Medical Teacher
ISSN: 0142-159X (Print) 1466-187X (Online) Journal homepage: http://www.tandfonline.com/loi/imte20
Structured clinical teaching strategy Sharon S. Allen, Carole J. Bland, Ilene B. Harris, David Anderson, Gregory Poland, Leon Satran & Wesley Miller To cite this article: Sharon S. Allen, Carole J. Bland, Ilene B. Harris, David Anderson, Gregory Poland, Leon Satran & Wesley Miller (1991) Structured clinical teaching strategy, Medical Teacher, 13:2, 177-184, DOI: 10.3109/01421599109029028 To link to this article: http://dx.doi.org/10.3109/01421599109029028
Published online: 03 Jul 2009.
Submit your article to this journal
Article views: 44
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=imte20 Download by: [McMaster University]
Date: 10 April 2016, At: 23:36
Medical Teacher, Vol. 13, No. 2, 1991
177
Downloaded by [McMaster University] at 23:36 10 April 2016
Structured clinical teaching strategy
SHARON S. ALLEN, CAROLE J. BLAND, ILENE B. HARRIS,’ DAVID ANDERSON: GREGORY POLAND? LEON SATRAN: WESLEY MILLER5, ‘Medical School, University of Minnesota, 2Department of Neurology, Hennepin County Medical Center, Minneapolis, Minnesota, 3Division of General Internal Medicine, Mayo Clinic, Rochester, Minnesota, 4Department of Pediatrics, Medical School, University of Minnesota Q SDepanment of Medicine-Hematology, Medical School, University of Minnesota
SUMMARY This study investigated the impact of a structured exercise on low back pain, as part of a second year ambulatory course, on students’ low back pain examination skills. One-hundred and eighty-eight medical students participated in one of four types of instructional intervention: 1) smrctured clinical exercise and reading, 2) random clinical experience and no reading, 3) reading only, and 4) no clinical experience or reading. At the end of the year, students completed an Objective Structured Clinical Exam (OSCE) in which two stations assessed back pain history and physical exam skills. A n analysis of variance of the OSCE scores showed no significant difference in students’performance in relation to the type of instructional intervention. The General Professional Education of the Physician Report argues for the importance of structured clinical education. In medical education, unpredictable patient exposure and crowded curricula frequently result in students having none or only one structured learning opportunity to acquire critical skills. Unfortunately, this study found that assuring at least one structured clinical experience for a specific common problem seen in ambulatory care did not enhance student ability to select and use specific history or physical exam skills for that problem as assessed by an OSCE, as compared with students who did not have this expen’ence. To assure that essential clinical skills are acquired, it most likely requires that both systematic instructional strategies and repeated learning opportunities are available to reinforce learning.
Introduction Medical students’ initial experiences in clinical medicine profoundly affect their
Downloaded by [McMaster University] at 23:36 10 April 2016
178
Sharon S. Allen et al.
professional development. The General Professional Education of the Physican Report (GPEP), published in 1984, discussed the need for clinical education to be carefully structured, and for medical faculty to identify specific clinical knowledge, skills, values, and attitudes that students should develop during their professional education. The GPEP report, therefore, implies that systematic and structured instructional and evaluation strategies will enhance learning. It also describes the types of clinical settings, supervision, and evaluation of clinical performance which would provide that structure (GPEP Report, 1984). Incorporating structured learning and meaningful objective evaluation of clinical competence into clinical courses is, however, both difficult and challenging, especially in the ambulatory setting, because of the characteristic one-on-one learning process, variability in patient experience, and problems in obtaining reliable and valid evaluations of clinical performance in this setting. Students’ clinical performance is usually evaluted by faculty who use rating scales to assess global skills such as history taking and physical examination skills. Such ratings are often not accurate or reliable (Bowman, 1982). Observational studies of clincial teaching (Mattern, Weinholtz & Freidman, 1984; Payson & Barchas, 1965) have shown that most faculty members do not directly observe students’ clinical performance. In response to the GPEP report we attempted to standardize a part of students’ clinical work by designing a structured clinical exercise on a specific problem (back pain) during their Family Medicine ambulatory experience. In turn, we wanted to determine if such a structured clinical exercise affects performance of students in their approach to the same clinical problem. Several investigators (Petrusa et al., 1987; Harden ef al., 1975; Harden, 1979; Noble, Hoare & Elmslie, 1981; Norman, 1985) have described the use of objective structured clinical examinations (OSCE’s) as a means of increasing precision of clinical competence evaluation. We, therefore, used the OSCE to assess students’ clinical competence in this area, and to evaluate the success of our instructional strategy. We hypothesized that students who completed a structured clinical exercise in back pain, specifically designed to improve history taking and physical exam skills needed for this particular problem, would perform better on OSCE stations testing these clinical skills than students who did not participate in the exercise; thus indicating that structured clinical learning improves performance, as suggested in the GPEP report. Methods
Subjects The entire class of medical students (N=188) at the University of Minnesota participated in the study at the end of Year Two. These students had learned physical diagnosis skills in two introductory clinical courses (Clinical Medicine I and 11) at the end of Year One and the start of Year Two. These introductory courses were followed in Year Two. by four 6 week clinical rotations (Clinical Medicine Ill) in Internal Medicine, Family Medicine, Neurology and Pediatrics, consisting of 2 half-days per week during which clinical skills were refined in small groups and individually with a physician tutor. Students were randomly assigned to rotation sequence and clinical site. In Family Medicine, a required structured exercise was selected by students from one of five common clinical problems presenting in their clinics. The choices were: 1) headache, 2) cough, 3) hypertension, 4) fatigue, and 5) low back pain. Fifty-seven (57) out of 188 students (30%) chose back pain for their exercise.
