A transfusion medicine physician in Ohio asks about the appropriate action to take when a monitored device has repeated alarm activations. He wants to know if there is a 'standard' for the number of unscheduled alarm activations that should trigger a comprehensive analysis of the cause and implementation of corrective actions. He was taught that an alarm's unscheduled activation should be minimal/rare .... and he still tries to follow that guideline.

However, recently he came across a laboratory where a monitored device's alarm goes off at least 2-3 times per month. The manager of the laboratory in question reportedly approaches the alarm activations in the following manner: "Well, we responded to it right away, so what is the problem"?

The following comments have been received.

ADDENDA Mar. 31, 2005

1. A colleague in the Midwest reports that, in his experience, alarm activations on monitored blood bank devices (e.g., refrigerators and freezers) are rare occurrences when alarms are properly calibrated and set. Frequent alarm activations require users to look closely at the device being monitored as well as the alarm doing the monitoring. He comments that the 14th edition of the AABB Technical Manual covers the frequency of alarm testing and gives general guidelines for conducting alarm tests. The AABB Technical Manual suggests that corrective action be taken when an alarm is unexpectedly activated, but that it is up to the lab staff to determine what action is necessary. He adds that during periods of extreme equipment usage (opening refrigerator or freezer doors frequently) alarms may activate even if they are properly adjusted and maintained. In his hospital they have experienced alarm activations caused by improper adjustment of temperature alarm set points. After adjusting their alarm test protocol and utilizing an alarm simulator they are now able to accurately and reproducibly test alarms, and have eliminated nuisance alarms. He offers the following website as a source of information on this subject www.scientificelectronicsllc.com. He concludes by strongly encouraging colleagues to carefully read the AABB Technical Manual chapter covering alarm testing, and to pay close attention to the footnotes regarding pitfalls and cautions.

2. A Quality Assurance Manager at a California blood bank reports that reacting to an unscheduled activated alarm is an appropriate immediate corrective action, but not necessarily sufficient. In her opinion, one must go beyond the immediate action and investigate for a root cause. When unscheduled alarms happen in her facility, the person responding to the alarm documents the probable cause of the alarm, and the documentation is reviewed for increased alarm activation frequency or abnormal trends. A root cause for an abnormal trend or change in alarm activation frequency might be related to a problem with the equipment being monitored. Alternatively, the monitored equipment might be fine but the alarm trigger settings need to be adjusted.

3. A colleague who has had extensive experience in more than one transfusion service laboratory and who currently serves as a regulatory consultant for an environmental monitoring system manufacturer agrees with the comments made by the individuals in addenda 1 and 2 above. In her experience when a monitored device begins to set off alarms with greater frequency, either the device being monitored might be at risk of failing or the alarm system needs adjustment. In either case, one or the other are in need of further attention. If necessary, a repair or adjustment can be made to the device being monitored, so that the panic of immediately needing to move products or losing them altogether can be avoided, in case the device suddenly fails. Finally, she cautions that it seems unwise for the manager (described by the Ohio physician) to demonstrate a reactive rather than proactive approach to troubleshooting.

ADDENDA Apr 2, 2005

4. A Quality Manager at a blood bank in Santander, Spain is concerned that the laboratory mentioned by the Ohio physician may have two problems:

1. A potentially failing device (refrigerator or freezer) - Since the device has repeated alarms, something might be failing, which could represent a potential risk to blood stored within the device.
2. An ineffective quality system - The staff do not seem to know what to do in cases such as the one described, and the facility seems to have no plans for preventing the recurrence of an incidence.

The Quality Manager adds the following comments regarding a failing refrigerator or freezer, based on his local practice:

• The immediate action is to check if the monitored device is breaking down, or is the problem merely due to the door being ajar, or other easily correctable situation. One should take into account that the intense usage of a piece of equipment (door opening and closing, particularly when the device is not full) can cause temperature shifts.
• Check whether the alarm limits are properly set.
• Calibrate alarms.
• If no device or alarm failure is apparent, call the maintenance team or the supplier, and have the machine tested and repaired. It may be leaking coolant gas.
• Make plans to have backup equipment available.
• Check how long the contents may have been "outside of acceptable limits", and evaluate the suitability of blood products within.
• Consider buying a new device.

The Quality Manager adds the following comments regarding the use of a quality system based on his local practice:

• Put in place a contingency plan for equipment failure specifying as much as possible, when, who, how, where, what... to do if failures occur.
• It is useful to define the maximum timeframe allowed for products to remain "outside of acceptable limits" in case an accident happens. For example, at his blood center they allow FFP to be out of limits (that is, temperatures higher than -30ºC) 1% of their life. Define what to do with those products (i.e. test them, label with a new expiration date...). Check it periodically.
• Check the state of equipment, some pieces could be near the end of their useful lives.
• He would seriously consider installing a centralized, computer-managed, alarm system.
• Above all, educate and train staff in these matters.

ADDENDA Apr. 8, 2005

5. A Vice President of a wireless monitoring system manufacturer reports that he agrees with much of what has been said in response to the Ohio physician's question about repeated alarm activations.  In particular, he agrees that alarm activations should be minimal, but whenever repeated alarm activations occur an investigation should be undertaken to determine a root cause.

He cautions that conclusions as to a root cause must be backed up with evidence, and a good monitoring system with a graphical representation of temperature over time can be useful in helping to isolate a root cause. As a supplier of wireless monitoring systems, his experience has been that a variety of conditions and/or problems can be seen and understood quite easily by a quick review of graphical evidence: defrost cycles that cause the unit to exceed the upper temperature limit; poor condenser performance (usually due to old age or lack of maintenance and cleaning) that causes an unacceptable variance from the average temperature; conditions where the variance from average temperature is acceptable, but the average temperature itself needs to be adjusted up or down to keep the temperature within the maximum and minimum limits; and temperature fluctuations that are due to usage-related activities such as restocking, if the temperature is stable when these activities are not taking place. Repeated alarm activations are seen most often with his own company's systems when initially installed or when customers have had no prior experience with a continuous, real-time monitoring system. Once such a system has been in place for a couple of weeks, it is possible to review data and make adjustments that will resolve substantially all of the repeated alarm cases. In a select few cases, the introduction of continuous monitoring has resulted in the evidence needed to justify the purchase of replacement refrigerators and freezers.

He concludes by recommending that blood banks and transfusion services consider the relative merits of monitoring systems that continuously and automatically collect data and can generate trend graphs on demand, versus systems that rely on collecting and analyzing data manually.

6. Editor's Note: Colleagues familiar with the use of continuous automatic temperature monitoring systems are encouraged to share their experiences.

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator