SimSurg: A Low-cost Cricothyroidotomy Simulator with Real-Time Feedback System

[AshLu]

SimSurg:

A Low-cost Cricothyroidotomy Simulator

with Real-Time Feedback System

Abdul Aref (Hardware Engineer)

Joey Peroski (Systems Engineer)

Rob Przybylski (Software Engineer)

Dan Westcott (Software Engineer)

Introduction

Many trauma related surgical procedures cannot ethically be practiced by medical students or

inexperienced doctors. Therefore, medical simulators that provide high anatomical and procedural fidelity

are used. One of the most important things to monitor during such a procedure is the vital signs of the

patient. One specific procedure that this is important for is a cricothyroidotomy, in which an emergency

incision through the skin and cricothyroid membrane is made to secure a patient's airway during certain

emergency situations, such as an airway obstructed by a foreign object or swelling and a patient who is

not able to breathe adequately on his own. Current simulators that provide live feedback to the trainee

including the most necessary vital signs are very expensive. The amount of cases per doctor is further

amplified in many developing countries, with many of these clinicians not being able to practice before

being in the real life situation. High fidelity trauma simulators are in high demand in the developing

world, yet the training institutions in these countries do not have the means to acquire them. Therefore, a

low cost and high fidelity cricothyroidotomy simulator with a live feedback system that tells the clinician

the vital signs of the patient including heart rate, blood pressure, respiratory rate, oxygen content, and

ECG was developed.

Experimental Design

The hardware portion of the system consisted of a foam model of a human head, larynx and

trachea made of rubber-based materials, a differential pressure sensor positioned to take readings within

the trachea, and a circuit to increase signal strength from the sensor as well as to reduce signals with

frequencies higher than 3Hz. The output from the circuit to the software was fed into a data acquisition

board and analyzed using National Instruments LabVIEW 8.6 (a screen shot of the LabVIEW block

diagram can be found in Figure 1 in the Appendix). Two emergency medicine doctors, an

anesthesiologist, and an expert in clinical simulation helped to write the algorithms for the software. The

two different types of patients that algorithms were written to simulate were a healthy 25 year-old male

and a 75 year-old male with COPD and taking beta blockers. A user interface that displays what the

patient monitor information a clinician would see in the real life situation was also developed. Vital signs

consisted of heart rate and associated ECG chart; systolic and diastolic blood pressure; oxygen saturation;

and frequency of contraction of respiratory muscles. The simulation began at the onset of an airway

obstruction. Trainees started the experiment by making a 1 cm vertical incision through the skin and

cricothyroid membrane. Following opening of the hole via a 90-degree turn of the scalpel, a 6 mm

internal diameter tracheostomy tube was inserted. The cuff was inflated and the tube was secured. A bag

valve device was attached, and the trainee provided ventilation by compressing the bag once every four

seconds.

Vital sign information was displayed depending on existence and duration of an unobstructed

airway. Trainees began the experiment by making an incision into the airway and inserting a

tracheostomy tube. A resuscitation bag device was attached. LabVIEW acquired data from the

differential pressure sensor and circuit and displayed it on a voltage-time chart in successive 10-second

windows. Peak voltage values were obtained for each window with or without the bag being compressed,

representing the successful and unsuccessful establishment of an airway, respectively. Likert scale

surveys were completed by each trainee to determine usefulness and accuracy of the simulator.

Results

An image of the user interface that the physician sees can be found in Figure 2 in the Appendix.

The voltage data output from the differential pressure sensor was collected and analyzed. The average

Vmax value obtained while the clinician was not ventilating the simulator (no flow) was 2.46 ± 0.005 V.

The average Vmax value obtained while the clinician was ventilating the patient (airflow present) was

2.66 ± 0.010 V. Performance of an unpaired, two-tailed Student’s t-test, with an alpha value of 0.05,

yielded a p-value of 1.87E-20, indicating that there is a significant difference in Vmax over a 10 second

window between scenarios with no airflow and airflow present and that the simulation is repeatable. The

design and medical application of the simulator was approved by physicians via a survey that was

conducted (a copy of the survey conducted can be found in Table 1 in the Appendix). Through the

analysis of results discussed above, it was deduced that the acceptance criteria, which were differentiation

between no airflow and airflow, reproducibility, and physician approval, were met. Thus, the simulator

was suggested to accurately mimic realistic physiological responses and anatomical structure.

Discussion and Conclusion

As always, there are sources of error in this experiment. First, there is the differential pressure

sensor, which has to be calibrated every time we start using it. If calibration is not performed, then the

LabVIEW program may pick up false positives in terms of breaths. Also, another source of error for this

simulator is that fact that some of the electronic components have 5% tolerances, meaning that the value

of any of these components could be off by 5%. This would ultimately be another source of error in the

system. Yet another source of error in the system is introduced since the procedure is dependent on the

performance of a trainee. If the trainee inserts the tracheostomy tube improperly, or does not apply

airflow in the correct manner, then the results would be skewed, thus introducing another source of error

to the system.

The system can be improved in several ways to yield more accurate results. First, it would be

helpful to incorporate a live feedback feature in the LabVIEW program, where the physician can monitor

the patient’s vitals in real time instead of every 10 seconds. Also, another improvement that can make this

simulator better is incorporating more algorithms for various types of patients and conditions instead of

just the two current situations. This will allow for the residents to train with simulations of patients more

like what they would encounter in the field; not every patient is going to be a healthy 25 year-old male or

an unhealthy elderly person. Another improvement to the simulator would be to cover the chest cavity to

achieve a more anatomically correct model as well as to address safety concerns with the freely accessible

circuit.

Applications

Future work includes writing multiple algorithms for various types of patients and the

development of low-cost and high fidelity simulators with live feedback systems for other various trauma

procedures. Algorithms for passive patient recovery could be formulated to simulate the recovery of the

patient after the execution of a successful cricothyroidotomy. Also, a CO2 meter that displays the CO2

percentage of air released from the lungs could be incorporated to inform the physician if the patient is

exhaling or not. Another future direction would be to make the simulator into a portable setup. This

would allow the training of residents in remote locations of underdeveloped countries.

This system will allow the training of residents and physicians to be completed under a low-risk

environment. This will result in better-trained physicians who will be able to perform the

cricothyroidotomy procedure accurately and with less risk to the patient. The feasibility and impact of

training clinicians and medical students in low resource settings such as Ghana and other underdeveloped

countries in West Africa has shown much promise and work is continuing to grow in this area of research.

Attachment: Original Report