Structured clinical teaching strategy
179
Downloaded by [McMaster University] at 23:36 10 April 2016
internention For this study, four types of instructional situations for the clinical problem of low back pain were defined: 1) structured clinical exercise and focused reading (the intervention which is the subject of this paper); 2) random clinical experience and no reading; 3) reading only; or 4) no clinical experience (random or structured) and no reading. As described above, students could select to participate in the low back pain structured experience group. Students fell into one of the three other groups as a result of random exposure to back pain problems during their ambulatory experience in Family Medicine, or personal reading, or neither; and were classified based on a student self-report of learning opportunities they had on back pain collected via a survey given immediately after the OSCE administration. The intervention which is the focus of this study, the structured back pain exercise, consisted of readings and application of the concepts and skills described in the readings with a patient in their clinic. The readings provided specific information to help students elicit a directed history. Algorithms were provided in the readings related to three questions, intended to be helpful and directive in identifying the etiology of back pain: 1) Where is the pain? 2) Is the back pain acute or chronic? and 3) Is the pain mechanical or nonmechanical? Interviewing with the aid of the algorithms would help students classify low back pain into three general subsets of mechanical low back pain: 1) ominous low back pain, 2) radicular low back pain, and 3) simple mechanical low back pain (Reilly, 1984). The readings also focused on (emphasized) examining patients with low back pain in five major aspects: 1) observation, 2) evaluation of motion of the back, 3) palpation of the spine and paraspinal structures, 4) traction maneuvers, and 5) neurological exam (Reilly, 1984). Readings were reinforced by clinical experience; each participating student identified a patient with low back pain, developed a reasoned hypothesis regarding back pain etiology, and identified specific cues in the history and exam to support this hypothesis. On-site preceptors provided feedback, although no formal checklist of appropriate questions or exam-skills was used by preceptors. In addition, a case report on their back pain patient emphasizing appropriate questions and/or skills was turned in by the student and graded by a faculty tutor (not the student’s clinical preceptor) using a rating sheet which provided students with feedback on appropriate identification of cues for back pain in history and exam and problem oriented record format. This rating sheet was mailed to the student within 2 weeks. It should be noted that this rating sheet did not assess the same specific items as were used on the OSCE checklists. The second group is labeled “random clinical experience and no reading”. Some students who did not choose back pain as their specific structured exercise still examined back pain patients in the ambulatory setting with their tutor. This experience was not structured, as described above, and was left up to the circumstances of the clinical setting. All students had available to them the packet of materials on low back pain used by the intervention group. However, we know by our records that only students in the intervention group picked up packets at the departmental office. In this second group no write-ups were turned into faculty, no checklist was done and the student received no formal feedback. Forty-four (27%) and 39 (24%) of students reported a random clinical experience for back pain history and physical exam respectively. The third group labeled “reading only” includes students who read about back pain but did not examine patients with that complaint. Thirty-three (18%) of
180
Sharon S.Allen et al.
students were included in this group. Students were informed that back pain would be on the OSCE examination. They therefore had the opportunity to read about back pain prior to the exam. This may account for this group of students who reported they had read about back pain but had not picked up packets on back pain and had no clinical experience. The fourth intervention group, “no clinical experience and no reading” includes students who did not do the structured exercise and reported no random clinical experience and no reading on back pain problems. Thirty (17%) students reported no opportunity for learning about the history of back pain; and 37 (21%) students reported no opportunity for learning about the physical exam of back pain.
Downloaded by [McMaster University] at 23:36 10 April 2016
Objective structured clinical exam An OSCE, comprised of 22, 5 minute stations, was administered near the end of Year Two to evaluate student progress in acquiring clinical skills, and related to content from each of the four specialty rotations in Clinical Medicine 111. Two of the Family Medicine stations were designed to test history and physical examination skills dealing with the problem of back pain. In order to accommodate 188 students in a one-day period, students were randomly assigned to one of three 2 hour testing sessions. During each session the OSCE was run simultaneously in three locations. This resulted in the use of three raters and three simulated patients for each station. The same rater and patient staffed a station at each location for all sessions. Simulated patients and examiners were recruited from 3rd and 4th year medical students, as well as from Family Medicine faculty, who were not involved in the structured exercise (and therefore were not aware of which student examinees did the structured clinical back pain exercise). The back pain history and physical exam stations were set up to simulate an office examining room. The simulated patient portrayed a middle-aged man with low back pain due to a herniated disc. Students at each station were informed that the patient had low back pain. In the history station, they were instructed to ask the patient direct questions in order to obtain a detailed history of the problem and to then inform the rater of their diagnosis. In the physical exam station, they were instructed to perform a limited exam specific to the history of back pain, tell the rater reasons for examination manipulations, and to communicate the findings of the exam. The rater was instructed to have no discussion or comment on the student’s report during the stations other than to record appropriate information on the rating sheet. All feedback on stations was provided to students at the end of the entire OSCE exam. Simulated patients and raters were trained by faculty at orientation sessions. History patients and physical exam patients were given detailed instructions as to specific details for simulation of herniated disc. For the history station, raters used an objective checklist to rate student performance on specific history-taking skills as “satisfactory,” “attempted but not satisfactory,” or “not attempted”. Students were rated on their inquiry about specific characteristics of the low back pain, including: onset, course, duration, site, radiation, character, severity, aggravating factors, relieving factors, numbness, weakness, and the effects of activity, bowel movements, cough, straining or micturition. These history items were discussed in readings and were described as helpful in determining the etiology of back pain (Reilly, 1984). The student was asked to provide a diagnosis at the closure of the history and that answer was recorded. Student performance in the physical exam station was rated in the following areas corresponding to the skills
Structured clinical teaching strategy
181
taught in the structured clinical exercise: 1) observation, 2) range of motion, 3) palpation, 4) traction maneuvers, and 5) neurological exam. Student scores for the entire class of 188 are detailed in Table I and Table 11. The total possible score for each station was 45 points.
