 89%. Of 109 respondents, 38.5% reported having a formal EBM curriculum at some time during the third year and/or fourth year. Variability existed in curricular materials and evaluation tools used. The most common barriers to implementing EBM were lack of time and inadequately trained faculty.

Conclusions. Evidence-based medicine has been formally incorporated into relatively few internal medicine clerkships. Faculty development programs in conjunction with innovative teaching methods may help overcome these barriers.

**********

EVIDENCE-BASED MEDICINE (EBM) is not only a way to practice medicine, but also is a strategy that promotes lifelong learning skills. The Society of General Internal Medicine/Clerkship Directors of Internal Medicine (SGIM/CDIM) core clerkship guide defines nine basic generalist competencies as the highest priority to be taught whenever appropriate. (1) Three of these competencies--self-directed learning, diagnostic decision making, and therapeutic decision making--may be enhanced by the addition of EBM to the internal medicine clerkship curriculum. (2) Although EBM plays an important role in critical thinking and medical decision making, no data reflect the extent of its implementation in undergraduate internal medicine curriculum. The EBM Special Interest Group of the CDIM designed a questionnaire to be included in the organization's annual membership survey. The primary objective of this portion of the survey was to determine the extent of EBM implementation in the core internal medicine clerkships of medica l schools in the United States. Also, this special interest group sought to delineate the experiences, barriers to implementation, and evaluation tools currently being used by internal medicine clerkship directors (CDs). Thus, the main question posed by this survey was: "To what degree is EBM a formal curricular component internal medicine clerkships on a national basis?"

Broadly defined, EBM is the "conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients." (3) Operationalizing this definition for third-year medical students on a clinical clerkship predominantly involves students learning to develop a clinical question and searching available resources, and only then beginning to learn how to critically appraise the validity of retrieved sources and how to apply this information to an individual patient. Thus, the practice of EBM requires an amalgamation and integration of everything students have learned thus far using a problem-solving approach: pathophysiology, pharmacology, history-taking skills, physical examination skills, and development of a differential list, in addition to critically appraising information gained from systematic research.

METHODS

In 1999, the CDIM Evaluation and Research Committee developed a four-part membership survey. The first three sections of the survey focused on demographic information of internal medicine CDs, administrative issues, and grading and evaluation practices in internal medicine clerkships. The CDIM EBM special interest group developed the EBM section of the survey. This section focused on several main categories: (1) demographics of EBM use in third- and fourth-year internal medicine clerkships; (2) EBM curriculum and evaluation; (3) barriers to implementing EBM; and (4) presence of an EBM-centered faculty development program.

The entire survey was reviewed and approved by the CDIM council. The confidential survey was mailed to all 123 people listed as CDIM institutional members as of 1999, which included 122 medical schools in the United States and 1 in Canada. The initial mailing occurred in April 1999. Nonresponders received a second mailing and telephone and/or electronic mail reminders.

Variables in EBM included the year EBM was introduced into the curriculum, whether EBM was a formal component of Third-Year Medicine Clerkships, whether EBM was a formal component of Fourth-Year Medicine Clerkships, how long EBM had been a formal part of the curriculum, whether EBM was specific to medicine clerkships, what other clerkships taught EBM, the number of faculty teaching EBM, the number of hours dedicated to EBM in the medicine clerkship, who developed clinical questions, the types of experiences used to formulate EBM questions, whether students were taught to do literature searches, who taught literature searches, the types of evaluation used for EBM, whether a formal faculty development program in EBM was offered, the types/format for EBM faculty development programs, settings in which EBM is taught during the medicine clerkship, barriers encountered when incorporating EBM in the curriculum, and the number of hours devoted to formally teaching critical appraisal skills.

Demographic variables for medicine CDs included the following: age, years as clerkship director, percentage of salary for being a CD, percentage of time spent on clerkship, number of clinic half-days per week, number of months on inpatient services, number of other courses directed, number of other courses taught, number of medical school committees served, number of hospital committees served, number of peer reviewed papers, number of awarded grants, length of clerkships, number of students per year, number of inpatient clerkship sites, number of outpatient clerkship sites, and the composition of clerkships (inpatient, ambulatory, or combined).

