Tired of following manufacturer recommendations for preventive maintenance? Here’s how to develop, implement, and monitor your own in-house guidelines
In February 2014, the Centers for Medicare and Medicaid Services (CMS) finalized changes to 42 CFR 482.41 (c)(2), the federal regulation that provides guidance on the performance of hospital maintenance activities to ensure “an acceptable level of safety and quality.” The revisions, initially issued in a December 2013 memo known as S&C 14-07-Hospital, were implemented in collaboration with The Joint Commission (TJC) and introduced significant flexibility to the existing maintenance requirements.
Historically, there has been no one-size-fits-all approach to ensuring medical devices were maintained properly. The implementation of risk-based, evidence-based, or time-based maintenance—or procedures derived from some conglomeration of these strategies—has been the norm for some time. Unfortunately, there are many organizations that have either failed to properly develop these programs or simply failed in their execution, resulting in less than satisfactory outcomes.
The changes made by CMS, if implemented in accordance with the agency’s intent, are fairly impactful. While simply stating that “hospitals comply with this regulation when they follow the manufacturer-recommended maintenance activities and schedule,” 42 CFR 482.41 (c)(2) goes on to say that “A hospital may, under certain conditions, use equipment maintenance activities and frequencies that differ from those recommended by the manufacturer.” So, what are the “conditions” under which we can deviate from published recommendations?
The simplest solution would be maintaining your entire inventory according to manufacturer recommendations, either through contract or your in-house program. However, this is both impractical and unnecessary. It is fairly safe to assume that many hospitals that claim to be maintaining their equipment in accordance with manufacturer recommendations aren’t. That may be because they have simply overlooked a procedure they feel may be unnecessary, unwittingly adopted an incorrect inspection frequency, or regularly use test equipment that is not specifically identified in the manufacturer’s literature. For each of us to maintain every device in our respective inventories while adhering strictly to manufacturers’ recommendations would be prohibitively expensive, according to published research on the subject. Let’s face it: Manufacturers’ requirements are not always that easy to achieve. In many cases they are simply instituted to absolve manufacturers of some form of liability.
The most efficient use of your time and resources is to trim the fat where safe and practical. CMS has given us this opportunity in the alternate equipment management (AEM) program. So, what is the AEM and how do we use it effectively?
As the name implies, the AEM is an alternative means (differing from that of the manufacturer) by which to maintain your equipment. This alternative does, however, come with a fairly significant level of responsibility and documentation, both necessary to ensure an effective program and to meet the intent of the standard. From here, we will discuss many aspects of implementing an AEM program, using the following document (hereafter referred to as A-0724) for reference: CMS Revised State Operations Manual Appendix A for standard §482.41(c)(2), A-0724 (Rev. 103, issued: February 21, 2014; effective: February 21, 2014; implementation: February 21, 2014). The following insights are based on policies we have successfully put in place at WakeMed Health and Hospitals in Raleigh, NC.
Having a clean inventory will pay dividends when making critical decisions regarding your AEM. By “clean,” we are referring to its accuracy and consistency. Do you have several device categories for the same device type (for example, monitor, physiological; monitor, vital signs; monitor VSM; etc)? Do you have model inconsistencies (such as 12-345b; 12 345b; 12 345 B; etc)? Do you have categories or models that need to be archived? If so, now is the time to make changes.
TJC EC.02.04.01 EP7 stipulates that “the hospital identif[y] medical equipment on its inventory that is included in an alternative equipment maintenance program.” Simply put, your inventory must identify in some fashion any device that is being maintained outside of parameters established under the manufacturer’s recommendations. We found the easiest method, based on our unique circumstances, was to add the prefix “AEM” to the description field of the asset record. The field is searchable, and we can run reports on it if necessary. You may choose to add a unique identifier to the device category or control number field instead, but you need some method of identifying AEM equipment on your inventory.
So how do you determine what equipment will be moved to the AEM program? A better first question is, how do you determine what makes it into your inventory at all? Some devices don’t even need to be on your inventory, and therefore will not be subject to inspection at all. Think of this as a prefilter for the AEM program. We found that by developing a flow chart using simple true/false logic and questions based on risk and patient impact, we could quickly establish if a device met the criteria to be inventoried.
Expanding on this, we developed a second flow chart to determine if a device should be categorized as “high-risk” or “nonhigh-risk,” which incidentally is another requirement called out in TJC EC.02.04.01 EP3. Finally, a third flow chart establishes the AEM or non-AEM status of the device, once again using true/false logic and simple questions based on service history, regulatory requirements, and data analysis.
