1998 OPEN FORUM Abstracts
Lung simulators: How Important is it to have one in your department?
Thomas D. East Intermountain Health Care, University of Utah
There is a need to accurately test the performance of mechanical ventilators. This testing may be part of new prototype product evaluation, quality assurance during the manufacturing process, calibration in the field or as part of a routine service program. The following are published standards for the minimum performance standards for mechanical ventilators: ANSI Standard Z79.7, ISO Standard 5469.:1987, ASTM Standard 1100. Any testing device for medical devices should meet or exceed the minimum performance standards. The commercial test lungs manufactured by Michigan Instruments (Grand Rapids, MI), Ingmar Medical (Pittsburgh, PA), and Bio-Tek (Winooski, VT) all exceed existing standards requirements. These devices use a spring loaded bellows to create a mechanical lung simulation. Some of the models have two bellows to allow for a two compartment lung simulation. A few discrete resistances can be modeled using fixed resistive elements that are added in the tubing of the test lung. Compliance is adjusted by changing the spring loaded force on the upper plate of the bellows. Lung volume is measured by observing the displacement of the upper bellows plate, translating pressure into volume assuming fixed compliance, or by direct measurement with a flow meter. All of these vendors offer computer interfaces to automate data collection. These systems are specified to have an accuracy of 3-5%. The disadvantages of these systems are:
1) Accuracy of volume and flow measurement is either dependent on accurate knowledge of compliance or on the performance of an external flow meter. If this is not accurate large errors can occur.
2) The system is primarily passive. It is possible to mechanically lift the bellows and to simulate spontaneous respiration; however, this provides only a very simple spontaneous breathing model. The practical result is a system that is very difficult to use to test trigger sensitivity and spontaneous modes of mechanical ventilation.
3) The system cannot duplicate coughs, sighs, ventilatory dysynchrony, and other spontaneous diaphragm and chest wall components of respiration.
4) The choices of resistance are limited.
5) An entirely different set of bellows must be used for pediatric simulations.
6) The lung model will not tolerate ventilation with heated humidified gas
Essentially there are no tools that can be used to reliably test modern mechanical ventilators throughout the spectrum of their capabilities. Several years ago we designed a computer controlled lung simulator to model the mechanical properties of a human lung, chest wall and diaphragm. The initial version of the simulator provides a two compartment mechanical lung simulation with the ability to have chest wall and diaphragm activity modeled as triggering and spontaneous inspiration. Coughs, sighs, bucking the ventilator and other active exhalation maneuvers can be created. The system calculates real-time pressures, volumes and flows with 1% accuracy. A full analysis package is included which contains modules for work of breathing, triggering and vent servo performance. A full demonstration and an opportunity for hands on use will be provided.
The 44th International Respiratory Congress Abstracts-On-Disk®, November 7 - 10, 1998, Atlanta, Georgia.