General context
The Neurology Course at the Department of Neurology and Neuroscience at the Medical Center of the University of Freiburg, Germany, usually takes place during the students’ fourth or fifth year of study and is the first contact with clinical neurology. The mandatory 3-week block course includes disease-oriented lectures (12 × 1.5 h, groups of 80 students), symptom-oriented seminars (4 × 1.5 h, groups of 20 students), practical teaching of the neurological examination (2 × 3 h, groups of 6 students) and neurological bedside teaching (7 × 3 h, groups of 6 students). The course ends with a summative multiple choice question examination for all participants covering all course sections.
Design of the study
The Ethics Committee of the Medical Center of the University of Freiburg, Germany, approved our study and all students participating in the key feature problem examination provided written, informed consent. The study was performed using a static-group comparison design: the TBL-class covering the topics of the seminars was offered as a voluntary supplementary class for all participants of the neurology course. The key feature problem examination was offered as a voluntary formative examination to all participants of the TBL-class (TBL-group) and to all students of the neurology course not participating in the TBL-class (non-TBL-group). The summative multiple choice question examination was mandatory for all participants of the neurology course (Fig. 1).
Sample
Of the 123 students attending the neurology course in the winter semester 2012/2013.
28 students applied for the TBL-class. Due to the tight schedule of students’ fourth or fifth year of study at our university only 17 students (10 fourth-year and 7 fifth-year students; 7 male and 10 female) could participate at the selected time points of the voluntary supplementary TBL-class. The non-TBL-group consisted of 15 students from the same neurology course (10 forth-year and 5 fifth-year students, 6 male and 9 female) with 11 of them having also applied for the TBL-class but could not participate due to their individual schedule (see above).
Seminars
The seminars with groups of a maximum of 20 students took place in the regular context of the neurology course. The 90-min units covered the topics “vertigo”, “back pain”, “first epileptic seizure” and “acute altered mental status”. The seminars were held interactively and included case-based teaching of specific diseases, forcing students to recollect knowledge about relevant neuroanatomical background and utilize concepts for diagnosing and treating these diseases. Voluntary individual advance preparation was not requested but made possible since the slides used in the seminars were available in advance via the university’s learning management system. The seminars were mandatory for all students, but local study regulation allowed an authorized absence of one seminar. Only long-time experienced board-certified neurologists with postdoctoral lecture qualification acted as teachers for all seminars.
Team-based learning
The following description of the TBL activities used in this study is based on the proposed guidelines by Haidet et al. [18]. The TBL-units lasted 90 min each. There was one trained instructor (J.B.) for all TBL units who had been teaching TBL for 2 years in voluntary settings [19]. The teams were randomly distributed into groups of five to seven participants. The permanent assignment for all units without possibility of switching was pointed out to all participants. Since TBL is not used in other fields at our Medical School the first TBL-unit about neuroanatomical localization utilizing previous knowledge about neuroanatomy was held to introduce the new teaching method. Subsequently one TBL-unit to each of the above named topics of the seminars was taught.
Preparation (Phase 1) was recommended. Students were asked to read the corresponding seminar slides that were available in advance via the university’s learning management system. Each TBL unit (Phase 2) began with a paper-based, 5-min individual Readiness Assurance Testing, which consisted of three multiple-choice questions on clinical presentations, diagnostics, and therapy. These questions were constructed to cover important issues of clinical reasoning. Each question was subsequently discussed for 5 min in teams. The teams were then responsible for generating an answer and appointing a team spokesperson (team Readiness Assurance Testing). After a prearranged signal, the teams simultaneously held up their answers on colored paper sheets. A discussion moderated by the instructor started among the spokespersons about the different team responses during which the elimination of the alternative answers needed to be actively justified. The instructor gave immediate oral feedback during and after the discussion, which lasted about 10 min and ended with a short summary of the underlying concepts. The application exercises (Phase 3) were each comprised of a clinical case closely based on real cases (“significant problem”) with one to two related multiple-choice questions. The answering options were diagnostic and therapeutic steps, so that the teams had to come up at first with a preliminary diagnosis in order to solve the questions. One application exercise per unit was given to all of the teams to be worked on and discussed for 5–10 min within the teams (“same problem”). Afterwards all teams reported their group results to the audience when given the signal (“specific choice” and “simultaneous reporting”). The teams discussed the selected as well as discarded answers among themselves supervised by the instructor. After the discussion the instructor provided immediate oral feedback and also gave a brief summary of the underlying concepts if necessary pointing out the critical steps in clinical reasoning. No grading or peer evaluations were conducted.
Key-feature problem examination
The key feature problems were developed with regards to the contents of the seminars according to the steps recommended by Page et al. [11]. They were written by didactically and clinically experienced neurologists of the Department of Neurology and Neuroscience at the University Medical Center Freiburg and reviewed and adapted by two physicians with long-term clinical expertise in neurology who were not involved as authors. Diseases discussed in the TBL application exercises were excluded. The key feature problem examination was intended as a voluntary, formative examination at the end of the neurology course that took place 5 days before the summative multiple-choice examination. Motivation for taking part in the key feature problem examination was encouraged by offering a book voucher as a reward.
The key feature problem examination was conducted in the faculty’s computer lab using a computer-based exam system [20]. Each participant was assigned a unique login and password. Each key feature question could only be answered once, as the correct answer was mostly implied by the following item. Therefore, backward navigation was only possible to review information, not for editing. All answers were centrally recorded on the system’s server. After a short introduction to the test procedure and the test tool, the students had to answer 13 key feature problems in 60 min (four to each of the three seminars “back pain”, “first epileptic seizure” and “acute altered mental status” and five to the seminar “vertigo”). Each problem consisted of three to four key features. As a total, the students had to answer 51 key features, 25 in short menu question format and 26 in long menu format [14, 21]. Each key feature was scored one point, with partial credits of equal weights if not all correct answers were given.
Multiple choice question examination
The multiple choice question examination (MCQE) closing the neurology course consisted of 40 Type-A multiple choice questions with a set of five options each testing factual or conceptual knowledge. Six questions referred to the four topics addressed in the seminars and TBL units, the remaining 34 questions referred to complementary topics of the lecture and the bedside-teaching, such as multiple sclerosis, dementia, muscle diseases, neuro-oncology and the clinical neurological examination. Three experienced neurologists reviewed all questions internally.
Statistical analysis
Item analyses were computed for key feature problem examination and MCQE, using Cronbach’s α to determine its internal consistency. Differences between the TBL- and the non-TBL-group were tested by means of t tests for the normally distributed results of the key feature problem examination (verified with the Kolmogorov–Smirnov test) and Mann–Whitney-U tests for non-normally distributed results of the MCQE. Effect size was calculated using Cohen’s d. All statistical analyses were performed with SPSS, version 21 (IBM).