Virtual Coach for Practicing Clinicians
We conducted a study to compare the performance of a virtual coach (also known as intelligent tutor system, ITS, or integrated tutor) built into the central venous access simulator to that of an average human instructor. The data from this study, performed on resident trainees, are aiding the development of a virtual coach for practicing clinicians.
Average Human Instructor vs Virtual Coach Study for Central Venous Access
- Group A (n=27) trains with virtual coach
- Group B (n=27) trains with average human instructor
- Ultrasound-guided short and long axis techniques used for internal jugular (IJ) vein; landmark-based technique for infraclavicular subclavian (SC) venous access
- Participants were evaluated on 3 trials each for IJ access and SC access
- Success was defined as venous access within one try – 3 attempts (miss, pull needle back, redirect) allowed without pneumothorax or arterial puncture
- Proportion is the number of successful accesses divided by the total number of tries
The performance of the virtual coach group is non-inferior to the human instructor group for ultrasound-guided internal jugular (IJ) access (trials 2 and 3), and for infraclavicular subclavian (SC) access (trial 6). Most notable is the result for short axis ultrasound-guided IJ access because this is the most common approach used in our clinical setting. Having established evidence of non-inferiority, we are using the central venous access simulator and the virtual coach for a system-wide mastery training of all anesthesiology, surgery and emergency medicine residents performing ultrasound-guided IJ central venous access, annually.