Electrical Fires in Hewlett-Packard Monitors Due to Saline Contamination 

A 64-yr-old woman undergoing coronary artery bypass grafting was monitored using our department's standard monitoring. All electronic monitoring was performed using the HP Component Monitoring System (CMS or Merlin) using both a six- and an eight-module satellite rack (Figure 1). Approximately 1 h after connecting the patient monitors and before instituting cardiopulmonary bypass, the monitoring screens on the CMS went blank. The smell of burning plastic was noted, and smoke was seen emanating from a power/data cable that connected the CMS to the satellite monitoring rack. The power cable was disconnected immediately from the satellite rack, and subsequently a yellowish white flame arced approximately 1 cm from the cable to the satellite rack. The main power switch to the CMS was turned off, which extinguished the fire. A CO²fire extinguisher was obtained from outside the operating room but was not needed. A new satellite monitoring rack (HP M1278A), a new power/data cable (HP 8120–5236), and all new monitoring modules were obtained and power was restored. The monitoring resumed and the remainder of the case was uneventful.

A similar event occurred in two other cases within 18 months. In one case, the fire occurred during similar circumstances. In the other, no patient was in the operating room, and the fire occurred when the anesthesia monitoring technician was setting up for the next case.

In accordance with the Safe Medical Device Act of 1990** and departmental policy, all of the electronic equipment involved in each fire was immediately withdrawn from clinical use. Visual inspection of the damaged components revealed that the fire had been confined to the monitor connector cable (HP 8120–5236), satellite rack, and male and female input/data connector (Figure 1, arrow). No extension to the printed circuit board was apparent. One corner of the male and female connectors corresponding to the location of the 60-volt power supply and ground pins was burned (Figure 2). Also noted on visual inspection of the rack and cable assemblies was a dried white powder found at each connector, metal connector shield, and on the cable where the fire occurred.

A written report was sent to the manufacturer and at their request, the affected components were forwarded to the HP Materials Engineering Failure Analysis Laboratory for immediate analysis. The white powder found at the burn areas was scraped from the connector and analyzed using an ISI/DS 130 scanning electron microscope with Kevex EDS system (Analytical Answers, Woburn, MA) and with an elemental analysis tool. This substance was determined to be sodium chloride.

The conclusion of the HP Failure Analysis Laboratory was that all three fires were caused by a combination of the build up of saline solution inside the mated connection including the 60-volt pin to ground pin potential. Current flowing through a connection contaminated by saline caused a short circuit and overheating occurred, leading to ignition of plastic connector components. It is possible that the close proximity between the power and ground pins on the connector contributed to the short. The ignition of plastic components produced the yellowish white flame noted on disconnection of the power/data cable from the satellite rack in all three fires.

The monitoring fire occurred while the transport monitor rack (HP M1276A) was powered by the CMS. This CMS monitor can provide up to 3.2 amps at 60 volts to the rack, whereas the transport monitor (HP M1277A) provides only 0.46 amps. The monitoring system was designed so that the transport monitoring rack (HP M1276A) can be powered by the CMS during the operating room period. For patient transport, the monitoring rack can then be transferred to the transport monitor, disconnecting the power cable (HP 8210–5236) before leaving the operating room. The differences in the lower available to the rack by the CMS versus the transport monitor may have been a contributing factor in causing the fire. Allowing unconnected cables to rest on the floor with subsequent contamination by spilled liquids also may have been a contributing factor.

Any fire in the operating room is of grave concern. In the two cases in which patients were involved, the electronic patient monitoring was also lost temporarily. We also recognized an additional potentially serious fire hazard related to the proximity of the fire to the breathing circuit, which contained 100% oxygen (Figure 1).

Three components must come together to produce a fire, and these are readily present in the operating room. These three components are said to make up a fire triangle [3]and consist of (1) an oxidizing agent, (2) a source of heat, and (3) a flammable substrate.#

Regarding the presence of an oxidizing agent, oxygen and nitrous oxide are not flammable themselves but readily support combustion. Their presence markedly increases the potential for a seemingly minor incident to degenerate into a conflagration. These components of the fire triangle have been and probably will continue to be in constant use in the operating room in the foreseeable future.

Since 1936, various types of electrical equipment have been introduced into the operating room, and saline-induced short circuits have been reported. [4–7]Each piece of electrical equipment has the potential to be a source of ignition because they are sources of heat.

Cotton towels, surgical drapes, plastic tapes and tubing, gauze sponges, eye ointments, cleaning solutions, adhesives, bowel gas, operating room Table foamcushions, and plastic housing of electrical equipment have all been implicated as flammable substrates. [8,9] 

Our efforts toward subsequent fire prevention have been directed toward three areas: equipment modifications, changes in protocol, and education.

We have worked in association with the equipment manufacturer to improve the safety of the monitoring equipment in the operating room environment. The six-module transport monitor rack assembly (HP M1276A) was modified to include a plastic hood (now standard on all monitor rack assemblies) over the connection wells to reduce the likelihood of spilled saline entering the connection wells.** The satellite rack, however, requires the user to connect and disconnect the power cable with every case. Contamination of the cable's electrical connection with saline still potentially could produce a fire. We also have suggested engineering changes to the manufacturer:(1) the CMS could be provided with current and module load-sensing capabilities such that when connected to a transport monitoring satellite rack, the current supply can be regulated; and (2) the proximity of the 60-volt direct current pin and ground pin in the satellite rack connector and cable connector could be separated to make arcing less likely.

A more rigorous visual inspection protocol for monitoring equipment before use has been adopted by anesthesiologists and monitoring technicians. Every effort is made to keep the monitoring cables off the floor, clean, and free of liquid contamination per recommendation of and manufacturer. Monitoring cables are replaced every 6 months on a regular schedule. Satellite racks have been positioned away from the patient's intravenous fluids, monitoring transducers, and breathing circuit.

Fire safety and prevention education has been added to the anesthesia department curriculum for staff and residents. The quality assurance committee conducts a lecture on fire safety and prevention twice yearly. This includes a review of the principles of basic fire science, personal safety at a fire scene, procedures for patient evacuation, and fire suppression methods, including the use of water and CO²fire extinguishers.

Our experience with operating room fires related to electrical monitoring equipment has heightened our awareness of a fire hazard. Our approach of procedural changes, equipment modifications, and staff education is directed toward providing greater safety for patients and personnel in the operating room.

The authors thank Hewlett-Packard Company for laboratory support and Jill LaRue for assistance with preparation of the manuscript.

# National Fire Protection Association: Health Care Facilities Handbook, Fifth edition. Edited by Klein BR. Quincy, National Fire Protection Association, 1996.

** Code of Federal Regulations, Title 21, Department of Health and Human Services, Food and Drug Administration, Parts 803 and 807, 1994.