Featured Articles

Fatal Gas Line Mix-Up: How to Avoid Making This "Gastly" Mistake

Problem: Recently, we learned that an oxygen flow meter had been forced into a nitrous oxide wall outlet that was directly adjacent to an oxygen outlet in a radiology suite. The oxygen flow meter's index safety system, designed to assure connection only to oxygen wall outlets, was broken at the time of insertion. (All gas cylinders, flow meters, and wall outlets utilize a PIN index safety system, DIAMETER index safety system, or other system meant to avoid cross utilization of sources and components) Additionally, the technician was unable to distinguish blue (nitrous oxide) from green (oxygen) because the radiology suite was dimly lit in preparation for a CAT scan. Sadly, instead of receiving oxygen, the patient died of nitrous oxide poisoning. Although the index safety systems and color-coding of gas cylinders have been in use for years as constraints or cues to prevent this type of error, there have been numerous similar events of hypoxic injury reported anecdotally and appearing in the literature. Broken or modified pins used with gas cylinders also have allowed misconnections.

At first glance, you may ask, "Why is nitrous oxide available in radiology?" Indeed, it may be possible to eliminate nitrous oxide in some locations outside of the operating room (OR). However, mix-ups between other gases accessible through wall outlets, including air (and even vacuum outlets), would still be possible. Also, many anesthesiologists now administer anesthetics in locations outside the OR, so it may not be possible to remove this useful anesthetic from all non-OR settings. Invasive procedures are being performed more frequently in radiology, sometimes replacing the need for surgical procedures, and anesthesia coverage for endoscopic procedures has become more common. General anesthetics are also delivered in bronchoscopy suites, cardiac catheterization laboratories, plastic surgery clinics, and other locations where invasive procedures are performed. As one of our consulting anesthesiologists pointed out, "A basic principle in anesthesia is that no single agent is optimal, so we use multiple agents and, through their synergy, obtain a more favorable benefit-side effect profile. Removal of nitrous oxide in locations outside of the OR could contribute to two different standards of anesthesia care."

Separating nitrous oxide and oxygen wall outlets is one option, but it may not be economically feasible to relocate all of the gas lines within the walls in existing facilities. Eliminating access to nitrous oxide from the wall outlets in locations outside the OR is another option, but this could result in new risks and concerns. First, anesthesiologists would have to use cylinders of nitrous oxide if needed. With wall outlets, the multiple gas connections are spaced close to each other so that the tubing running from the gas source to the anesthesia equipment is localized and compact. If gas sources were separate, the tubing would not be localized, thus creating new hazards such as difficult access around equipment, tripping over tubing, accidental disconnection, obstruction from kinking, and tethering of equipment. The replacement of nitrous oxide cylinders and the risk of an empty tank also could create a new set of concerns.

Administration of the wrong gas has also been attributed to connecting the patient's tubing to the wrong flow meter, especially if color-coded "Christmas Tree" adaptors are used. Green adaptors are meant for oxygen, yellow adaptors for air. Although flow meters may be designed to connect to just one type of gas, the threading for connection between the flow meter and the adaptor is universal, so the wrong color adapter could be used. Thus, if staff rely on the color of the adapter to guide connections, the oxygen tubing could be misconnected to an air flow meter if a green adapter has been used in error.

Safe Practice Recommendation: Since it's likely that nitrous oxide will continue to be available in various locations throughout the hospital, action must be taken to prevent a deadly mix-up. First, be sure to standardize the type of flow meters, regulators, and connectors used throughout the facility, and use only those with intact index safety systems to help prevent misconnection. Assure that gas connections are observable (not hidden under a table or behind a drape) and that the labeling of all gas connections and sources is prominent and visible under the conditions that are actually present during use (e.g., dim lighting, crowded spaces). If a patient does not respond as expected to treatment with supplemental oxygen, consider the possibility that the wrong gas (or no gas) is being administered and check the flow meter and tubing connections. Also, consider using clear Christmas Tree adapters, forcing the user to look at the flow meter and wall connection itself. Also, a bad outcome from a nitrous oxide/oxygen switch would be less likely if an oxygen saturation monitor was used to provide an early alarm of hypoxia.

Biomedical engineering experts should perform regular preventive maintenance on gas wall outlets, gas cylinders, flow meters, and other related equipment to assure that all connections and connectors are intact and in good working condition. Only trained and certified personnel should be allowed to service, maintain, and use this equipment. Perhaps it's also time to consider new standards. While the index safety system has worked reasonably well, in light of adverse events like the one cited above, new standards may be needed to make it impossible for gas line misconnections. Changing the materials used to manufacturer pins and attachments so they are less likely to break or wear out is just one example of an improvement that could be made with renewed scrutiny. The Association for the Advancement of Medical Instrumentation (AAMI) has just approved a committee to develop national standards for tubing connectors, which will also address this important safety issue.

We thank Matthew B. Weinger, MD, San Diego Center for Patient Safety (CA), Allan Frankel, MD, Partners HealthCare and Institute for Healthcare Improvement (Boston, MA), and John Gosbee, MD, VA National Center for Patient Safety (Ann Arbor, MI) for their input into this article.