Central venous catheter placement is an essential skill for critical care physicians, and competency is required by the American Board of Internal Medicine (1). Internal medicine residents are no longer required to place a minimum number of catheters, so incoming fellows may not be competent in this procedure. Ultrasound guidance is now standard for many vascular procedures, and the American Board of Internal Medicine strongly recommends proficiency in this area (1). Furthermore, the Agency for Healthcare Research and Quality rated ultrasound guidance for central venous catheterization highly as a patient safety practice (6).

Use of simulation for training and assessment of central venous catheter skills is well described (7–10). Although anatomically accurate phantoms are commercially available (Figure 1), less expensive phantoms can be created with readily accessible materials (11). Vascular access curricula and assessment checklists are described in the literature (7, 9, 10, 12). Although there is no established ideal checklist for training, validated tools do exist, and key components have been described (13, 14).

Figure 1. Phantom model for ultrasound-guided central venous catheter placement (Simulab). Image courtesy of David Fobert.

Download Figure | Download Powerpoint

[More] [Minimize]

Evidence supporting simulation for catheter placement is robust, demonstrating not only increased operator comfort and behavior in simulated settings but also improved patient outcomes (7, 15–21). The application of simulation-based mastery learning, which requires trainees to reach a minimum passing score before performing the procedure on patients, is an effective method of improving skills. Barsuk and colleagues demonstrated that this form of training improved confidence and technical skills compared with traditional training for internal medicine residents (7). Additional work by this group has demonstrated reduced catheter-related complications in the ICU. Trainees receiving simulation-based instruction performed fewer needle passes (mean, 1.32 vs. 1.74) and arterial punctures (1 vs. 14%). In addition, the simulation cohort had higher success rates (95 vs. 81%), and their catheters required fewer adjustments (17). There is further evidence that simulation-based training can reduce catheter-related bloodstream infections, a finding with major implications for ICU outcomes (15, 16). Although the literature supports simulation for central venous catheter placement, the quality of available studies is variable. Many are limited by small sample sizes from single institutions and lack controlled designs. Furthermore, there is no standardized definition of competency or universal approach to simulation-based training. Curricula typically consist of approximately 4 hours of training and include didactic sessions, hands-on practice, deliberate practice with expert feedback, and a checklist assessment of competence.

High-quality models for central venous catheter training are widely used, and evidence supports simulation-based methods. In addition to improving knowledge and confidence, studies have demonstrated a reduction in patient complications (Table 1). We support the incorporation of central venous catheter simulation training in light of available data and the paramount importance of patient safety.

Airway management is fundamental to resuscitation in the critically ill. Management is often urgent or emergent and performed on patients with little reserve, with significant comorbid illness, and under suboptimal conditions. It is believed that proper training in an optimal setting can position the practitioner to handle the airway in diverse circumstances. No guidelines exist for teaching airway management, and there is no consensus as to the training standards to achieve or maintain competency.

Current training methods are variable regarding content and hands-on experience; simulation has the potential to fill in gaps in these programs. A survey of critical care medicine program directors demonstrated that more than half of programs included a dedicated airway rotation, which varied in duration and was most frequently in the first month of training (22).

Techniques for airway management include supraglottic devices, intubating stylets, video laryngoscopy, fiberoptic intubation, and asleep and awake flexible scopes. Available models for simulation include high-fidelity computerized simulators, simple mannequins, cadavers, and standardized patients (Figure 2) (22, 23).

Figure 2. Phantom models for airway management (Laerdal). Image courtesy of David Fobert.

Download Figure | Download Powerpoint

[More] [Minimize]

More than 70% of program directors had educational resources to teach airway management that included didactic sessions, direct supervision, and simulation. More than half of the programs had a minimum requirement for proficiency. Unfortunately, the cases performed by trainees under direct supervision failed to meet minimum levels set to attain competence for the majority of the advanced techniques (22).

There is extensive literature regarding simulation-based training to improve knowledge and skills in airway management. A recent systematic review and metaanalysis of simulation education for airway management included 76 papers and 5,226 study participants (23). Comparison of simulation to nonsimulation showed a clear preference for simulation with respect to learner satisfaction, skills, and patient outcome but no difference in knowledge. Although greater learner satisfaction was derived using biologic models over synthetic models, this did not translate into meaningful clinical outcomes. Skill retention was augmented by postcourse interventions such as periodic skill evaluation, refresher courses, and self-directed practice. Interestingly, educational interventions coupled with operating room experience did not add to maintenance of skills (23).

Airway education data are limited in their ability to demonstrate an impact on clinically relevant patient outcomes. The frequency of retraining and the most effective strategy have yet to be elucidated. In addition, the methodology required to demonstrate meaningful change remains unclear. We recommend the use of didactic and simulation-based methods to acquire initial airway management skills followed by regularly scheduled refresher courses. The need for formal skills maintenance is based on the infrequent use of many rescue techniques in daily practice and the lack of time to reacquaint oneself with these techniques in the setting of an airway emergency.