Analyses were done using SPSS version 10.0 (SPSS Inc, Chicago, Ill). Qualitative data included responses to open-ended questions, and these were sorted first by question number, response, and common themes. The Mann-Whitney U test and the Fisher exact test were used to compare CDs with formal EBM curriculum to those without a formal EBM curriculum. Kruskal-Wallis analyses were used to compare EBM variables for programs that had a formal EBM curriculum in the third year, in the fourth year, or both. For other EBM specific variables, Mann-Whitney U and cross-tabulations along with Mantel-Haenszel and chi-square analyses were used where appropriate to compare clerkships with a formal EBM curriculum to those without a formal EBM curriculum.

RESULTS

The overall survey response rate was 89% (109/123). Of the 109 returned surveys, 42 CDs (38.5%) reported having a formal EBM curriculum at some time during the third and/or fourth year. Of these 42 CDs, 28 reported having a formal EBM component in the internal medicine clerkship during the third year only, 4 CDs had a formal EBM component only during the fourth year, and 10 had a formal EBM curriculum during both third and fourth years.

Of the 42 clerkships with a formal third-year internal medicine EBM curriculum, 26 had responded that EBM had been formally introduced into the curriculum before the third year of medical school, ie, during the first year in 17 of the 26 and introduced during the second year in the 9 other programs. Of the 63 clerkships that did not have a formal EBM curriculum in their medicine clerkship, 18 reported having a formal EBM curriculum during the first year and 10 during the second (preclinical) year of medical school.

Comparing those clerkships with EBM and those without EBM showed no statistically significant differences in CD demographics except in the number of clinic half-days per week (Table 1). Clerkship directors who had a formal EBM curriculum spent 3.6 half-days in the clinic compared with 2.9 half-days for those who did not (P=.046). Neither the CDs' academic rank nor scholarly activity, as measured by peer-reviewed publications or receipt of grants, predicted the use of a formal EBM curriculum in the clerkship. Similarly, clerkship characteristics (such as clerkship length, number of training sites, or the presence of an ambulatory component) did not predict implementation of a formal EBM curriculum in the medicine clerkship.

For those clerkships using EBM, the mean length of time it had been incorporated into the third-year clerkship curriculum was 2.5 years, 1.7 years for fourth-year clerkships, and 3.7 years for both third and fourth years (Table 2). The average number of hours devoted to teaching EBM during third- and fourth-year internal medicine clerkships was 6.7 hours. On average, approximately three faculty members taught the EBM curriculum, though more faculty participated when EBM was in both the third- and fourth-year clerkships (5.2 faculty vs 3.3). However, written narrative comments reported ranges of 1 to more than 30 preceptors participating in teaching EBM. There were no statistically significant differences between the clerkship years with respect to number of faculty participating in EBM (P = .06). Despite the number of faculty teaching EBM, EBM-based faculty development was provided by only 20% (n = 8) of the clerkships. Workshops (50%, n = 4) were the most common format used for faculty development. Conferenc e series (n = 3), faculty journal clubs (n = 3), and/or other formal courses (n = 2) were some of the other formats reported.

Common curricular materials for EBM included generating clinical questions and performing literature searches. For clerkships with EBM curricula in the third year only, students developed clinical questions 48.1% of the time, faculty 22.2% of the time, faculty and students together 22.2%, and others 7.4% of the time (Table 2). For clerkships with EBM curricula in only the fourth year, students developed clinical questions 50% of the time and faculty and students together the remaining 50%. For clerkships with EBM curricula in both third and fourth years, clinical questions were developed by the students (11.1%), faculty (44.4%), faculty and students together (33.3%), and others (11.1%). There was no statistical significance between the third year, fourth year, and both clinical years with regard to who developed the clinical questions (P = .38).

Narrative comments from the surveys indicated that students primarily based their clinical questions on direct patient interactions and written patient cases. Problem-based learning cases, standardized cases, and article discussions were some of the other formats used. Clinical questions were discussed in oral presentations and via e-mail.

For clerkships with EBM curriculum, 85% of the third year, 100% of the fourth year, and 66.7% of combined third and fourth years incorporated courses on literature searches to help students learn how to most effectively solve their clinical questions. Librarians and trained faculty were the main instructors for these courses (Table 2).

