Volume 9, No. 2 • Summer 1994

Medical Gas Contamination An Unrecognized Patient Danger

Erwin Moss, M.D.

Anesthesia personnel usually automatically assume that the medical gases delivered from the wall outlets in the OR are clean, correct, and safe. While crossed-pipeline accidents always receive significant publicity, an under-appreciated patient safety issue concerns possible contamination of these medical gases with substances or materials that could possibly harm anesthesia equipment and, directly or indirectly, the patient breathing these gases.


The original purpose of the APSF Subcommittee on Medical Gas Systems was to make anesthesiologists aware of the design, location, and problems of the life support system with which they work everyday, but which is beyond the walls and out of sight.

Coincidentally, the ECRI, a non-profit health services research agency, formerly the Emergency Care Research Institute, has devoted a special double issue January-February 1994 vol. 23 No. 9 1-2) to medical gas and vacuum systems (MGVS). This publication should be in the library of every anesthesia department since it contains a crash course in what anesthesiologists should know about their MGVS. The address of ECRI is 5200 Butler Pike, Plymouth Meeting, PA 19462-1298.

ECRI correctly identified issues that this committee recognized early in its research. There is a mass of regulations, codes, and standards published by organizations such as the NFPA JCAHO, OSHA, CGA (Compressed Gas Institute), ANSI, AIA (American Institute of Architects), UL, and at least a dozen others. These regulations, standards, and codes address every aspect of the MGVS in hospitals and ambulatory care facilities from design to the concentration of agents used to clean pipes and valves. Why then are there reports of cross connections and contaminated pipes or why are fortunes needed or spent to correct errors in construction and design?

The ECRI, in its article, asked is there “a paper tiger’ in all of these published regulations, standards, and codes? It discusses the “Problem in Enforcing Compliance’ and ‘The Devil is in the Details.” Even the JCAHO which updated its MGVS standards this year (APSF Newsletter Winter 93-94, Tom Nagle) “looks for evidence of a properly installed and routinely inspected MGVS only in the form of proper record keeping; it does not look for in-depth adherence to the standard during on site visits even insurers require that a MGVS meet certain tests and have documents pertaining to use and care … however, enforcement is lax and only documentation is required.’ ECRI further states that ‘although state and local building and fire codes may also regulate the construction and operation of MGVS and most state or local departments of health require certification of new installation before occupancy permits are issued … rigorous enforcement of the standard (NFPA) is spotty and depends on the interpretation of ‘authorities having jurisdiction’ and the ‘responsible facility authority.’ Those authorities having jurisdiction .usually rely on information from independent inspectors or contractors, who may or may not be fully knowledgeable about the current details of NFPA-99 or even know how to perform complete testing of the MGVS … notably in the United States, no nationally recognized agency certifies MGVS inspectors as competent.”

The article by ECRI confirms in no uncertain terms what the Committee on Medical Gas Systems of the APSF early identified and that is a lack of an accountable authority to coordinate and enforce the many codes and standards now in place. There should be an organizational chart with a specific agency at the top!

Included in the structure should be the education and credentialing of those involved in MGVS construction from design architects to plumbers. ECRI identified two private organizations involved in training of installers and verifiers, PIPE (Piping Industry Progress and Education Trust Fund, Los Angeles, California and Medical Gas Management, Bethany, Oklahoma). Mr. Fred Evans, President of MGM is a member of the APSF committee on Medical Gas Systems. Another organization cooperating with this committee is the American Medical Gas Institute, a nonprofit organization, located in Metairie, Louisiana. There is a common frustration expressed by these companies in that they deal daily with the problems of faulty design, construction, installation, inspection, and certification, only to have deaf ears turned to them when recommendations to correct the faults to meet NFPA-99 standards are made to administrators.

California has strong MGVS construction requirements resulting from the Sylmar Earthquake of 1971. ECRI explains that ‘state law demands that MGVS adhere to the requirements of NFPA-99 as well as other codes defined by such agencies as AWS, CGA, ANSI, OSHA, and UL. Contractors must be certified as competent by recognized agencies such as AWS (American Welding Society), PIPE and ACIA (American Construction Inspection Association).

Inspect the Inspectors!