Since data supporting the use of simulator training in flexible bronchoscopy were first published by Colt and colleagues in 2001, there has been a shift away from the apprenticeship model. Simulation provides a zero-risk environment for the provider to become familiar with airway anatomy and reach a level of proficiency in speed, accuracy, and dexterity required to perform flexible bronchoscopy (24).

Bronchoscopy simulators are classified as high or low fidelity. Low-fidelity models vary widely from nonanatomic “choose-the-hole” boxes to anatomically accurate representations of the bronchial tree. These models are less expensive, more readily available, and allow use of a standard bronchoscope. High-fidelity simulators consist of a flexible bronchoscope introduced into an interface that transmits movements to a computer and allows the operator to navigate through computer-generated airway images (Figure 3). Devices allow for realistic patient responses, including breathing, coughing, bleeding, and changes in vital signs. High-fidelity models offer clinical scenarios of varying difficulty and objective metrics for assessing performance. The cost of high-fidelity simulators often exceeds $100,000, making this method cost prohibitive for most training programs (25).

Figure 3. View of carina on screen of high-fidelity bronchoscopy simulator (Immersion). Image courtesy of David Fobert.

Download Figure | Download Powerpoint

[More] [Minimize]

Emerging data support the value of simulators in bronchoscopy training. Bronchoscopy simulators have been shown to accurately discriminate between expert, intermediate, and novice bronchoscopists with respect to procedure time, wall collisions, and percentage of segments visualized. One study demonstrated that novices trained with a simulator outperformed conventionally trained control subjects as assessed by procedure time and percentage of segments correctly identified/procedure time for their first patient bronchoscopies (26).

A recent systematic review and metaanalysis of simulation-based bronchoscopy studies showed that, in comparison with conventional instruction, simulator training was associated with significant benefits in procedural skills and behaviors (pooled effect size, 1.21 [95% confidence interval, 0.82–1.60]) and time (0.62 [95% confidence interval, 0.12–1.13] (27). A prospective multicenter study similarly found that incorporation of 20 simulated bronchoscopies resulted in significant improvement in speed of skill acquisition assessed by the Bronchoscopy Skills and Tasks Assessment Tool. The impact of this intervention was particularly evident during the first 30 bronchoscopies (28). There is a paucity of studies directly comparing high- and low-fidelity simulators. One study by Davoudi and colleagues directly compared these two types of simulators for transbronchial needle aspiration and showed that low-fidelity simulators were preferred in terms of ease of learning, realism, and overall as a model for needle aspiration (29).

Flexible bronchoscopy training varies considerably across teaching programs, and the optimal curriculum that integrates simulation needs to be more clearly defined. We recommend that simulators be included for bronchoscopy training. Given the limited data directly comparing high- and low-fidelity simulators, a recommendation cannot be made regarding the preferred system.

Robust data have proven the efficacy of endobronchial ultrasound in the diagnosis and staging of lung cancer, creating a need for effective training in this area (30–32). In this section we review existing literature addressing simulation-based endobronchial ultrasound training.

Simulation technology in endobronchial ultrasound is available in both low-fidelity and high-fidelity models. Low-fidelity simulators consist of models with realistic tubular airway structures surrounded by silicone spheres simulating lymph nodes in the mediastinum and, therefore, allow tactile feedback of how the scope and needle operate. High-fidelity simulators are computer-based machines consisting of a proxy scope, a robotic interface, and a computer with a monitor (Figure 4). The bronchoscope is inserted into a plastic face and is maneuvered in a three-dimensional image recreation of the airways. The interface tracks motion and reproduces the force felt during an actual bronchoscopy. The “virtual” patient breathes and coughs, and vital signs are monitored in real time. Systems mimic the upper airway to allow practice advancing the scope through the vocal cords. The learner can choose to examine lymph nodes, vascular structures, and surrounding anatomy as well as sample nodes with simulated needles. The software tracks performance metrics such as time of procedure, lidocaine used, wall collisions, percentage of segments entered, and success obtaining a sample (33).

Figure 4. High-fidelity endobronchial ultrasound simulator showing station 7 lymph node (Immersion). Image courtesy of Darren Tavernelli.

Download Figure | Download Powerpoint

[More] [Minimize]

National societies have published number-based guidelines for endobronchial ultrasound procedures, but there is currently no evidence-based consensus on how to train clinicians (34–36). Several studies have paved the way toward a systematic and validated approach by assessing the learning curve for this procedure and testing objective assessment tools (29, 37, 38). Stather and colleagues validated the use of an endobronchial ultrasound simulator to discriminate between different levels of experience and later showed that high-fidelity simulator training led to more rapid acquisition of skills compared with conventional training (39, 40). A study using simulation to assess skills showed that simulation-based metrics can be used as reliable measures of basic competence in endobronchial ultrasound (41). In a multicenter study, Wahidi and colleagues showed that a training curriculum with didactics and simulation resulted in a successful pathway to independent endobronchial ultrasound performance among trainees (38). Despite positive results from these small studies, randomized trials are needed.