Evidence-based medicine was taught in a variety of settings (Table 2). During the third year, EBM was taught in inpatient settings 30.8% of the time, ambulatory settings 26.9% of the time, and both settings 42.3% of the time. During the fourth year, EBM was taught in the inpatient setting 75% of the time and in the outpatient setting 25% of the time. For the clerkships that used offered EBM in both third and fourth years, the inpatient setting was used 44.4% of the time, while a combination of both inpatient and outpatient settings was used 55.6% of the time.

Various evaluation methods used to assess a student's EBM skills at the end of the clerkship (Table 3) included critical appraisal skill exercises (53%, n = 17), papers (22%, n = 7), and multiple-choice questions (16%, n = 5). Essay questions were the least used method (9%, n = 3) and were mostly used during the fourth year. Other unspecified evaluation methods were used 22% of the time (n = 7).

The perceived barriers to implementing EBM curricula were, in order of frequency, as follows: inadequately trained faculty (83% of respondents), lack of time (65.7%), lack of faculty enthusiasm (43%), lack of curriculum (34%), and lack of student interest (34%). (Numbers do not add to 100% since respondents could identify as many barriers as appropriate.) Adequate access to the National Library of Medicine and interest among site directors were not significant barriers to implementing EBM. From written comments, lack of time to grade exercises was also indicated as a barrier. Seven (10%) of the CDs without a formal EBM curriculum also responded to this question, though they were not required to do so, and identified the same barriers with the same frequency.

Finally, EBM was not limited to the internal medicine clerkships. Written comments indicated the following other clerkships offered EBM courses: family medicine (n = 2), obstetrics/gynecology (n 3), psychiatry (n = 2), surgery (n = 2), primary care (n = 2), pediatrics (n = 2), neurology (n = 1), and all clerkships (n=4).

DISCUSSION

Evidence-based medicine has been recently characterized as a paradigm shift in medical education. (4) In an effort to provide medical students and residents necessary skills, both the Association of American Medical Colleges and the Accreditation Council for Graduate Medical Education have required inclusion of critical appraisal skills as a curricular subject. (5,6)

Despite accreditation requirements, our survey found a minority of the internal medicine clerkships used a formal curriculum for teaching EBM; only 35% of third-year internal medicine clerkships and 13% of fourth-year internal medicine clerkships used such a formal curriculum. The majority of the clerkships with a formal EBM curriculum report having EBM introduced early in the preclinical years. However, there seems to be little longitudinal application of EBM during the clinical years of medical school curriculum or across disciplines during the core clerkship rotations of the third year. Lack of time in busy inpatient and outpatient clinical settings and lack of adequate EBM skills of the house staff and faculty might explain this finding. A recent national survey of internal medicine residency programs by Green (7) showed that only one third of the residency training programs offered a freestanding EBM curriculum. Residents do play a major role in medical student education. However, our survey was not desi gned to determine whether lack of resident training was an independent determinant of EBM in the medicine clerkship.

Also, EBM is not solely within the domain of internal medicine. Nevertheless, according to our survey results, it would appear that internal medicine clerkships have taken a significant role in implementing EBM curricula in clerkships and have the experience and expertise to lead. We recognize that EBM may be taught in informal ways by internal medicine and taught by other clinical disciplines.

We found that, on average, only three faculty members taught EBM in third-year medicine clerkship, one in the fourth year, and five [n both years. Librarians provided the most support on "how to do literature searches." However, we could not determine whether the survey respondents' concerns regarding the lack of trained faculty reflected the lack of availability of a librarian to assist with or lead instruction or the lack of faculty to apply EBM to clinical settings or both. The availability of such library personnel to coordinate instruction with clinical faculty appears to be more effective during clinical years than when it occurs during preclinical instruction. (8)

