Impact of simulation-based training in surgical chest tube insertion on a model of traumatic pneumothorax

Study

This was a prospective, randomized, controlled, bicentric study, carried out from May 2013 to October 2013 in the Simulation Laboratory of the Faculty of Medicine of Poitiers and in the Medical Center of Cayenne, French Guyana, after approval by the Institutional Research Board (Scientific Committee of the Faculty of Medicine of Poitiers, file number 11–37). All participants signed an informed consent form. Results were kept anonymous.

Objectives

The primary objective was to evaluate the success rate of surgical insertion of a chest tube in a task trainer simulator of traumatic pneumothorax.

The secondary objectives were (1) to assess the performance of insertion procedure according to a performance assessment scale, (2) to measure the global procedure time and dissection time during insertion, (3) to seek out a correlation between the learner’s status, experience, and performance and success rate, and (4) to survey the learners for their evaluation of the realism of the model, self-confidence, and satisfaction after simulation experience.

Population

Sixty-five healthcare providers (18 residents and 47 senior emergency physicians), representing two groups of registered participants (Poitiers and Cayenne) for the pediatric emergency procedure course at the University of Poitiers (carried out in Poitiers and in Cayenne), were given a chance to participate.

Intervention

Since the surgical approach is rarely taught and practiced in France, all participants received a 1-h academic lesson, prior to the simulation session, on current international recommendations for surgical chest tube insertion for traumatic pneumothorax [1, 21–24]. This didactic lesson was performed by three Advanced Trauma Life Support (ATLS)-certified supervisors. The session consisted in a presentation of the currently recommended safest approach to chest tube insertion in case of traumatic pneumothorax: landmarks on the medio-axillary line, in the fourth or fifth intercostal space [21, 22, 25], use of a tube without chuck or handle, dissection of chest wall muscular layers with a Kelly clamp, followed by insertion of a gloved finger probing into the chest cavity to confirm pleural placement, and strip any adhesions facing the insertion site [21]. In small children, it is recommended to tunnel the chest tube by a skin incision at the underlying intercostal space. This provides a subcutaneous path for securing the tube and avoids its falling out due to a thin chest wall [22].

Comparison

Randomization was based on a list of 65 random numbers—even numbers for SIM+ and odd ones for SIM− group allocation. The SIM+ group participants (n = 34) were assigned for deliberate practice on the simulator just after the didactic lesson, whereas the SIM− group participants (n = 31) were not assigned any simulation practice. The SIM+ group participants had the opportunity to practice several times (mean = four times) on the simulator with immediate feedback from the supervisor. The SIM− group participants were exposed 5 min to the simulator and its environment (webcam, laptop, chest tube equipment) but did not perform chest tube insertions. One month later, all participants (SIM+ and SIM−) were evaluated during an assessment session on the simulator.

Every participant received a questionnaire on clinical experience in chest tube insertion—especially surgical approach—before the didactic lesson and during the month that followed prior to the assessment on the simulator. We arbitrarily distinguished experienced participants who had inserted at least five chest tubes (with a surgical approach) during the last 5 years from novices who had inserted less than five or none during the same span time.

At the end of the study, all participants were offered a chance for deliberate practice on the simulator with supervision.

Outcomes

The scenario consisted in an emergency surgical chest tube insertion for a traumatic pneumothorax in a teenager. The model was the one we had previously developed and published [26]. Briefly, the model assembled a lamb half chest tightly fixed on a box cover (Fig. 1). A spread-out two-layer plastic film simulated the pleural membranes. Inside the box, a webcam connected to a laptop made it possible to assess the intrathoracic steps of the procedure. All the required disposable equipment was set on a table: a 24-Fr-wide Joly drain (diameter 5.4 mm, length 40 cm, with two lateral holes, without chuck or handle) (Fig. 1). The landmark of the fourth rib was indicated to the participant. All simulation sessions were videotaped.

Success was assessed on the laptop connected to the webcam using two criteria: (1) sudden loss of resistance while passing through pleural membranes (one could hear the “pop” of the penetration in the pleural space) and (2) insertion of the chest tube without resistance over at least 10 cm (at or over the third black bar). Assessment of performance was carried out by an independent observer expert in surgical approach for chest tube insertion (ATLS-certified or ATLS instructor), who was neither a supervisor nor a research investigator. He was unaware of the breakdown into two groups. Four observers participated in the assessment (two emergency physicians, one intensivist, and one pediatric intensivist). All were trained for surgical chest tube insertion assessment according to a specific scale that we had previously designed and validated [27]. It included eight steps, detailing the procedure: asepsis, local anesthesia, cutaneous incision and dissection of the thoracic wall, confirmation of insertion, introduction of the chest tube with a Kelly clamp, securing the water seal tubing, securing the chest tube, and location of the incision site (which was examined by the observer after the procedure had been completed). Each item was ranked 1 (correctly performed) or 0 (not done or incorrectly performed). The maximum total score was over 20 points. This scale had acceptable internal coherence (Cronbach’s alpha coefficient = 0.747) and high inter-observer reproducibility (intra-class correlation coefficient = 0.966). Because of the webcam connected to a laptop inside the model, the observer could assess the performance of the extra- and intrathoracic steps of the procedure.