Although there are limitations in sample size and study design in the available data, the risk of patient harm by training on live patients versus the proposed safety benefit of simulation was considered in the working group recommendations. We advocate for educational programs that involve acquisition of endobronchial ultrasound knowledge via a structured curriculum and development of skills using a low- or high-fidelity simulator before performing procedures in a patient care setting.

Pleural procedures are commonly performed and include thoracentesis, chest tube placement, tunneled pleural catheter placement, and thoracoscopy. Studies show that operator experience and ultrasound use are associated with higher success and decreased complication rates (42). Emerging data support the benefit of simulation-based training for pleural procedures (43–45).

Mannequin simulators are available for thoracentesis, chest tube placement, and tunneled catheter placement. Low-fidelity models are composed of synthetic skin, a chest wall, pleural cavity, ribs, and fluid to simulate pleural effusions (Figure 5). They can be placed in multiple positions to mimic real-life circumstances. Mannequins are reusable and cost approximately $5,000.

Figure 5. Phantom model for ultrasound-guided pleural procedures (Blue Phantom).

Download Figure | Download Powerpoint

[More] [Minimize]

A small body of studies has demonstrated the benefit of simulation-based education for pleural procedures. There are currently no negative studies or studies that suggest simulation is not beneficial. The Ultrasound-Guided Thoracentesis Skills and Tasks Assessment Test (UGSTAT) has been validated as a tool to determine the adequacy of thoracic ultrasound training on mannequins before clinical practice (46). In addition, improved skills were demonstrated with the use of simulation and deliberate practice in medical residents. Residents underwent a training session including simulated ultrasound-guided thoracentesis. Ninety-three percent of trainees achieved mastery within 4 hours (44). Duncan and colleagues demonstrated a decrease in complication rates with thoracentesis in a large academic pulmonary practice via experiential training in a zero-risk environment (43). Pulmonologists underwent training that included both cadaveric and mannequin models. In addition, ultrasound guidance was required for all thoracenteses, and a structured proficiency and competency standard was established. Experienced proceduralists directly supervised an additional 10 thoracenteses on human subjects after simulation training. The volume of procedures increased dramatically from 58 to 186 per year (P < 0.05), and the rate of pneumothorax decreased from 8.6 to 1.1% (P = 0.0034).

Although further high-quality research is needed in this area, current data support simulation training in developing mastery of thoracentesis skills and reduction in complication rates. Given the positive data, and the lack of negative data, we believe that training via simulation on mannequins before clinical practice should become standard of care in pleural procedures.

Critical care ultrasonography is now a standard skill, and key competencies have been defined (47). The Accreditation Council of Graduate Medical Education requires competence in pleural and vascular access ultrasonography and demonstration of knowledge of ultrasonography as an imaging technique for pulmonary and critical care fellowship training programs (48). Training must address three aspects of point-of-care ultrasonography: image acquisition, image interpretation, and clinical applications. Simulation is relevant to all three aspects.

The use of normal human subjects as simulated patients for training in image acquisition is the present standard in large ultrasound courses. Typically, instructors work with a small group of learners and one model, using deliberate practice with expert feedback. This method is the recommended approach for initial training in critical care ultrasound (49). Limitations include cost, logistics, and lack of pathology in normal models.

High-fidelity mannequins are an alternative to human models and available for both transthoracic and transesophageal echocardiography. These simulators couple a manual transducer with a computer-screen interface; as the learner moves the transducer, a corresponding image is displayed on the screen (Figure 6). A variety of pathology may be displayed, allowing identification of both normal and abnormal images. Simulators are costly (approximately $100,000) and may be impractical in large courses.

Figure 6. Parasternal long axis view on high-fidelity echocardiography simulator (CAE). LA = left atrium; LV = left ventricle; LVOT = left ventricular outflow tract; RV = right ventricle. Image courtesy of Thomas Edrich, M.D.

Download Figure | Download Powerpoint

[More] [Minimize]

Small studies suggest that the use of human models in combination with supervised scanning of patients is an effective training method for image acquisition (50, 51). A randomized trial of approximately 60 anesthesiology residents demonstrated that the use of a simulator led to improved knowledge and performance of transthoracic echocardiography on human subjects compared with traditional didactics (52). The benefit persisted even after members of both groups were taught using human models. Similarly, the use of simulation compared with traditional didactics led to improved transesophageal echocardiography image acquisition on anesthetized patients among anesthesiology residents at a single center (53). A randomized study of 46 anesthesiology trainees and faculty demonstrated that a high-fidelity simulator is not inferior to the use of a human model based on practical and written examinations (54).

We recommend that initial training in image acquisition be accomplished through deliberate practice on normal human models or high-fidelity simulators. There is no evidence that mannequins are superior to human models, and the cost may be prohibitive. We recommend initial training in transesophageal echocardiography be performed using high-fidelity simulators, given the invasive nature of this test and risk for iatrogenic complications. We believe that training in image interpretation and clinical application is best accomplished by supervised review of image sets, case studies, and study of standard literature on the subject.