degli Studi di Siena in the year 1994 were invited to participate in the study. The course of Italian medical studies is structured so that this was the first course involving patient examination, except for a general course in Semeiotics, which had been held 3 months before, in which the students had limited practical experience. At the beginning of the course in Respiratory Medicine, the students were given a formal lesson on chest examination, including information on the classification and characteristics of lung sounds. Although our study was conducted within the time allotted for the course of Respiratory Diseases, participation in the study was freely chosen, and an informed consent form was given to all participants.
The course of Respiratory Medicine was held over a period of 2 months in March and April 1994, and included a weekly session of 4 h dedicated to patient examination, clinical case simulations, and attendance at clinical laboratories. To avoid overcrowding, the students were divided into five groups, all of which attended the ward for at least 1 h in each session, according to a schedule arranged so that no more than two groups were simultaneously present.
During the very first session of clinical practice, all the participating students were requested to report on a form their findings of chest auscultation of three selected patients. After the test, the students of three randomly selected groups attended a multimedia lung sound presentation session lasting approximately 1 h (see below). At the same time, the remaining two control groups attended 1 h of conventional bedside teaching on lung sounds, in which each student was allowed to listen to the chest of 4—5 patients, selected as representative of definite patterns of lung sounds, and the findings were discussed with the tutor. The following week, during the second clinical session, the students of all five groups repeated the auscultation test using the same method as on previous occasion, but with different patients
In subsequent weeks, the control students also attended the multimedia lung-sound session, without repeating the auscultation test. At the end of the course on Respiratory Medicine, all the students were requested to fill out an anonymous feedback questionnaire regarding various aspects of the organization of the course, including a few questions concerning their participation in this study.
Evaluation of auscultation skills
Each student was given a form, labeled with an anonymous identification code, on which he or she reported the auscultatory findings for each of a group of three patients, about whom no clinical information was given. The form contained a checkbox to specify whether the finding was normal or not, and separate rows to describe the characteristics and localization on the chest of breath sound abnormalities, wheezes, rhonchi and crackles, when present.
The testing cases for the study were selected among
the patients hospitalized in our 24 bed University Clinic for various respiratory diseases. Each patient was examined by three of the authors (ER, MR and PS) in the morning preceding the student’s practical session, and a consensus was obtained on the relevant features present on auscultation.
The criteria for patient selection were: 1) the auscultation findings were clear and definite and a consensus on the classification of the relevant features was easily obtained; 2) their clinical condition was such that they could stand multiple chest examinations; and 3) they gave informed consent to participate in the study.
Groups of three patients to be assigned to the different groups of students were assembled according to the following criteria: 1) each group of patients would include at least one case in which crackles were present, at least one with continuous adventitious sounds (wheezes and/or rhonchi), and one with either decreased breath sound or a completely normal finding; 2) each patient would be examined by an approximately equal proportion of students of the multimedia and control groups; 3) none of the patients would be assigned twice to the same student group; and 4) the difficulty of the auscultation findings would be reasonably balanced between the different groups of patients.
We recruited seven groups of patients and each group was examined by an average of 14 students. The number of relevant features identified during the preliminary consensus auscultation as the ones that had to be detected by each student was similar in the patients examined by the multimedia and the control group, both
in the preliminary assessment.
The test was performed in groups of 5—10 students accompanied by a tutor, and approximately 5 min were allotted to each student for each patient. The tutor also examined the patient at least once during the session, to verify that the auscultatory findings were unchanged in patients who were considered potentially unstable.
All the forms were scored independently by two of the research team (ER and PS), without knowledge of which group they belonged to. Each researcher first considered each relevant finding, as identified in the previous consensus auscultation, evaluating whether it had been detected, undetected or misclassified (when a wrong name was given to a present feature). When localized bilateral basal crackles were present, right and left crackles were considered as two separate findings, whereas diffuse rhonchi (low pitched continuous adventitious sounds), wheezes (high pitched), reduction of breath sounds and widespread crackles were each considered as a single finding. Misclassification of crackles as pleural rubs was not considered, and no distinction was made between fine and coarse crackles, as these distinctions were considered too difficult a task for students at this stage. Irrelevant misclassifications, such as reports of modest reduction of breath sounds where it was not present, were not considered unless the patient presented the opposite abnormality or they were reported as the only abnormality in that patient. Those findings fabricated by the students (reported but not present) were also counted.