TABLE I. Percent of students correctly performing specific back pain history items ( N = 188)
Downloaded by [McMaster University] at 23:36 10 April 2016
Satisfactory Onset Course Duration Site Radiation Character Severity Aggravating factors Relieving factors Numbness Weak n ess Movement/activity Bowel movement Coughlstraining Micturition
94% 78% 92% 90% 85% 85% 63% 73% 90% 38% 24% 82%
11% 7% 7%
TABLE 11. Percentage of students correctly performing specific back pain exam items ( N = 188) Satisfactory Observation ROM flex/ext ROM latlflex ext Palpation Straight leg raising Hip flexion Hip extension Hip adduction Hip abduction Hip internal rotation Hip external rotation Reflex, knee Relfex, ankle Reflex, Babinski Foot dorsiflexion Foot plantar flexion Foot eversion Foot inversion Dorsiflexion, big toe
74% 72%
51% 76% 78% 53% 87% 13% 17% 26% 28% 94% 91% 52% 22% 20% 7% 7% 2%
182
Sharon S.Allen et al.
Downloaded by [McMaster University] at 23:36 10 April 2016
Data analysis Several concerns were addressed before the analysis was done. One concern was selfselection of students for the structured clinical exercise. Since, as reported in the next section, the self-selected group did not perform better than the other groups, it is unlikely that the self-selected group was “ a prior?’ different than the other groups, based on such factors as interest in the problem of back pain or perceived strengths or weaknesses in skills. One could hypothesize that the self-selected group was less able before the structured exercise, and through it brought themselves to the level of students who had no exposure to readings or patients in the area. However, this is unlikely. Another concern was that significant differences were found among OSCE sites on the average score for each station. Since students were randomly assigned to OSCE sites, the significant differences in mean ratings across sites suggested some inter-rater unreliability across sites and that more training of patients and raters is necessary to achieve comparable results using multiple raters in different sites. As described above, the same rater and simulated patient were used the entire time at each site. There were no significant differences in mean ratings within sites over the three OSCE time-slots and this suggested intra-rater reliability within each site. Therefore, within each site, an analysis of variance was used to compare average student scores on the back pain histories and physical exams for the four levels of instructional intervention. Also, Chi Square analyses were performed for each specific item on the history and physical examination checklist. One person, the first author, read the written diagnoses by all students at the end of the history and categorized them medically as to be more differentiated, meaning nerve involvement, vs. less differentiated, meaning low back strain, mechanical strain or muscle strain. A Chi Square was then performed comparing the level of diagnosis and the instructional level. This time the instructional level was grouped as structured learning group vs. other groups combined. Thirty-eight students were not used in this analysis because of inability to track the original choice exercise since exercises had been mailed back to the students.
Results There were no statistically significant differences in students’ performances on each OSCE station, overall, in relation to the types of learning situation (Table 111). In addition, Chi Square analyses show no significant differences for each item on the history and the physical examination checklist in relation to learning situation. Across sites and intervention groups, mean scores on the physical examination for back pain were consistently lower than mean scores for the back pain history. An inspection of the specific history items revealed that students did well in obtaining the basics of the history, including onset of the pain, course, duration, site, radiation, character, severity, and aggravating and relieving factors. However, they frequently failed to address neurological areas such as numbness, weakness, effects on bowel movement and micturition (Table I). There were no significant differences between the structured learning group, and the other groups combined, in their ability to give a diagnosis at the close of the back pain history OSCE station. Seventy-nine percent (79%) of students (45157) who did the structured back pain exercise mentioned a nerve involvement as their diagnosis at the end of the history, while 12/57 (21.1%) mentioned a less differentiated diagnosis such as low back strain, mechanical strain or
Structured clinical teaching strategy
183
TABLE 111. Mean OSCE scores at three sites for back pain exam and history according to self-report of learning experience* **(N=173 for history; N = 1 7 5 for exam) ~~
