Although internal medicine residents commonly perform bedside procedures, technical proficiency and knowledge regarding procedures such as thoracentesis may be lacking. The risk of complications from thoracentesis may also be increased when performed by physicians‐in‐training. Our aim was to use a thoracentesis simulation to assess graduating residents' proficiency and evaluate the impact of an educational intervention on the development of skills to mastery standards.


This was a pretest‐posttest design without a control group. Subjects were all 40 of the third‐year internal medicine residents in a university‐affiliated program. A 25‐item checklist was created for the thoracentesis procedure using established methods for checklist development. Each resident underwent baseline assessment of thoracentesis skills using a standardized history, radiograph, and a thoracentesis simulator developed by MediSim Inc. Performance was assessed at baseline by 1 of 2 faculty raters using the checklist. The minimum passing score (MPS) was set by a panel of 10 clinical experts using the Angoff and Hoffstee standard setting methods. A 50% random sample of pretest sessions was rescored by a third rater from videotapes to assess interrater reliability. After baseline testing, residents received two 2‐hour education sessions featuring a videotaped didactic presentation and deliberate practice with the thoracentesis simulator supervised by 1 of 2 faculty instructors. Residents were then retested. Those who did not achieve an MPS on the thoracentesis posttest had additional deliberate practice and were retested until the MPS was reached. Residents were asked to rate the curriculum using a 5‐point Likert‐type response system (from 1 = strongly disagree to 5 = strongly agree). Checklist score reliability was estimated using the Kappa coefficient. Pretest and posttest scores were analyzed using paired t tests.


Interrater reliability for the thoracentesis checklist was very high (k = 0.94). Performance improved significantly after simulation training (pretest mean score = 51.7, posttest mean score = 88.3, P < .001). The MPS used as the mastery achievement standard was 80%. Thirty‐seven of the 40 medicine residents (93%) achieved mastery within the standard 4‐hour training. To reach mastery, the remaining 3 residents needed extra time, ranging from 20 to 90 minutes. Residents agreed that practice with the thoracentesis model boosted their skill (mean rating = 4.3 ± 0.8 out of 5) and helped prepare them to perform the procedure better than did clinical experience alone (mean rating = 4.0 ± 1.0).


A curriculum featuring deliberate practice with a thoracentesis simulator dramatically increased the skills of residents in performing this procedure. The educational program received high ratings from learners and provides a reliable assessment of procedural competence.

Author Disclosure:

J. H. Barsuk, None; K. J. O'Leary, None; D. B. Wayne, None; M. J. Fudala, None; W. C. McGaghie, None.