By William A. Hyman, ScD
The issues of preventive maintenance (PM) have been with us for as long as I have been involved in clinical engineering—which is a long time. The key technical questions are what maintenance tasks to perform and how often, with the sometimes hidden question of: “For what purpose?” There is also the administrative question of “Who if anyone is making us do it?”—an issue that often seems to have little to do with the technical questions.
Part of who is making us is whether the requirement is real or a myth. Thus, the answer to the question of “Why are you doing this?” may often be “because it’s required” without the person answering being able to tell you who it is that requires it, and being otherwise unable to substantiate the claim.
Or, they may cite a non-governmental standard, without being able to tell you an authority having jurisdiction that has adopted that standard as a binding requirement. Or, they might give a grandiose citation—such as “the government”—but no particulars.
Routine electrical safety testing falls into this category; “because it’s required” is often stated, even when this may not be the case. If there is an actual, citable requirement, there still is a potential difference between this and having a sound technical basis.
In addition, there may also be a legal risk management component to levels of maintenance, anticipating a scenario in which a piece of equipment fails and causes an injury and the level of maintenance was less than or different from what the manufacturer recommended.
Then, the plaintiff argument might be that if you had only followed the manufacturer’s recommendations, the event would not have happened. The manufacturer will likely agree because any alleged failure by the equipment owner may offset any liability the manufacturer might have.
Fortunately, device maintenance-related patient injuries are extremely rare. This is why the recent and ongoing hubbub about maintenance has been met in part with the rejoinder that there is no real problem that needs to be solved.
Returning to the issue of “what maintenance and how often,” it is good to review the technical bases for PM. One basis is that the condition of a medical device might deteriorate such that it becomes unsafe and that by checking, changing, and/or adjusting certain things, the path to unsafety can be interrupted.
For this to be true, there must be something that is observable or measurable that is related to the device becoming unsafe—and there must be a time at which the device is still safe, but the deteriorating condition can be detected.
Alternatively, there may be a situation in which a forthcoming failure cannot be measured, but some action can reduce its occurrence. In this case, it is good to actually know how this is the case. Changing consumable parts may fall into this category if it is actually known that certain parts have a limited life and that replacing them in a timely manner can prevent significant field failures.
Proactive changing of consumable batteries is an example here. Note, however, that such actions are related to safety only if there is an actual hazard associated with the failure, as opposed to, for example, inconvenience. Here, risk analysis plays an important role.
When assessing inconvenience, it is also appropriate to remember that a scheduled PM may also be inconvenient for the end user—although, hopefully, it does not come as a surprise. If the failure can be easily corrected, perhaps even by the end user, then preventing the failure may not be necessary. This is the run-to-failure model, which generally sounds worse than it is and perhaps should be called “run-till-service-required.”
Another reason to replace consumable parts that are still working is the efficiency that might arise from sweeping an area, rather than going to that area on an on-call basis. Even so, the efficiency of area sweeps depends, in part, on the true time cost of a service-on-demand event. This is not just the actual time of touching the device, but includes the time to get to the device—time that might be lost in some accounting systems.
In large, far-flung, multi-building mazes, the time from shop to device can be considerable. And that’s not even counting the conversations that might take place along the way, although such conversations may have value.
In determining the what and how, there have been some recent mandates to follow the manufacturer’s instructions—but whether the device manufacturer actually knows how much maintenance is required is an open question. If it is a new device, the manufacturer will not have had any direct field experience when the instructions were written, although some lessons might be learned from previous devices. It is also clear that manufacturers have reason to keep maintenance activities on the high side.
One such reason is that routine maintenance doesn’t cost them anything. Secondly, if the manufacturer is providing the service for a fee, or parts are being replaced, then they’re incentivized to keep the amount of service relatively high. These days, they also seem to be motivated to prevent others from servicing. Depending on who you believe, this is either to protect their service income or to protect patients from bad service providers.
Another non-safety reason for PM may be to control repair costs and to prolong the life of the equipment. For instance, mechanical devices may have serviceable components that, if they were to fail, could cause a high level of damage and associated repair or replacement.
Still, such an assertion should be based on actual data. In this regard, it seems to be increasingly apparent that repair can cost more than replacement. This is part of planned obsolescence, which seems to have accelerated in recent years. In fact, one could argue that out-of-control obsolescence is at an all-time high as exemplified by certain electronics that many of us have in our pockets.
There may also be value in PM that goes beyond the specific items on the work order—for instance, is there an eyes-on value from being face to face with the device? Such value, if any, would depend on whether anything was likely to be seen, along with the observational skills and effort of the servicer. It would be preferable if servicers did not fail to see or actively ignore observable conditions that might be present. It is also sometimes appropriate to not just fix an observed condition, but to investigate it with questions such as: “How did it get that way?” “Is it likely to happen again?” and “How can it be prevented?”
Unfortunately, this non-specific category also invites speculative second-guessing when adverse events occur. The question that then arises is: “If there had been a PM at an appropriate time before the event, would the servicer have seen the condition and acted on it such that the adverse outcome would not have occurred?”
Of course, anyone seeking to blame the facility or servicer would say “yes” to this question, although it is generally impossible to know the answer. At a minimum, the allegedly missing PM would have to have occurred between when the device was damaged and the harm incurred.
Equally important is the assertion that the servicer would have seen the defective condition that “must” have been present had they been there to conduct the PM, even though that condition was not on their worklist. Another issue with this non-specific “other” category is that there is no definable interval for such inspections to be both practical and effective.
To return to the title, it seems reasonable that “maintenance” should have technical value. If there is no value, then maintenance is as much of a misnomer as “preventive”. After all, a PM that has no value is simply a waste of resources. If we like the acronym “PM” for historical reasons, however, perhaps we should separate the letters “P” and “M” from having any meaning.
There is a precedent for this: ASTM once meant American Society for Testing and Materials, but it now officially has no meaning beyond its name. Some might remember JCAHO, which once was an acronym, but then became a name.
Similarly, I spent most of my academic career at Texas A&M University, where “A&M” once meant “agricultural and mechanical” but later dropped that meaning. We might note that other words could be given to “PM” without using “preventive” or “maintenance.” “Periodic meddling” comes to mind.
So, we are back where we started: Trying to answer the question of: “What procedures should be done, at what intervals, to achieve what specific goals?” Then there’s the question: “How will we measure if those goals are met?” Unless, however, someone with authority is making us do it—in which case, technical questions do not apply.
William A. Hyman, ScD, is professor emeritus, biomedical engineering, at Texas A&M University, College Station, Texas, and adjunct professor of biomedical engineering at The Cooper Union, New York. Questions and comments can be directed to 24×7 Magazine chief editor Keri Forsythe-Stephens at firstname.lastname@example.org.