Pre-Use Anesthesia Check Fails to Find Faults
Hazard [Health Devices Sep 1988;17(9):274-6]
A patient suffered cardiac arrest during surgery when an excessive amount
of anesthetic agent was delivered because of two malfunctions in the anesthesia unit that
went undetected during pre-use check.
The anesthesia system was equipped with both a flowmeter-controlled
vaporizer and a concentration-calibrated vaporizer. When the concentration-calibrated
vaporizer did not seem to produce adequate anesthesia, it was shut off and the
flowmeter-controlled unit substituted. However, an excessively high concentration of agent
was delivered, which was not discovered until later.
In ECRI's investigation, the primary fault was found to be a manifold
disconnection (or partial disconnection) at the outlet of the calibrated vaporizer. This
disconnection would ordinarily be easy to find during pre-use testing; however, a second
fault, a leak of approximately 400 mL/min from the flush valve into the common outlet,
masked the first fault, so that the system appeared to be working properly.
The system passed the tests used by the anesthetist because, even with
complete disconnection of the manifold, the flush valve leakage was sufficient to produce
up to 60 cm H2O pressure at the common outlet or in the breathing circuit during the leak test. (Note
that the U.S. Food and Drug Administration [FDA] checklist tests tolerate leakages of a few
hundred mL. If the FDA procedure had been used, it would have detected this problem, but
only if done carefully and the results thoroughly analyzed to confirm the problem. It is
not known what checklist was used by the reporting hospital.) The flow into the breathing
system was also sufficient to keep the ventilator bellows filled.
The tests in ECRI's Pre-Use Checklist for Anesthesia Units were devised to
locate such faults by separate leak tests for the machine and complete system. During
ECRI's investigation, a leak test of the machine (ECRI test 7), excluding the breathing
system, disclosed only that there was excessive leakage into the system manifold. This
leakage could not be occurring through the flowmeters, which indicated no flow. An
alternative source had to be found.
The second confirming test, which also explained what happened during the
procedure, was checking vaporizer output. A rough test of vaporizer output using an oxygen
analyzer (ECRI test 8) is described in our checklist. Had our test been used, the
calibrated vaporizer output (measured at the fresh gas outlet) would have been found to be
zero as the vaporizer dial was turned in steps up to 5%. However, the flowmeter-controlled
vaporizer output would have been found to be excessively high because the only diluent
flow would have been that portion of the flush valve leakage that went to the common
outlet and breathing circuit.
- Photocopy and use the ECRI Pre-Use Checklist for
- If the hospital uses the FDA checklist or similar procedures, perform
the tests or their equivalents completely and carefully. Be sure that you interpret the
results accurately. Add the tests for leakage in the anesthesia machine alone and for both
percent and vaporizer output using Tests 7 and 8 from ECRI's checklist or
- Anesthesia Unit Vaporizers [10-144]
- Anesthesia Units [10-134]
Cause of Device-Related Incident
Device factor: Device failure
Support system failure: Lack or failure of incoming and pre-use
Mechanism of Injury or Death