Site 1
Downloaded by [McMaster University] at 23:36 10 April 2016
Learning Experience
Site 2
Site 3
Hx
Px
Hx
Px
Hx
Px
None
31.8 N=9
23.5 N=10
28 N=ll
20 N=12
26.3 N=10
17.1 N=15
Reading Only
25.8 33.1 N = 10 N = 1 2
26.4 N=8
18.2 N=9
27.3 N=15
17 N=12
Random clinical Experience
32.6 21.1 N = 19 N = 1 6
28.9 20.4 N = 12 N = 10
24.5 N = 13
16.6 N=13
Structured Experience
31.9 N=18
24.2 N=18
31.1 N=24
20.1 N=24
26.1 N=15
17 N=15
32.3
23.5
29.3
19.8
26.1
16.9
Average Mean
**Fifteen students failed to report information, *Total possible score=45.
muscle strain. Of the 79%, several even localized the nerve involvement to L-4, L-5, or S-1. This diagnosis was made in light of the fact that students failed to address neurological areas in their history (Table I). Of students who did not do the back pain exercise, 79% of them (73193) also mentioned nerve involvement as their diagnosis at the end of the history. Twenty-two percent (20193) had less differentiated types of diagnoses. In addition, basic parts of the back physical exam-observation, palpation, range of motion and straight leg raising were consistently performed, but a large percentage of the class neglected important neurological components of the back examination and some of the musculoskeletal items, such as the hip exam. Two neurological items, knee reflex and ankle jerks were done more consistently (Table 11). A large percentage of the students who did the back pain exercise and who did not, 79% respectively, mentioned nerve root involvement in their diagnosis and yet failed to do important parts of the neurological back pain exam.
Discussion The low mean scores on the history of back pain, and even lower mean scores on the physical examination of back pain, for students in all intervention groups, suggests that students have not learned to conduct a thorough history and examination for back pain utilizing the principles discussed above. This is surprising, especially for the group who participated in the exercise, since the readings did emphasize appropriate neurological questions and areas for neurological physical examination associated with mechanical low back pain identified as radicular low back pain. A large percentage of students (79%), regardless of learning strategy, could accurately determine the etiology of the back pain at the history station and mention nerve root involvement and yet a large percentage of all instructional groups did not then proceed to do appropriate neurological questions and physical exam. This suggests that a single clinical structured exercise has little impact. It is puzzling that students could accurately hypothe-
Downloaded by [McMaster University] at 23:36 10 April 2016
184
Sharon S. Allen et al.
size a neurological problem but fail to perform appropriate questioning and physical examination. It appears from this study that a single structured instructional strategy (including specific readings, clinical experience with immediate feedback from their preceptor, a prepared write-up on the clinical experience with written feedback from faculty) is not adequate to ensure student learning pertinent history and physical exam skills even though many clinical courses do not include even one such strategy for many important clinical problems. Teaching students how to use a specific data collection strategy and when to use it (most likely) requires supplementary instructional strategies. In conclusion, this study assessed an effort to follow important recommendations of the GPEP report: to identify some specific clinical knowledge and skills needed by students and to provide a structured learning exercise designed to help students acquire (develop) these skills. Given the time constraints in medical school curricula, most curricula do not include even one such structured learning opportunity for important medical problems. Unfortunately, however, this study was unable to demonstrate that a single structured clinical exercise for a specific common problem seen in the ambulatory setting enhanced students’ ability to select and use specific history or examination skills for that problem. These observations are important since time and information overload frequently dictate limited learning opportunities for important clinical skills. It would seem incumbent on medical educators to identify important skills and knowledge and to design curricula which will allow multiple longitudinal integration of structured experiences to reinforce these skills and knowledge areas.
Correspondence: Sharon S . Allen, MD, PhD, Assistant Professor, Department of Family Practice and Community Health, University of Minnesota, 6-240 Delaware Street SE, Box 381 UMHC, Minneapolis, Minnesota, 55455, USA. REFERENCES ASSOCIATIONOF AMERICAN MEDICALCOLLEGES (1984) Physicians for the Twenty-First Century. The GPEP Report. Report of the Panel on the General Rofessional Education of the Physician and College Presentation for Medicine (Washington, DC 20036). BOWMAN, W.C. (1982) Evaluating performance effectiveness on the job: how can we generate more accurate ratings? in: LLOYD,J.S. (Ed.) Evaluation of non-cognitive skills and clinical performance (Chicago, American Board of Medical Specialties). F.A. (1979) Assessment of clinical competence using an objective structured HARDEN,R.M. & GLEESON, clinical examination (OSCE), Medical Education, 13, pp. 41 -54. HARDEN,R.Mc.G., STEVENSON, M., DOWNIE,W.W. & WILSON,G.M. (1975) Assessment of clinical competence using Objective Structured Examination, British Medical Journal, 1, pp. 447-45 1. MARERN, W.D., WEINHOLTZ, D. & FREIDMAN, C.P. (1984) The attending physician as teacher, New England Journal of Medicine, 208, pp. 1129-1 132. NEWBLE, D.I., HOARE,J. & ELMSLIE, R.G. (1981) The validity and reliability of a new examination of the clinical competence of medical students, Medical Education, 15, pp. 46-52. NORMAN, G.R. (1985) Objective measurement of clinical performance, Medical Education, 19, pp. 43-47. PAYSON, H.E. & BARCHAS, J.D. (1965) A time study of medical teaching rounds, New England Journal of Medicine, 237, pp. 1468-1471. PERKOFF, G.T. (1986) Teaching clinical medicine in ambulatory setting, New England Journal of Medicine, 314(1), pp. 27-31. PETRUSA, E.R., BLACKWELL, T.A., ROGERS,L.P., SAYDJARI, C., PARCEL,S. & GUCKIAN, J.C. An objective measure of clinical performance, American Journal ofhfedicine, (83) (July), pp. 34-41. REILLY,B.M. (1984) Low back pain, in: Practical Strategies in Out Patient Medicine, pp. 1-70 (Philadelphia, W.B. Saunders).