With regard to time being a barrier to implementing EBM, our survey suggests that the amount of time devoted to teaching EBM during the clerkship amounts to less than 1 hour per week, based on the internal medicine clerkship median length of 12 weeks. (9) However, this time must also incorporate local resources, including faculty, librarian support, and computers that have access to MEDLINE and the National Library of Medicine. Our survey did not identify all the specific resources used for teaching EBM (such as Cochrane's Library or Best Evidence), but adequate access to the National Library of Medicine and MEDLINE were not significant barriers to implementation. Finally, these perceived barriers to implementing and practicing EBM are not unique to academic settings. A lack of knowledge, lack of familiarity with basic skills, and lack of time continue to be the primary barriers to practicing EBM by general internists. (10) Unfortunately, limited data are available on effective methods to integrate EBM curric ulum into training programs or clinical practice. One might assume that the presence of well-trained faculty, access to the National Library of Medicine and other databases, availability of computer resources in distant learning sites, and faculty/resident buy-in might all be necessary for an EBM curriculum to function effectively. However, this is unproven. Even with an understanding of these barriers, we could not identify any specific CD or clerk-shop characteristics--other than the number of clinic half-days per week that predicted the incorporation of a formal EBM curriculum into the medicine clerkship. These included characteristics such as the CD's academic rank, length of time as CD, length of the clerkship, the number of clerkship sites, or scholarly activities. This finding might have several reasons. First, the survey inquired only about the use of a formal curriculum in EBM. A small percentage of clerkships without a formal EBM curriculum did report offering sessions on critical appraisal skills t o their students, and hence we may not have identified other clerkship projects or initiatives that involve teaching principles of EBM, without the use of a formal EBM curriculum. In addition, we asked only those clerkships with a formal EBM curriculum to identify the barriers they encountered implementing EBM, and thus we missed an opportunity to determine the perceptions of clerkships without EBM. This may have shed light on other perceptions about or barriers to initiating EBM curricula. Nevertheless, from the few CDs without EBM (n = 7) who did respond to the question, and the consistency of our findings with other studies, (10) it seems likely that the barriers of adequately trained faculty and time would be common ones. Finally, although CDs who incorporated EBM spent one more half-day of clinic per week than those without EBM, we are uncertain about the importance of this difference. It could be that these CDs are more directly involved with ambulatory student education and emphasize the value and role of EBM in this setting. This finding might also simply be a statistically significant but clinically insignificant finding.

Among the goals for incorporating EBM into the clerkship curriculum is that these skills enable the learner to identify a clinical question, identify literature that best answers the question, critically appraise the gathered data, and effectively apply it to real life scenarios. To that end, most of the clinical questions used in the clerkship EBM curricula were developed by students who used direct patient interactions and written patient cases. This active involvement is a key component of adult learning theory. Multiple evaluation methods were used to assess student learning and application of knowledge, including overall critical appraisal skill exercises, multiple choice questionnaires, and essay questions. The evaluation strategies used measure various levels of cognitive ability from basic understanding, as with multiple choice questions, to synthesis and applications as measured by the critical appraisal skill exercises.

It is important to remember that the process of learning ultimately results in changing behavior, based on experiences. Although several evaluation methods have been used to assess students' understanding of EBM principles and application to clinical scenarios, our survey did not address whether any change in student behavior was effected. However, a previous study of interactive seminars on critical appraisal skills showed that students exposed to the seminars improved their knowledge of EBM principles but did not improve their use of these skills in their appraisals of their patients when compared with controls. (11) Although that study was of short duration, it highlights the critical need for programmatic evaluation of EBM initiatives and establishing appropriate outcomes for measuring behavior change after an intervention.

In addition to what has already been mentioned, this survey has other limitations. First, the survey was designed primarily to gather data on the current implementation of EBM in internal medicine clerkships. Even though this survey reviewed the presence of EBM in pre-clinical years, it was not specifically designed to include EBM that may have been informally taught during courses such as epidemiology. Second, the survey was not designed to assess the success of EBM programs at enhancing knowledge or changing behavior or to examine the effectiveness of any given evaluation method. Future research in this area needs to focus on methods of student evaluation as well as programmatic evaluation, including short-term and long-term follow-up of these students. Third, the survey did not inquire about the nature and consistency of EBM curricular goals and objectives among the clerkships. Consistency of goals among internal medicine clerkships may help broaden the implementation of EBM and help ensure that students r eceive similar training and skills. The variety of curricular and assessment methods might reflect a lack of clear EBM goals and objectives for graduating medical students. Finally, as with any survey that requires multiple choice or numeric answers, there is the potential for bias in the responses, since CDs may have been constrained in describing their curriculum. In addition, this type of survey allows for reporting without consideration to quality or effectiveness. However, given the nature of the survey questions, we believe this likely had little impact on the results.

