Since the composition of the five original groups of students attending the clinical session was not properly randomized, we first checked the homogeneity of the performance of the different groups in the baseline test. No significant differences between the groups were found in the number of detected, undetected, misclassified and fabricated findings or in the LSDIS. Minor differences were found in the localization and timing fault scores, with one group being significantly worse in the first and
better in the latter, compared to the other groups.
In the preliminary test, no significant difference in the LSDIS or in any of the parameters of the study was found between intervention (multimedia seminar) and control students (table 1). By contrast, in the postintervention assessment, the mean LSDIS of students who attended the lung sounds session was significantly decreased compared to baseline, and was significantly better than that of controls (fig. 1), whereas LSDIS of the controls did not change compared to baseline.
Among the various parameters used to compute the LSDIS, the number of detected features in the postintervention assessment was slightly higher and the number of undetected and misclassified features and that of localization and timing faults were lower in the intervention group than in the controls (table 1). However, only the difference in the number of fabricated features reached statistical significance (p<0.0l). The analysis of the different type of fabricated sounds showed a reduction of undefined and breath sound abnormalities reported by the multimedia group, whereas adventitious sounds were less affected (table 2).
The feedback questionnaire was filled out by 55 students, 52 of who had attended the lung sounds session. Of these, only one reported that the session had a deleterious effect on the learning of respiratory sounds, four that it had no effect, 39 reported a positive, and seven a very positive effect. Similarly, 85.5% of the students thought that the association of lung sounds with their visual image was useful, and 87.5% would recommend attending the seminar to other students, with the remaining 12.5% indifferent, and none suggesting not going. Regarding a possible computer program to be used by the students singly, only 56% of the students thought that it could be useful, whereas 13% disagreed and 31% were uncertain, sometimes adding as a note that they had no experience with computers. Finally, regarding their participation in the study, 5! of the students reported that performing the auscultation test was interesting or fun, one student that it was indifferent, and none found it tiring or boring.
Our data indicate that the exposure of inexperienced medical students to a multimedia presentation of acoustic and graphic characteristics of lung sounds significantly boosts their learning process compared to students receiving only conventional teaching.
Several factors might have contributed to this difference. The effectiveness of bedside teaching could have been reduced by the limited availability of representative patients, by a shorter time dedicated to each student because of the need for individual auscultation, or by a lower motivation of the tutors compared to the teacher
performing the multimedia presentation. However, the