ISSN: 0142-159X (Print) 1466-187X (Online) Journal homepage: http://www.tandfonline.com/loi/imte20
Structured clinical teaching strategy Sharon S. Allen, Carole J. Bland, Ilene B. Harris, David Anderson, Gregory Poland, Leon Satran & Wesley Miller To cite this article: Sharon S. Allen, Carole J. Bland, Ilene B. Harris, David Anderson, Gregory Poland, Leon Satran & Wesley Miller (1991) Structured clinical teaching strategy, Medical Teacher, 13:2, 177-184, DOI: 10.3109/01421599109029028 To link to this article: http://dx.doi.org/10.3109/01421599109029028
Published online: 03 Jul 2009.
Submit your article to this journal
Article views: 44
View related articles
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=imte20 Download by: [McMaster University]
Date: 10 April 2016, At: 23:36
Medical Teacher, Vol. 13, No. 2, 1991
177
Downloaded by [McMaster University] at 23:36 10 April 2016
Structured clinical teaching strategy
SHARON S. ALLEN, CAROLE J. BLAND, ILENE B. HARRIS,’ DAVID ANDERSON: GREGORY POLAND? LEON SATRAN: WESLEY MILLER5, ‘Medical School, University of Minnesota, 2Department of Neurology, Hennepin County Medical Center, Minneapolis, Minnesota, 3Division of General Internal Medicine, Mayo Clinic, Rochester, Minnesota, 4Department of Pediatrics, Medical School, University of Minnesota Q SDepanment of Medicine-Hematology, Medical School, University of Minnesota
SUMMARY This study investigated the impact of a structured exercise on low back pain, as part of a second year ambulatory course, on students’ low back pain examination skills. One-hundred and eighty-eight medical students participated in one of four types of instructional intervention: 1) smrctured clinical exercise and reading, 2) random clinical experience and no reading, 3) reading only, and 4) no clinical experience or reading. At the end of the year, students completed an Objective Structured Clinical Exam (OSCE) in which two stations assessed back pain history and physical exam skills. A n analysis of variance of the OSCE scores showed no significant difference in students’performance in relation to the type of instructional intervention. The General Professional Education of the Physician Report argues for the importance of structured clinical education. In medical education, unpredictable patient exposure and crowded curricula frequently result in students having none or only one structured learning opportunity to acquire critical skills. Unfortunately, this study found that assuring at least one structured clinical experience for a specific common problem seen in ambulatory care did not enhance student ability to select and use specific history or physical exam skills for that problem as assessed by an OSCE, as compared with students who did not have this expen’ence. To assure that essential clinical skills are acquired, it most likely requires that both systematic instructional strategies and repeated learning opportunities are available to reinforce learning.
Introduction Medical students’ initial experiences in clinical medicine profoundly affect their
Downloaded by [McMaster University] at 23:36 10 April 2016
178
Sharon S. Allen et al.
professional development. The General Professional Education of the Physican Report (GPEP), published in 1984, discussed the need for clinical education to be carefully structured, and for medical faculty to identify specific clinical knowledge, skills, values, and attitudes that students should develop during their professional education. The GPEP report, therefore, implies that systematic and structured instructional and evaluation strategies will enhance learning. It also describes the types of clinical settings, supervision, and evaluation of clinical performance which would provide that structure (GPEP Report, 1984). Incorporating structured learning and meaningful objective evaluation of clinical competence into clinical courses is, however, both difficult and challenging, especially in the ambulatory setting, because of the characteristic one-on-one learning process, variability in patient experience, and problems in obtaining reliable and valid evaluations of clinical performance in this setting. Students’ clinical performance is usually evaluted by faculty who use rating scales to assess global skills such as history taking and physical examination skills. Such ratings are often not accurate or reliable (Bowman, 1982). Observational studies of clincial teaching (Mattern, Weinholtz & Freidman, 1984; Payson & Barchas, 1965) have shown that most faculty members do not directly observe students’ clinical performance. In response to the GPEP report we attempted to standardize a part of students’ clinical work by designing a structured clinical exercise on a specific problem (back pain) during their Family Medicine ambulatory experience. In turn, we wanted to determine if such a structured clinical exercise affects performance of students in their approach to the same clinical problem. Several investigators (Petrusa et al., 1987; Harden ef al., 1975; Harden, 1979; Noble, Hoare & Elmslie, 1981; Norman, 1985) have described the use of objective structured clinical examinations (OSCE’s) as a means of increasing precision of clinical competence evaluation. We, therefore, used the OSCE to assess students’ clinical competence in this area, and to evaluate the success of our instructional strategy. We hypothesized that students who completed a structured clinical exercise in back pain, specifically designed to improve history taking and physical exam skills needed for this particular problem, would perform better on OSCE stations testing these clinical skills than students who did not participate in the exercise; thus indicating that structured clinical learning improves performance, as suggested in the GPEP report. Methods
Subjects The entire class of medical students (N=188) at the University of Minnesota participated in the study at the end of Year Two. These students had learned physical diagnosis skills in two introductory clinical courses (Clinical Medicine I and 11) at the end of Year One and the start of Year Two. These introductory courses were followed in Year Two. by four 6 week clinical rotations (Clinical Medicine Ill) in Internal Medicine, Family Medicine, Neurology and Pediatrics, consisting of 2 half-days per week during which clinical skills were refined in small groups and individually with a physician tutor. Students were randomly assigned to rotation sequence and clinical site. In Family Medicine, a required structured exercise was selected by students from one of five common clinical problems presenting in their clinics. The choices were: 1) headache, 2) cough, 3) hypertension, 4) fatigue, and 5) low back pain. Fifty-seven (57) out of 188 students (30%) chose back pain for their exercise.