It is important that anesthesiologists understand that new construction may be inspected and certified by an individual who may not have himself or herself been credentialed for the task. It is not unusual for a facility to request certification just before opening its doors. Construction has been finished, the walls erected and the verifier (certifier) is limited to what he can see and do! Hidden behind the wall may be errors in design, incorrectly joined pipes which are improperly hung or supported and unclean. It is important not only to identify the gas flowing from each outlet, but to have an analysis of purity including particulate, chemical, and bacterial contamination. Logically, inspection of the MGVS should be continuously performed during each step of construction by credentialed inspectors and before the walls are put up. Certification should be by a disinterested third party as is required by Canadian Standards. It is not unusual for the contractor to be the certifier of his own work. Tennessee, through efforts of Mr. Fred Evans and Mr. Pete Winbourne, retired from Ohmeda, is in the process of requiring third party certification and MGVS regulations much like California has. Interestingly, Armed Forces facilities require certification by a third party.

Again, anesthesiologists must involve themselves during the construction phase of their MGVS. They must be knowledgeable as to the NFPA-99 codes. An excellent reference is the ‘Health Care Facilities Handbook’ Fourth Edition published by the NFPA in which each code is explained in easy to understand terminology. Anesthesiologists should not hesitate to don a hard-hat and enter the construction area. They are the end users of the MGVS and should understand the complexity of this life support system of their hospitals.

Anesthesiologists, during their workday, turn on gases and the vacuum systems with little thought as to the purity of the gases or the complexity of the MGVS. At the same time, in other units of the facility, gases are flowing to infants in incubators or on ventilators, or to patients in the ICU, CCU and even the Emergency Room. Suction is in use in all parts of the hospital. Although deaths are rare compared to the total number of patients using MGVS, when they do occur, the event gains nationwide attention and is followed by awards in the millions. In a ten year period one company, Medical Gas Services of Lenexa, Kansas, reported 205 instances of cross connection of which 81 involved a cross connection of a gas to the vacuum system. The excellent Canadian Standard is a result of 23 deaths in Sudbury, Ontario, in 1973 due to errors in construction of the MGVS.

California, according to PIPE, had as of December 1993, 368 certified inspectors and 22 certified verifiers as compared to many states that have none of either. California also hired an engineer to evaluate MGVS plans while in other states, approval of plans may be based on what the contractor tells the state.

Although there may be other organizations teaching and certifying installers, inspectors, and verifiers, the number is small. PIPE and MGM have training centers as well as the American Medical Gas Institute.

All three organizations offer programs in all parts of the United States. Although performing a vital service, they admit that they have developed their own curriculum and certification criteria and that there is no supreme body that sets educational standards as exists in our medical education system.

The worse possible scenario, other than crossed pipe lines, is an error by the manufacturer in the filling of tanks with the wrong gases at the manufacturing site. This possibility is responsible for the recommendation of the constant use of an oxygen analyzer on the machine even though an oximeter is in use. An oxygen monitor would have alarmed when a cylinder filled with nitrogen in error instead of oxygen was put on he. However, these monitors are not routinely used in other parts of the hospital. Therefore, it is important to understand the regulations placed on the manufacturer of the gas supply who incidentally are responsible for maintaining the bulk gas supply at hospitals.

When gas outlets are certified, the concentration or purity of the oxygen, nitrous oxide, nitrogen, or medical air is documented. What is not routinely identified is contaminants that may be present in acceptable or unacceptable levels. Particulate, foreign bodies, and bacteria are not the usual part of a certification of medical gases.

Included in the list of contaminants are metal fillings, flux, teflon, carbon, carbon oxide, oil and its breakdown products, halogenated solvents, methane, carbon monoxide, nitrogen oxide, hydrogen fluoride, hydrogen sulfate, carbon dioxide, cement, dirt, vermin, copper, copper oxide, copper carbonate, iron oxide, sand grains, wood chips, sodium crystals, chlorine, halogenated refrigerants, desiccant dust, fibers, aldehyde, lint, water and odor.

CO Monitored But Odor Banned

While there is an acceptable level for carbon monoxide (5PPM), and as of 1993 there must be a carbon monoxide monitor on the medical air system, there is no acceptable level for odor. Any odor originating from a medical gas system must be traced to its source. It is often the result of bacterial contamination or oil in the system. The medical air system, because of moisture, is the most common site of bacterial contamination. However, bacteria can grow in spaces left in improperly joined pipes. Culturing of medical gas is rarely performed. While ventilators and respiratory care systems are a known source of infection due to bacterial contamination, the idea that the source could be beyond the

walls in the pipe systems is not easily accepted by owners or administrators possibly due to liability issues and the need to clean the systems once the contamination is identified.

