This study used a rigorous PPSMA analytic framework to present empirical evidence for implementing the FC teaching model to teach EBM competency. For all aspects of EBM competency, including Ask, Acquire, Appraise, and Apply, students in the FC teaching group performed significantly better than the students in the LB group. This study also found that for higher-level competencies, such as Appraise (written exam) and Apply (oral exam) the evidence, the difference in performance between the two groups was more pronounced. These findings indicate that the FC model is a useful and efficient teaching method for all EBM competencies. In our previous experience, students could not overcome barriers to learning EBM, because they lacked practice and real-time problem solving skills. In the FC method, students have discussions with their peers through pre-class video clips, ask questions through the study platform, and resolve their doubts when sharing their answers in class. The interactive approach reduced learning gaps and helped students utilize the knowledge for clinical decisions, facilitating further utility of EBM in medical care. Moreover, our study endeavored to avoid the selection bias that often exists in non-randomized controlled studies. The matched control group made our results more convincing than most of the observational educational studies that utilized only pre- and post-study comparisons.
The current study findings on the use of the FC teaching model add to the body of evidence on teaching the tenets of EBM. A number of studies investigating different teaching methods were evaluated in one systematic review on the effectiveness of EBM teaching methods. This review concluded that the body of evidence available to guide educators on how to teach EBM to medical trainees is small, and further research is required to determine the effects of timing, content, and length of EBM courses and teaching methods [24]. However, studies included in this review were limited to randomized control trials (RCTs), but many empirical studies involving real-world teaching situations cannot be implemented using random assignment, so they were excluded. In other studies that used meta-analysis to analyze the effectiveness of FC teaching, less than 20% of the included studies were RCTs [25, 26]. Excluding non-randomized controlled studies may leave out a lot of available evidence from good quality research.
There are some reasons underlying the significant differences in the learning outcomes between the FC and LB groups. We observed the implementation process for teaching activities in the two groups and noticed that the characteristics of the curriculum may contribute to such differences. For example, in the FC group, by allowing students to engage in interactive discussions between teachers and peers after studying the pre-class materials, students may have more in-depth content knowledge, which can facilitate the discussion about the abstract concept of EBM principles. In contrast, students in the LB group simply received knowledge in the classroom, without having interactions and discussions beyond the classroom; the learning situation of students was similar to that of our EBM courses in the past before the innovative teaching was attempted [18]. Therefore, a higher degree of internalization of the knowledge and stimulating reflections from each student could not be achieved through the traditional learning method. EBM is not only a type of static knowledge, but a high-level skill that transforms knowledge into effective thinking, judgment, and decision-making [12, 27]. Curriculum design based on the FC teaching model may be more conducive to the development of EBM's ability to apply meta-knowledge.
Our study used the PPSMA method to control for the influence of multiple possible confounding factors on the evaluation of learning effectiveness in quasi-experimental and observational research, which helps enhance the confidence in the internal validity of the research results. However, there are still some limitations in this study. First, the sample size of that have finally included in the analysis was relatively small. However, the results demonstrated that there was still sufficient statistical power. Besides, it is still possible that there may still be other unobserved variables that might bias the evaluation of the intervention effect. Theoretically, it is very difficult for the PPSMA to include all possible sources of learning performance in the assessment. However, we included as many potential confounding factors as possible in the analytic framework, including age, sex, university admission route, student loans, part-time jobs, past academic performance in the preclinical years, personality dimension scores from the Big Five, and pre-course EBM scores.
In conclusion, the empirical evidence for teaching effectiveness demonstrated in this study provides an important reference to support the large-scale application of the FC method in EBM teaching. The results of this study may help EBM educators select the most appropriate teaching method. We believe that the application of this method to other clinical education fields may also have considerable potential. With the use of PPSMA, the evaluation of learning effects can be presented with a much more rigorous approach, for data processing and analysis. The present study represents an important step for informing clinical educators of a useful educational strategy, by sharing the successful experience of implementing the FC model in pre-clinical EBM curriculum, as well as the establishment of a rigorous framework for evaluating teaching effectiveness.