Structured clinical teaching strategy
179
Downloaded by [McMaster University] at 23:36 10 April 2016
internention For this study, four types of instructional situations for the clinical problem of low back pain were defined: 1) structured clinical exercise and focused reading (the intervention which is the subject of this paper); 2) random clinical experience and no reading; 3) reading only; or 4) no clinical experience (random or structured) and no reading. As described above, students could select to participate in the low back pain structured experience group. Students fell into one of the three other groups as a result of random exposure to back pain problems during their ambulatory experience in Family Medicine, or personal reading, or neither; and were classified based on a student self-report of learning opportunities they had on back pain collected via a survey given immediately after the OSCE administration. The intervention which is the focus of this study, the structured back pain exercise, consisted of readings and application of the concepts and skills described in the readings with a patient in their clinic. The readings provided specific information to help students elicit a directed history. Algorithms were provided in the readings related to three questions, intended to be helpful and directive in identifying the etiology of back pain: 1) Where is the pain? 2) Is the back pain acute or chronic? and 3) Is the pain mechanical or nonmechanical? Interviewing with the aid of the algorithms would help students classify low back pain into three general subsets of mechanical low back pain: 1) ominous low back pain, 2) radicular low back pain, and 3) simple mechanical low back pain (Reilly, 1984). The readings also focused on (emphasized) examining patients with low back pain in five major aspects: 1) observation, 2) evaluation of motion of the back, 3) palpation of the spine and paraspinal structures, 4) traction maneuvers, and 5) neurological exam (Reilly, 1984). Readings were reinforced by clinical experience; each participating student identified a patient with low back pain, developed a reasoned hypothesis regarding back pain etiology, and identified specific cues in the history and exam to support this hypothesis. On-site preceptors provided feedback, although no formal checklist of appropriate questions or exam-skills was used by preceptors. In addition, a case report on their back pain patient emphasizing appropriate questions and/or skills was turned in by the student and graded by a faculty tutor (not the student’s clinical preceptor) using a rating sheet which provided students with feedback on appropriate identification of cues for back pain in history and exam and problem oriented record format. This rating sheet was mailed to the student within 2 weeks. It should be noted that this rating sheet did not assess the same specific items as were used on the OSCE checklists. The second group is labeled “random clinical experience and no reading”. Some students who did not choose back pain as their specific structured exercise still examined back pain patients in the ambulatory setting with their tutor. This experience was not structured, as described above, and was left up to the circumstances of the clinical setting. All students had available to them the packet of materials on low back pain used by the intervention group. However, we know by our records that only students in the intervention group picked up packets at the departmental office. In this second group no write-ups were turned into faculty, no checklist was done and the student received no formal feedback. Forty-four (27%) and 39 (24%) of students reported a random clinical experience for back pain history and physical exam respectively. The third group labeled “reading only” includes students who read about back pain but did not examine patients with that complaint. Thirty-three (18%) of
180
Sharon S.Allen et al.
students were included in this group. Students were informed that back pain would be on the OSCE examination. They therefore had the opportunity to read about back pain prior to the exam. This may account for this group of students who reported they had read about back pain but had not picked up packets on back pain and had no clinical experience. The fourth intervention group, “no clinical experience and no reading” includes students who did not do the structured exercise and reported no random clinical experience and no reading on back pain problems. Thirty (17%) students reported no opportunity for learning about the history of back pain; and 37 (21%) students reported no opportunity for learning about the physical exam of back pain.
Downloaded by [McMaster University] at 23:36 10 April 2016
Objective structured clinical exam An OSCE, comprised of 22, 5 minute stations, was administered near the end of Year Two to evaluate student progress in acquiring clinical skills, and related to content from each of the four specialty rotations in Clinical Medicine 111. Two of the Family Medicine stations were designed to test history and physical examination skills dealing with the problem of back pain. In order to accommodate 188 students in a one-day period, students were randomly assigned to one of three 2 hour testing sessions. During each session the OSCE was run simultaneously in three locations. This resulted in the use of three raters and three simulated patients for each station. The same rater and patient staffed a station at each location for all sessions. Simulated patients and examiners were recruited from 3rd and 4th year medical students, as well as from Family Medicine faculty, who were not involved in the structured exercise (and therefore were not aware of which student examinees did the structured clinical back pain exercise). The back pain history and physical exam stations were set up to simulate an office examining room. The simulated patient portrayed a middle-aged man with low back pain due to a herniated disc. Students at each station were informed that the patient had low back pain. In the history station, they were instructed to ask the patient direct questions in order to obtain a detailed history of the problem and to then inform the rater of their diagnosis. In the physical exam station, they were instructed to perform a limited exam specific to the history of back pain, tell the rater reasons for examination manipulations, and to communicate the findings of the exam. The rater was instructed to have no discussion or comment on the student’s report during the stations other than to record appropriate information on the rating sheet. All feedback on stations was provided to students at the end of the entire OSCE exam. Simulated patients and raters were trained by faculty at orientation sessions. History patients and physical exam patients were given detailed instructions as to specific details for simulation of herniated disc. For the history station, raters used an objective checklist to rate student performance on specific history-taking skills as “satisfactory,” “attempted but not satisfactory,” or “not attempted”. Students were rated on their inquiry about specific characteristics of the low back pain, including: onset, course, duration, site, radiation, character, severity, aggravating factors, relieving factors, numbness, weakness, and the effects of activity, bowel movements, cough, straining or micturition. These history items were discussed in readings and were described as helpful in determining the etiology of back pain (Reilly, 1984). The student was asked to provide a diagnosis at the closure of the history and that answer was recorded. Student performance in the physical exam station was rated in the following areas corresponding to the skills
Structured clinical teaching strategy
181
taught in the structured clinical exercise: 1) observation, 2) range of motion, 3) palpation, 4) traction maneuvers, and 5) neurological exam. Student scores for the entire class of 188 are detailed in Table I and Table 11. The total possible score for each station was 45 points.