Incorporation of EBM in the medical school curriculum is still in its infancy. Few would argue the value and necessity of learning and applying critical appraisal skills but this requires faculty time and development as well as access to resources. Integrating curricula across disciplines and across the clinical years of medical education may help enhance and reinforce lifelong learning skills, but carefully designed studies are needed to assess changes in practice behavior.

TABLE 1.

Selected Demographic Data for Internal Medicine Clerkship Directors and
Clerkships With and Without a Formal Evidence-Based Medicine (EBM)
Curriculum

                                               With EBM
        Variable                          (Mean [+ or -] SD)

Age                                   44.9 [+ or -] 7.8 (n = 41)
Years as clerkship director            6.1 [+ or -] 5.4 (n = 40)
Percent time spent on clerkship       30.9 [+ or -] 14.3 (n = 38)
Number clinic half-days per week       3.6 [+ or -] 1.8 (n = 39)
Number months on inpatient service     2.8 [+ or -] 1.8 (n = 41)
Number of other courses directed       1.2 [+ or -] 1.0 (n = 41)
Number peer-reviewed papers            1.8 [+ or -] 2.5 (n = 41)
Number of awarded grants               0.78 [+ or -] 1.3 (n = 41)
Length of clerkship (weeks)           11.29 [+ or -] 6.1 (n = 42)
Number of students per year          118.1 [+ or -] 49.4 (n = 42)
Number of clerkship sites
    Inpatient sites
[less than or equal to]4                        n = 33
[greater than or equal to]5                     n = 9
    Outpatient sites
[less than or equal to]5                        n = 14
[greater than or equal to]6                     n = 20
Clerkship composition
 Separate ambulatory                            n = 23
 Inpatient combined with ambulatory             n = 16

                                             Without EBM
        Variable                          (Mean [+ or -] SD)

Age                                   45.6 [+ or -] 7.6 (n = 64)
Years as clerkship director            7.0 [+ or -] 5.1 (n = 65)
Percent time spent on clerkship       26.4 [+ or -] 12.6 (n = 61)
Number clinic half-days per week       2.9 [+ or -] 2.0 (n = 65)
Number months on inpatient service     2.9 [+ or -] 2.1 (n = 64)
Number of other courses directed       0.92 [+ or -] 1.0 (n = 65)
Number peer-reviewed papers            2.3 [+ or -] 4.0 (n = 64)
Number of awarded grants               0.57 [+ or -] 1.0 (n = 65)
Length of clerkship (weeks)           10.45 [+ or -] 2.3 (n = 65)
Number of students per year          127.9 [+ or -] 53.0 (n = 65)
Number of clerkship sites
    Inpatient sites
[less than or equal to]4                        n = 51
[greater than or equal to]5                     n = 14
    Outpatient sites
[less than or equal to]5                        n = 29
[greater than or equal to]6                     n = 32
Clerkship composition
 Separate ambulatory                            n = 43
 Inpatient combined with ambulatory             n = 21


        Variable                     P Value

Age                                   .61 *
Years as clerkship director           .36 *
Percent time spent on clerkship       .12 *
Number clinic half-days per week      .05 *
Number months on inpatient service    .83 *
Number of other courses directed      .08 *
Number peer-reviewed papers           .10 *
Number of awarded grants              .56 *
Length of clerkship (weeks)           .32 *
Number of students per year           .34 *
Number of clerkship sites
    Inpatient sites
[less than or equal to]4              1.0 +
[greater than or equal to]5
    Outpatient sites
[less than or equal to]5              .67 +
[greater than or equal to]6
Clerkship composition
 Separate ambulatory                  .41 +
 Inpatient combined with ambulatory

* Mann-Whitney test.

+ Fisher exact test.

Results from annual survey of internal medicine clerkship directors
(CDs) conducted by the Clerkship Directors in Internal Medicine, 1999. A
total of 109 CDs (89%) responded to the survey; numbers in table may not
add up to 109, since not everyone answered every question or portion of
the survey.

