3.1 Descriptive statistics
A total of 140 students majoring in clinical medicine who had a one-month clinical practice in breast surgery of The First Affiliated Hospital of Zhengzhou University were recruited and randomly assigned to either the MDT teaching mode group (experimental group)(n = 70) or the traditional teaching mode group༈control group༉ (n = 70). The age of experimental group was (23.71 ± 0.127) years and that of the control group was (23.6 ± 0.14). There were 33 males and 37 females in the experimental group and 37 males and 33 females in the control group. The number of students in the experimental group and the control group who actually completed all course plans was 92.9% (65/70) respectively. The total absence time of students in the experimental group and the control group was 8 hours respectively. The comparison between the other factors of the experimental group and the control group students (height, weight, place of birth, whether glasses are worn, previous GPA, and whether there are clinical workers in the immediate family) is shown in Fig. 1. Among the baseline factors of the enrolled students in the experimental and control groups, only height and weight were correlated. The correlation coefficient (Corr) between height and weight in the experimental group was 0.812, while the Corr between height and weight in the control group was 0.830, all of which were highly correlated (Corr > 0.8).
First, for the scores of the two groups of students on the knowledge acquisition pre-test, skill acquisition assessment pre-test, clinical decision-making assessment pre-test, knowledge acquisition post-test, skill acquisition assessment post-test and clinical decision-making assessment post-test, we used the mean, standard deviation, median and other measurement indicators to perform descriptive statistics on the data to understand the distribution of the data and identify outliers. As shown in Supplementary Table 1, the descriptive statistics of all variables in the data set (gender, age, height, weight, native/non-native and wearing glasses/not wearing glasses) of the experimental group and the control group are shown respectively.
For the experimental group, the average scores of the knowledge acquisition pre-test, skill acquisition evaluation pre-test, clinical decision evaluation pre-test, knowledge acquisition post-test, skill acquisition evaluation post-test, and clinical decision evaluation post-test are 18.09, 17.09, 14.05, 28.31, 23.54, and 23.54, respectively. The standard deviations of these scores are 0.73, 0.74, 0.83, 0.67, 1.28, and 1.91, respectively. The median scores of these six tests are 18.15, 17.15, 13.95, 28.30, 23.60, and 23.35, respectively. The score ranges are 14.6–19.8, 13.7–18.5, 12.1–16.3, 26.6–30.0, 20.3–26.8, and 19.9–29.1, respectively. For the control group, the mean scores for the knowledge acquisition pre-test, skill acquisition assessment pre-test, clinical decision assessment pre-test, knowledge acquisition post-test, skill acquisition assessment post-test, and clinical decision assessment post-test are 17.86, 16.80, 14.16, 27.80, 22.33, and 22.14 respectively. The standard deviations for these scores were 1.03, 1.08, 1.17, 1.29, 1.04, and 1.27 respectively. The median scores for these six tests were 17.8, 16.8, 14.1, 27.8, 22.2, and 21.8 respectively. The score ranges were 13.3–20.3, 13.8–19.6, 12.1–18.2, 20.7–29.7, 20.8–25.8, and 19.6–26.8 respectively. Table 1 shows the mean scores of the pre- and post-test (knowledge acquisition tests, skill acquisition assessments, and clinical decision-making evaluations) for each group. The MDT teaching mode group had a higher mean score in both the multiple-choice questions and the practical skills evaluation. The difference in mean scores between the two groups was statistically significant (p < 0.05).
3.2 Descriptive statistics
Subsequently, we conducted a correlation analysis on different teaching modes in our dataset to explore whether there is a correlation between the six exam scores of two groups of students and other different variables (such as gender, age, height, weight, place of birth, and whether students wear myopia glasses). We conducted different correlation analyses based on the data types of the variables. For continuous variables such as height and weight, we performed Spearman correlation analysis (Figs. 2). For discrete variables, we examined the distribution of the six scores across different categories of each variable (Figs. 3).
In the correlation analysis between the six test scores of the traditional teaching model group and the MDT teaching model group and other variables, the correlation coefficient ranges from − 1 to 1, with values close to 1 indicating strong positive correlation, and values close to -1 Indicates a strong negative correlation. We found no significant correlations between any of the continuous variables and test scores regardless of the instructional mode (traditional or MDT). In the experimental group, student age had the strongest correlation with clinical decision assessment scores, but it was not statistically significant (R = 0.2, P = 0.094) (Fig. 2A). In the control group, students' height had the strongest correlation with pre-test knowledge acquisition scores, but it was also not statistically significant (R=-0.18, P = 0.14) (Fig. 2B).
In a correlation analysis of discrete variables, we found no variables that were correlated with student achievement (Fig. 3). These results suggest that students' individual baseline characteristics have little impact on their performance and that score differences between the two groups may be attributable to different teaching methods.
3.4 Analysis of differences in academic performance between experimental group and control group
We conducted a Wilcoxon rank sum test (also known as the Mann-Whitney U test), a non-parametric test, to analyze whether different teaching modes had an impact on student test scores. This test is suitable for comparing independent samples when the data does not follow a normal distribution.
First, we compared the pre-test scores for knowledge acquisition, skill acquisition assessment, and clinical decision-making assessment between the two groups of students. The results showed that the experimental group had higher pre-test scores for knowledge acquisition and skill acquisition assessment compared to the control group (P = 0.066, P = 0.067), while the scores for clinical decision-making assessment were lower (P = 0.8) (Fig. 5A-5C). However, these differences did not reach statistical significance.
Next, we compared the post-test scores for knowledge acquisition, skill acquisition assessment, and clinical decision-making assessment between the two groups. The analysis revealed that the experimental group had significantly higher scores in all three areas compared to the control group, with P-values of 0.0016, 2.8e-09, and 3.6e-07, respectively (Fig. 5D-5F). These findings suggest that the MDT teaching mode used in the experimental group had a better training effect on students compared to the traditional teaching mode in the control group.
3.5 Long term analysis
Students in both the experimental group and the control group chose to continue studying for a master's degree after completing their undergraduate studies. We conducted a statistical analysis of the master's majors they chose. Among the students who chose to study the master's degree in breast surgery, there were more students in the MDT teaching model group than in the traditional teaching model group, but the difference between the two is not statistically significant. There are also more students in the MDT teaching model group who choose surgical majors (including thoracic surgery, gastrointestinal surgery, thyroid surgery, plastic surgery, hepatobiliary surgery, neurosurgery, cardiac surgery, orthopedics, etc.) when pursuing their master's degree than those in the traditional teaching group, but the difference between the two is not statistically significant, too (Fig. 6, Table 2). These findings indicate the sustained effectiveness of the MDT teaching mode in enhancing student learning outcomes. For students pursuing a master's, no statistically significant differences in specialty choice were observed between groups, though a trend towards more MDT students choosing surgical specialties was noted. This indicates potential sustained benefits of MDT teaching on career trajectory.