TABLE I. Percent of students correctly performing specific back pain history items ( N = 188)
Downloaded by [McMaster University] at 23:36 10 April 2016
Satisfactory Onset Course Duration Site Radiation Character Severity Aggravating factors Relieving factors Numbness Weak n ess Movement/activity Bowel movement Coughlstraining Micturition
94% 78% 92% 90% 85% 85% 63% 73% 90% 38% 24% 82%
11% 7% 7%
TABLE 11. Percentage of students correctly performing specific back pain exam items ( N = 188) Satisfactory Observation ROM flex/ext ROM latlflex ext Palpation Straight leg raising Hip flexion Hip extension Hip adduction Hip abduction Hip internal rotation Hip external rotation Reflex, knee Relfex, ankle Reflex, Babinski Foot dorsiflexion Foot plantar flexion Foot eversion Foot inversion Dorsiflexion, big toe
74% 72%
51% 76% 78% 53% 87% 13% 17% 26% 28% 94% 91% 52% 22% 20% 7% 7% 2%
182
Sharon S.Allen et al.
Downloaded by [McMaster University] at 23:36 10 April 2016
Data analysis Several concerns were addressed before the analysis was done. One concern was selfselection of students for the structured clinical exercise. Since, as reported in the next section, the self-selected group did not perform better than the other groups, it is unlikely that the self-selected group was “ a prior?’ different than the other groups, based on such factors as interest in the problem of back pain or perceived strengths or weaknesses in skills. One could hypothesize that the self-selected group was less able before the structured exercise, and through it brought themselves to the level of students who had no exposure to readings or patients in the area. However, this is unlikely. Another concern was that significant differences were found among OSCE sites on the average score for each station. Since students were randomly assigned to OSCE sites, the significant differences in mean ratings across sites suggested some inter-rater unreliability across sites and that more training of patients and raters is necessary to achieve comparable results using multiple raters in different sites. As described above, the same rater and simulated patient were used the entire time at each site. There were no significant differences in mean ratings within sites over the three OSCE time-slots and this suggested intra-rater reliability within each site. Therefore, within each site, an analysis of variance was used to compare average student scores on the back pain histories and physical exams for the four levels of instructional intervention. Also, Chi Square analyses were performed for each specific item on the history and physical examination checklist. One person, the first author, read the written diagnoses by all students at the end of the history and categorized them medically as to be more differentiated, meaning nerve involvement, vs. less differentiated, meaning low back strain, mechanical strain or muscle strain. A Chi Square was then performed comparing the level of diagnosis and the instructional level. This time the instructional level was grouped as structured learning group vs. other groups combined. Thirty-eight students were not used in this analysis because of inability to track the original choice exercise since exercises had been mailed back to the students.
Results There were no statistically significant differences in students’ performances on each OSCE station, overall, in relation to the types of learning situation (Table 111). In addition, Chi Square analyses show no significant differences for each item on the history and the physical examination checklist in relation to learning situation. Across sites and intervention groups, mean scores on the physical examination for back pain were consistently lower than mean scores for the back pain history. An inspection of the specific history items revealed that students did well in obtaining the basics of the history, including onset of the pain, course, duration, site, radiation, character, severity, and aggravating and relieving factors. However, they frequently failed to address neurological areas such as numbness, weakness, effects on bowel movement and micturition (Table I). There were no significant differences between the structured learning group, and the other groups combined, in their ability to give a diagnosis at the close of the back pain history OSCE station. Seventy-nine percent (79%) of students (45157) who did the structured back pain exercise mentioned a nerve involvement as their diagnosis at the end of the history, while 12/57 (21.1%) mentioned a less differentiated diagnosis such as low back strain, mechanical strain or
Structured clinical teaching strategy
183
TABLE 111. Mean OSCE scores at three sites for back pain exam and history according to self-report of learning experience* **(N=173 for history; N = 1 7 5 for exam) ~~