Number clinic half-days per week refers to the number of outpatient
clinics the respondent participated in each week. Number of other
courses directed refers to whether the CD directed preclinical or
clinical courses other than the core internal medicine clerkship. Number
peer-reviewed papers and Number of awarded grants refers to the number
of peer-reviewed articles published or grants received in the 4 years
preceding the survey, respectively. Except for the number of half-days
in clinic per week, this Table summarizes that no specific demographic
variables of CDs or the clerkship itself was associated with the use of
a formal EBM curriculum in the third year internal medicine clerkship.

TABLE 2

Characterisitcs of Formal Evidence- Based Medicine (EBM) Curricula Among
Internal Medicine Clerkships

                                    All with
                                    EBM *
                                    (n = 42)

Number of years EBM in curriculum   2.7 [+ or -] 2.5 (n = 38)
 (mean [+ or -] SD)
Number of faculty teaching EBM      3.5 [+ or -] 5.3 (n = 32)
 (mean [+ or -] SD)
Number of clerkship hours devoted   6.7 [+ or -] 4.9 (n = 37)
 to EBM  (mean [+ or -] SD)
Number of hours used to teach       3.3 [+ or -] 3.3 (n = 33)
 critical appraisal  (mean
 [+ or -] SD)
Setting in which EBM is taught
   Inpatient                        38.5%
   Outpatient                       20.5%
   Both                             41.0%
Who develops EBM questions
   Faculty                          25%
   Students                         40%
   Both                             27.5%
   Other                             7.5%
Experiences used for EBM questions
   Patients                         54.1%
   Patient scenarios                21.6%
   Both                             18.9%
   Other                             5.4%
Formal instruction on conducting
 literature search
   Yes                              82.9% (34)
Who teaches literature search +
   Faculty                          58.8%
   Librarian                        67.6%
   Information specialist            5.9%
   Other                             5.9%
Academic year EBM first introduced
 to students
   M1                               44.7%
   M2                               23.7%
   M3                               15.8%
   M4                                2.6%
   Uncertain                        13.2%

                                    EBM in Third
                                      Year Only
                                      (n = 28)

Number of years EBM in curriculum   2.5 [+ or -] 2.1 (n = 26)
 (mean [+ or -] SD)
Number of faculty teaching EBM      3.3 [+ or -] 5.8 (n = 24)
 (mean [+ or -] SD)
Number of clerkship hours devoted   6.8 [+ or -] 5.3 (n = 26)
 to EBM  (mean [+ or -] SD)
Number of hours used to teach       3.5 [+ or -] 3.7 (n = 24)
 critical appraisal  (mean
 [+ or -] SD)
Setting in which EBM is taught
   Inpatient                        30.8%
   Outpatient                       26.9%
   Both                             42.3%
Who develops EBM questions
   Faculty                          22.2%
   Students                         48.1%
   Both                             22.2%
   Other                             7.4%
Experiences used for EBM questions
   Patients                         56.0%
   Patient scenarios                28.0%
   Both                              8.0%
   Other                             8.0%
Formal instruction on conducting
 literature search
   Yes                              85.7% (24)
Who teaches literature search +
   Faculty                          52.2%
   Librarian                        65.2%
   Information specialist            8.7%
   Other                             8.7%
Academic year EBM first introduced
 to students
   M1                               46.2%
   M2                               23.1%
   M3                               19.2%
   M4                                0%
   Uncertain                        11.5%

* Includes clerkships with formal EBM curriculum in third year (n = 28),
fourth year (n = 4), and both third and fourth year (n = 10).

+ Numbers do not add up to 100% since respondents listed all who taught
literature search course.

Results from annual survey of internal medicine clerkship directors
(CDs) conducted by the Clerkship Directors in Internal Medicine, 1999.
Numbers reflect those CDs who reported the presence of a formal EBM
curriculum during an internal medicine clerkship. Some CDs did not
answer all questions on the survey. For convenience, those with formal
EBM curricula in the third year, the fourth year, and both the third and
fourth years were collapsed into the category, "All with EBM."