Water that accumulates in medical air as a result of malfunctioning dryers can come out of air as dewpoint changes occur along the pipeline course. A dewpoint monitor and alarm is a part of a properly designed medical gas system. Remember, medical air is the result of compression of eight cubic feet of atmospheric air into one cubic foot of compressed air and that all contaminants in atmospheric air, including water and carbon monoxide, are therefore concentrated eight fold in compressed air.

The presence of iron or iron oxide (rust) is evidence of iron pipe being used against NFPA code somewhere in the MGVS. Once iron is documented after a nitrogen purge of a MGVS, a search should be made for the iron pipe. The iron pipe should then be removed and replaced by copper to meet the NRA code requirements.

Multiple Contaminants!

Documented contamination of pipelines includes dirt, sand, gravel, cement, rust, vermin, cigarette butts, and wood. This form of contamination results from the pipes and bulk gas containers being opened and exposed to construction debris and to the atmosphere. NFPA code now requires all pipes to be clean and capped at the factory. The bulk gas system is delivered and installed by the gas supplier and may be on the construction site with ports open to the atmosphere. The anesthesiologist would be wise to check the bulk oxygen site and the condition of the containers during construction. The largest particulate recently found was a bird that had been sucked into a medical gas system as a result of faulty construction of the medical gas pipeline. Chemical contamination can be the result of solvents used by the manufacturer to clean pipes and valves.

The major form of particulate matter contamination is the result of improper joining (brazing) of pipes and joints during construction. Brazing should be performed by a certified brazer, but in reality may be performed by a plumber unaware of NFPA code that requires brazing be performed with the interior, pipe purged with oil free nitrogen. When brazing is carried out in room air within the pipe, the oxygen content of the air can cause oxidation of the pipe at the temperatures required for brazing. The result is carbon, copper oxide, and carbon oxide.

Later, copper carbonate, recognized by its green color, is formed. AU contaminants can scale off the interior of the pipe, flow downstream and impair the function of equipment such as flowmeters, outlets, ventilators, and blenders in ventilators.

The brazing must be carried out at 1000 degrees F using a special joiner specified by NFPA-99 code. Nitrogen purging must continue until the pipe is cooled to touch, otherwise oxidation will occur. This relatively simple procedure, first put into code in 1993, can prevent the major share of particulate contamination in MGVS. Yet, new construction completed in 1994 has found particulate contamination not meeting NFPA-99 standards. Although the NFPA-99 describes the brazing technique, the contractor may have given the job to a low bid plumber who is not acquainted with the code.

Examination of a pipe system for particulate matter involves a white cloth being placed over an outlet permitting gases to flow through the cloth. The cloth acts as a filter. The color, amount, and size of the particulate matter will determine the degree of contamination while identification of the particles and size can be made by microscopic examination.

Purging Needed

Cleaning of a pipe system can be performed by purging with nitrogen or washing. Purging will blow out loose scale and is temporary while washing with an acid solution can be a permanent solution, but is expensive and complex in that washing is performed zone by zone with shut down of zones during the process or may require the complete shutdown of the medical gas system.

Once again, the anesthesiologist should involve himself or herself in any new construction or addition to present systems. The credentials of the designers, contractors, installers and certification expert should be verified. Even the brazer should be credentialed.

Inspection should occur during the various phases of construction. Certification should be by a credentialed expert. Pipe lines should not only be certified for gas identification, but for contamination. It is hoped that more can be learned about bacterial contamination and its prevention and treatment.

Coming Attractions

In future articles the following topics will be discussed:

1. Gas shutdown of hospitals during construction or enlargement of the gas systems in a facility.

2. The problems and responsibility of bulk gas supply.

3. The medical air system.

4. Recent updates in NFPA code.

5. Important points to watch for, as an anesthesiologist, in design, construction, inspection, and verification of a medical gas system.

Dr. Moss of Verona, NJ, has been very active with and is a consultant to the New Jersey State Society of Anesthesiologists.