Timing of the different steps of the procedure was recorded. A stopwatch was started at the beginning of the procedure (t0) to record different times: beginning of the cutaneous incision (t1), chest tube passing through the skin (t2), connection of chest tube to water seal tubing (t3), and end of the procedure after securing the chest tube on the skin (t4). Dissection time was consequently defined as t2 − t1.

After each chest tube insertion, a debriefing was performed by the observer, and all the observers were trained in debriefing by good judgment [28]. Videotape replay during debriefing could render participants aware of their gaps in performance [29].

After each assessment, the participant was asked to fill out a questionnaire on realism of the model, self-confidence, and overall satisfaction with simulation-based education using 10-point Likert scales.

Statistics

Analysis was carried out on Biostat TGV software and Excel 2010. Descriptive analysis included percentage, mean, and standard deviation (SD) of every variable. Comparative analysis used paired Student’s t test, with an ANOVA for repeated measures when necessary. Correlation between training and performance or success rate used Spearman’s test. Correlation between success and performance used Pearson’s test. A p value of <0.05 was considered as significant.

Population studied

Primary objective

Secondary objectives

There was no difference in the procedure time between SIM+ and SIM−, p = 0.71. By contrast, the SIM+ group had a shorter dissection time than the SIM− group, p = 0.047. There were no correlations between status and procedure time (p = 0.52) or dissection time (p = 0.16), nor between a participant’s previous experience and procedure time (p = 0.98) or dissection time (p = 0.65) (Table 2). Status was not correlated with either performance score (p = 0.24) or success rate (p = 0.94).

There was no difference between experienced participants and novices regarding performance score (p = 0.66) or success rate (p = 0.54) (Table 2).

Main results

On a traumatic pneumothorax model, simulation-based education was associated with higher success rate and performance score. These results were directly due to the simulation training, with no influence of status or previous clinical experience in surgical chest tube insertion. To our knowledge, this is the first study reporting the benefit of simulation-based education on success rate and performance for a surgical chest tube insertion.

Limitations

This study nevertheless presented some limitations. All participants were registered for the pediatric emergency procedure university course, which could have represented a confounding factor. The SIM− participants were exposed to the model only 5 min prior to the evaluation. Therefore, on evaluation day, they might have been surprised and experienced a lack in confidence, higher stress, and poorer performance due to ignorance of how the model would respond to their approach [17]. Nevertheless, the fact that assessment was performed 1 month after initial didactic and training reduced this déjà-vu phenomenon in the SIM+ group and might have counterbalanced the lack of familiarity with the task trainer of the SIM− group.

Discussion about the primary objective

The results of the present study on the benefit of simulation-based education for surgical chest tube insertion are similar to those reporting the increase of success rate for non-surgical chest tube insertion after simulation training [15–17, 19]. One distinctive feature of the present study is to have reported that simulation increased not only the success rate but also the performance assessment score [27]. This approach could reveal that, as expected, the specific “surgical” steps of the procedure were those that benefit the most from simulation training. In fact, there was a high correlation between success and performance score. During the validation process of the performance assessment scale, a score ≥14 was always found to be associated with a functional chest tube [27].

Discussion about secondary objectives

Simulation-based education was not associated with a decrease in procedure time for surgical chest tube insertion, as one would have expected. However, dissection time—the crucial step of the surgical procedure—significantly decreased in the SIM+ group. This implied that failure of surgical chest tube insertion was associated with poor mastery of this major step of the procedure [20–22].

Surprisingly, there was no effect of status or previous experience on the performance or success rate of surgical chest tube insertion [5, 30]. This fact might be related to the rarity of this technique (surgical approach) in France and a consequent need for adequate training, according to the recommendations [18]. Besides that, it underlines the fact that only simulation-based training was responsible for the gain in performance and success rate.

Compared with SIM− participants, SIM+ group participants judged the training model more realistic (with more anatomic features), probably because they had been trained on it repeatedly a month earlier. Similarly, there was a higher gain in self-confidence in the SIM+ group. But these differences were minimal. Indeed, self-confidence is a major issue in the performance of a stressful procedure or when the practitioner is uncomfortable with the procedure because of his/her scarce experience [11, 12]. The satisfaction rate was high in both groups—more pronouncedly in SIM+—thereby implying that participants are particularly receptive and attracted to simulation-based education as a means of improving technical skills [14, 26, 30, 31].

External validity

We think that simulation-based education for surgical chest tube insertion can be spread to other groups of learners with the same results, in initial learning (residents) as well as continuous education (emergency physicians, intensivists) [5, 6, 13–15]. We observed that one simulation attempt did not suffice to achieve 100 % success for the whole group. This observation underscores the great interest of repeated simulation training for low-volume/high-stake procedures [20–22]. Finally, the assessment session was scheduled 1 month after the initial didactic session in both groups and simulation training in the SIM+ group. This delay might have been short as regards the occurrence of clinical surgical chest tube insertion (in our occupational field) and performance scores may have been better in simulation than in reality [16, 20, 30, 32]. Future studies should investigate if practicing on a task trainer improves the performance of the technique when performed on human subjects.