TABLE 3

Evaluation Methods Used for Evidence-Based Medicine Curricula as
Reported by Internal Medicine Clerkship Directors

                              All with   EBM in Third
                                EBM       Year Only
                              (n = 42)     (n = 28)

Multiple choice questions    15.6% (5)    16.7% (4)
Critical appraisal exercise  53.1% (17)   50.0% (12)
Papers                       21.9% (7)    25.0% (6)
Essay                         9.4% (3)     4.2% (1)
Other                        21.9% (7)    20.8% (5)

Results from annual survey of internal medicine clerkship directors
(CDs) conducted by the Clerkship Directors in Internal Medicine, 1999.
Numbers at top of columns reflect those CDs who reported the presence of
a formal EBM curriculum during an internal medicine clerkship. Some CDs
did not answer all questions on the survey. For convenience, those with
formal curricula in the third year, the fourth year, and in the and in
the third and fourth years were collapsed into the category, "All with
EBM." Percentages (numbers) represent those CDs who stated they used the
method. Numbers do not add to 100% since CDs may have used more than one
method.

The results show that clerkship most frequently use critical appraisal
exercises and paper cases to evaluate students on their knowledge and
understanding of the EBM curriculum.

Acknowledgments. We thank Yuhuku Yamada for data entry and follow-up with CDIM members, and the CDIM staff, particularly Trudie Bruner and Tod Ibrahim for supervision of the survey process.

References

(1.) Goroll AH, Morrison G: core medicine clerkship curriculum guide: a cooperative project of the Society of General Internal Medicine and the clerkship Directors in Internal Medicine. SGIM/CDIM, Health Resources and Service Administration, March 1995

(2.) Bordley D, Fagan M, Theige D: Evidence based medicine: a powerful educational tool for clerkship education. Am J Med 1997; 102:427-432

(3.) Sackett D, Rosenberg WMC, Gray JAM, et al: Evidence based medicine: what it is and what it isn't. BMJ 1996; 312:71-72

(4.) Evidence Based Medicine working Group: Evidence based medicine: a new approach to teaching the practice of medicine. JAMA 1992; 268:2420-2425

(5.) Accreditation council for Graduate Medical Education: Program requirements for residency education in internal medicine: special educational requirements. The Graduate Medical Education Directory, 1996-1997. Chicago Ill, American Medical Association, 1996, p 79

(6.) Physicians for the twenty-first century: Report of the project panel on the general professional education of physician and college preparation for medicine. J Med Educ 1984; 59(pt2):127-128,155-167

(7.) Green ML: Evidence based medicine training in internal medicine residency programs: a national survey. J Gen Intern Med 2000; 15:129-133

(8.) Frasca MA, Dorsch JL, Aldag JC, et al: A multidisciplinary approach to information management and critical appraisal instruction: a controlled study. Bull Med Libr Assoc January 1992, p 80

(9.) Hemmer PA. Elnicki DM, Albritton TA, et al: The responsibilities and activities of internal medicine clerkship directors. Acad Med 2001; 76:715-721

(10.) Ramsey PG, Carline JD, Inui TS, et al: Changes over time in knowledge base of practicing internists. JAMA 1991; 266:1103-1107

(11.) Landry FJ, Pangaro L, Kroenke K, et al: A controlled trial of a seminar to improve medical student attitudes toward, knowledge about, and use of the medical literature. J Gen Intern Med 1994; 9:436-439

KEY POINTS

* Evidence-based medicine (EBM) can be used as an educational strategy to develop lifelong learning skills.

* A minority of internal medicine clerkships report having a formal EBM component in their curriculum.

* Faculty development, time, and access to resources are needed to implement EBM.

* Standardized evaluation tools and effect of change in student behavior as a result of EBM has not been described in the literature.

* Carefully designed studies are needed to assess the change in student behavior patterns.

From the Department of Internal Medicine and the Division of Educational Services, University of Illinois College of Medicine at Peoria; the Department of Medicine, Uniformed Services University of the Health Sciences, Bethesda, Md; the Department of Medicine, Medical College of Georgia, Augusta; the Department of Medicine, Albert Einstein College of Medicine, Bronx, NY; and the Department of Internal Medicine, University of Chicago Pritzker School of Medicine, Chicago, Ill.

The opinions expressed in this paper are solely those of the authors and do not necessarily reflect the opinions of the Department of Defense, the United States Air Force, or other federal agencies.

Data from the CDIM survey are the property of the CDIM and have been used with permission.

Reprint requests to Meenakshy Aiyer, MD, 530 NE Glen Oak Aye, Room 3420, Peoria, IL 61637.