Site 1
Downloaded by [McMaster University] at 23:36 10 April 2016
Learning Experience
Site 2
Site 3
Hx
Px
Hx
Px
Hx
Px
None
31.8 N=9
23.5 N=10
28 N=ll
20 N=12
26.3 N=10
17.1 N=15
Reading Only
25.8 33.1 N = 10 N = 1 2
26.4 N=8
18.2 N=9
27.3 N=15
17 N=12
Random clinical Experience
32.6 21.1 N = 19 N = 1 6
28.9 20.4 N = 12 N = 10
24.5 N = 13
16.6 N=13
Structured Experience
31.9 N=18
24.2 N=18
31.1 N=24
20.1 N=24
26.1 N=15
17 N=15
32.3
23.5
29.3
19.8
26.1
16.9
Average Mean
**Fifteen students failed to report information, *Total possible score=45.
muscle strain. Of the 79%, several even localized the nerve involvement to L-4, L-5, or S-1. This diagnosis was made in light of the fact that students failed to address neurological areas in their history (Table I). Of students who did not do the back pain exercise, 79% of them (73193) also mentioned nerve involvement as their diagnosis at the end of the history. Twenty-two percent (20193) had less differentiated types of diagnoses. In addition, basic parts of the back physical exam-observation, palpation, range of motion and straight leg raising were consistently performed, but a large percentage of the class neglected important neurological components of the back examination and some of the musculoskeletal items, such as the hip exam. Two neurological items, knee reflex and ankle jerks were done more consistently (Table 11). A large percentage of the students who did the back pain exercise and who did not, 79% respectively, mentioned nerve root involvement in their diagnosis and yet failed to do important parts of the neurological back pain exam.
Discussion The low mean scores on the history of back pain, and even lower mean scores on the physical examination of back pain, for students in all intervention groups, suggests that students have not learned to conduct a thorough history and examination for back pain utilizing the principles discussed above. This is surprising, especially for the group who participated in the exercise, since the readings did emphasize appropriate neurological questions and areas for neurological physical examination associated with mechanical low back pain identified as radicular low back pain. A large percentage of students (79%), regardless of learning strategy, could accurately determine the etiology of the back pain at the history station and mention nerve root involvement and yet a large percentage of all instructional groups did not then proceed to do appropriate neurological questions and physical exam. This suggests that a single clinical structured exercise has little impact. It is puzzling that students could accurately hypothe-
Downloaded by [McMaster University] at 23:36 10 April 2016
184
Sharon S. Allen et al.
size a neurological problem but fail to perform appropriate questioning and physical examination. It appears from this study that a single structured instructional strategy (including specific readings, clinical experience with immediate feedback from their preceptor, a prepared write-up on the clinical experience with written feedback from faculty) is not adequate to ensure student learning pertinent history and physical exam skills even though many clinical courses do not include even one such strategy for many important clinical problems. Teaching students how to use a specific data collection strategy and when to use it (most likely) requires supplementary instructional strategies. In conclusion, this study assessed an effort to follow important recommendations of the GPEP report: to identify some specific clinical knowledge and skills needed by students and to provide a structured learning exercise designed to help students acquire (develop) these skills. Given the time constraints in medical school curricula, most curricula do not include even one such structured learning opportunity for important medical problems. Unfortunately, however, this study was unable to demonstrate that a single structured clinical exercise for a specific common problem seen in the ambulatory setting enhanced students’ ability to select and use specific history or examination skills for that problem. These observations are important since time and information overload frequently dictate limited learning opportunities for important clinical skills. It would seem incumbent on medical educators to identify important skills and knowledge and to design curricula which will allow multiple longitudinal integration of structured experiences to reinforce these skills and knowledge areas.
Correspondence: Sharon S . Allen, MD, PhD, Assistant Professor, Department of Family Practice and Community Health, University of Minnesota, 6-240 Delaware Street SE, Box 381 UMHC, Minneapolis, Minnesota, 55455, USA. REFERENCES ASSOCIATIONOF AMERICAN MEDICALCOLLEGES (1984) Physicians for the Twenty-First Century. The GPEP Report. Report of the Panel on the General Rofessional Education of the Physician and College Presentation for Medicine (Washington, DC 20036). BOWMAN, W.C. (1982) Evaluating performance effectiveness on the job: how can we generate more accurate ratings? in: LLOYD,J.S. (Ed.) Evaluation of non-cognitive skills and clinical performance (Chicago, American Board of Medical Specialties). F.A. (1979) Assessment of clinical competence using an objective structured HARDEN,R.M. & GLEESON, clinical examination (OSCE), Medical Education, 13, pp. 41 -54. HARDEN,R.Mc.G., STEVENSON, M., DOWNIE,W.W. & WILSON,G.M. (1975) Assessment of clinical competence using Objective Structured Examination, British Medical Journal, 1, pp. 447-45 1. MARERN, W.D., WEINHOLTZ, D. & FREIDMAN, C.P. (1984) The attending physician as teacher, New England Journal of Medicine, 208, pp. 1129-1 132. NEWBLE, D.I., HOARE,J. & ELMSLIE, R.G. (1981) The validity and reliability of a new examination of the clinical competence of medical students, Medical Education, 15, pp. 46-52. NORMAN, G.R. (1985) Objective measurement of clinical performance, Medical Education, 19, pp. 43-47. PAYSON, H.E. & BARCHAS, J.D. (1965) A time study of medical teaching rounds, New England Journal of Medicine, 237, pp. 1468-1471. PERKOFF, G.T. (1986) Teaching clinical medicine in ambulatory setting, New England Journal of Medicine, 314(1), pp. 27-31. PETRUSA, E.R., BLACKWELL, T.A., ROGERS,L.P., SAYDJARI, C., PARCEL,S. & GUCKIAN, J.C. An objective measure of clinical performance, American Journal ofhfedicine, (83) (July), pp. 34-41. REILLY,B.M. (1984) Low back pain, in: Practical Strategies in Out Patient Medicine, pp. 1-70 (Philadelphia, W.B. Saunders).
