In mid-March 1996, the Associated Press reported on four deaths in a Texas hospital. The initial report attributed the deaths to contaminated oxygen. The A.P. release was not picked up by many major newspapers, but was reported in local Texas newspapers and other isolated publications.
Reports allege that the death toll has reached six with an additional 70 other patients under observation for hepatotoxic effects of the contaminant, which reliable sources have identified as trichloroethylene, a solvent used to clean pipes and gas tanks. This tragedy supports and emphasizes the APSF’s efforts to educate the anesthesia community about the dangers inherent in our gas pipeline systems, including the associated bulk supplies.
Trichloroethylene is an acceptable solvent for cleaning pipe to be used to carry oxygen but, according to National Fire Protection Association publication 99, this solvent can only be used in the factory in which pipes and valves are manufactured and assembled. The solvent is used to clean oil from the interior of valves and pipes. After cleaning, the pipes and valves are supposed to be exposed to the atmosphere until odor free, then capped and delivered to the construction site. Trichloroethylene is not to be used at the construction site, according to NFPA 99 code. It is important to use oil free pipes and valves since the miniscule amount of oil left by even a human fingerprint can, under certain circumstances, ignite and cause a fire fed by the oxygen in the pipe.
In the Texas case, trichloroethylene allegedly was used to clean a bulk oxygen tank. The hospital put the oxygen contract out to bid yearly. As a result, the bulk supply system was changed by the new low-bid contractor. The new tanks installed allegedly had been cleaned with trichloroethylene and some residual cleaning solution was left in the tank, which was then filled with oxygen. This new oxygen source was put online and the volatilized cleaning solvent reached patients receiving oxygen in critical care units and other sections of the hospital, according to the reports. Available reports do not mention deaths or injuries to patients undergoing anesthesia for surgery. According to the initial report, clinical hospital workers noted an odor within 15 minutes of the tank changeover, but no mention is made as to when patients were switched to oxygen cylinders.
The lesson to be learned is that all oxygen should be odor free and that whenever a medical gas system is breached for any reason, certification of the gases coming from the system when the work is complete should be performed. Also, the simplest test would be to smell the gas before administering it to patients. An administration system that causes changing of bulk gas suppliers frequently has an inherent risk as evidenced by this incident. In addition, bulk systems are regulated by NFPA 50 “Standard for Bulk Oxygen Systems at Construction Sites,” a completely different code publication than NFPA 99 which regulates pipelines. To further confuse the issue, bulk nitrous oxide is addressed in a Compressed Gas Association publication (G-8.1). Whenever any work is performed on the gas pipelines of a hospital or any medical facility, especially change of bulk suppliers, the anesthesiologists should demand to be informed so that proper certification of the gas pipeline system is performed before putting the system back in use for patient care.
A positive note from this tragedy was the result of an astute hospital administrator in another state having read of the Texas deaths and shutting down his facility’s oxygen system when patients complained of an odor in the gas. All patients were placed on oxygen cylinder supply and a consultant called. The odor was again trichloroethylene and the source was a contaminated valve that had been soaked in the solvent over a year before its installation. A valve from the same manufacturing lot was implicated in a death of a child in another part of the country. The near catastrophe occurred in April, 1996. The valve was taken out of the pipeline system and replaced. The system was cleaned of residual trichloroethylene by vacuum applied to the various zones until odor free. The trichloroethylene can remain in a metal valve for extended periods by penetrating lubricants or by combining with particulate contaminants such as copper oxide (also in pipes).
The lessons to be learned from the Texas tragedy case include that there is a need for free communication about incidents such as this, so that others can learn and be alerted to dangers such as odors in pipelines.
It seems logical that if, in fact, the contaminated valve described above was from a manufacturing lot implicated in a fatality a year earlier, there could be a mechanism for a recall of the plumbing fixtures. Attempts sometime ago to involve the F.D.A. in dealing with pipeline perils such as this received the answer that what is “Beyond the Walls” is not considered a medical device (and, thus, not subject to existing regulation).
There is a need for education of plumbers to correct improper brazing which results in particulate contaminants that are, in and of themselves, a danger and they may act as a reservoir for other contaminants such as trichloroethylene and bacteria. This tragic accident could have been avoided by testing of each component before connections to the hospital system – as required by NFPA, which states “whenever a medical gas system is breached for any reason, certifications of the gases shall be performed. The simplest test of smell could have averted this tragedy.
Finally, there clearly is a need for education of anesthesiologists in the complex subject of our gas storage and pipeline systems, including an understanding of the numerous applicable regulatory codes and the lack of any one overall authority amongst the numerous organizations and agencies concerned with medical gases.
Dr. Moss, Executive Medical Director of the New Jersey State Society of Anesthesiologists, is a member of the Board of Directors of the APSF and Chairman of the Working Group on medical gas safety; Mr. Evans is President of Medical Gas Management, Inc., Bethany, OK, and a member of the Working Group on medical gas safety.