These three flow charts were incorporated into two departmental policies named “inventory inclusion” and “inspection criteria,” both of which are now referenced in our medical equipment management plan. They help us satisfy several aspects of TJC and CMS record-keeping requirements, but the most important thing we gained from this approach is consistency. There is no question when a device hits the inventory about what its status will be or how we arrive at that determination.
In order to deviate from manufacturer procedures and frequencies, you must have some documented reasoning as to why you feel that changing how and when you perform a PM on a particular device will not adversely affect patient outcomes. There are some exclusions. Devices that cannot be placed on an AEM program include: those that have a specific maintenance requirement imposed by a federal, state, or local law; imaging/radiological equipment; medical lasers; and equipment for which you do not have sufficient maintenance history to support inclusion in an AEM program.
Essentially, regarding a specific piece of equipment, you either cannot deviate from the manufacturer’s recommendation, or you must have data to support your decision to do so. As stated in A-0724 regarding new equipment, “If a hospital later transitions the equipment to a risk-based maintenance regimen different than the manufacturers’ recommendations, the hospital must maintain evidence that it has first evaluated the maintenance track record, risks, and tested the alternate regimen.”
In an effort to establish a “track record,” we adopted a policy stipulating that a device new to our facility, or one for which we have no data, must have 2 years of maintenance history in order to be considered for the AEM program. During this period, the device will be maintained in accordance with the manufacturer’s recommendations. Incidentally, this requirement is also one of our decision points on our AEM flowchart, along with questions specific to federal, state, and local regulatory requirements and device type (imaging and lasers), which help tie the program together.
As for existing equipment, we developed a grandfather clause in our departmental policy for devices already in the inventory whose PM protocol deviated from manufacturers’ recommendations. Since they had previously been evaluated using an evidence-based program, these devices were transitioned directly to the AEM program, less the 2-year wait time. In every other respect, they are subjected to the same level of scrutiny as new devices. In conjunction with automated maintenance reporting from our maintenance management system, this clause gives us confidence that the AEM strategy employed is not detrimentally affecting device operation.
Alternate Maintenance Activities and Frequency
Once you have made a decision to include a particular device in your AEM program, you will be required to determine to what extent you will deviate from the manufacturer’s recommendations. This is where you have some autonomy as to how you establish your maintenance strategy, but you also will be burdened with accountability. There are three statements contained within A-0724 that are very important regarding your maintenance strategy:
1) “In developing AEM maintenance strategies hospitals may rely upon information from a variety of sources, including, but not limited to: manufacturer recommendations and other materials, nationally recognized expert associations, and/or the hospital’s own (or its third party contractor’s) own experience. Maintenance strategies may be applied to groups or to individual pieces of equipment.”
2) “The risk to patient health and safety that is considered in developing alternative maintenance strategies must be explained and documented in the AEM program.”
3) “The hospital is expected to adhere strictly to the AEM activities or strategies it has developed.”
The first statement is fairly benign—simply base the maintenance you choose to perform on best practices and your experience. One caveat I would add is that equipment suffers more wear and tear based on how and where it is used. You may find it necessary to have differing AEM strategies for the same device type based on application and environment. This also suggests that if you use a third-party vendor with an established AEM program based on a 300,000-piece inventory across five states, it may not meet your specific needs. Therefore, continuous monitoring of your program, which will be discussed shortly, is imperative.
Your knowledge of how and where the data you’re using was acquired also plays into the second statement. If you have chosen a strategy, what type of data did you use to make that determination and where did it come from? Was the data provided to you from an external source? If so, how can you be certain that you do not need to adjust PM frequency to accommodate your specific maintenance requirements? How and where are you maintaining the data for future reference?
The data used needs to be as thorough and accurate as possible. If the service history on a particular device suggests a minimal failure rate between PMs, it may be a candidate for a decrease in PM frequency. But what is the impact of the failure? Does the device employ a built-in self-test that would prevent its use on a patient? Are there consumable components that require replacement at the manufacturer’s suggested PM interval that, if allowed to expire, would be detrimental to the operation of the device? How about omitting certain steps to the PM process? What are the long-term effects? Are you unknowingly inducing a condition that would allow for a parameter to drift? If the device fails, does it fail in a safe mode or in a manner that prevents its use on a patient? What about run-to-failure devices?
There is a plethora of questions you can ask in order to suggest PM frequency or content changes, but the bottom line is determining the risk to the patient and having the data documented to support your decision. So, where do you get the data? Your documentation.
The third statement is something you should have learned over the years from TJC inspections. If you have something in your policy regarding an AEM program, do what you say you are going to do. There is nothing worse than an inspector calling you out for not following your own program. It should also be your call to routinely review and update your policies, as well as train your staff on what they include.
Documentation and Evaluation
A-0724 lists a number of must-haves related to documentation of equipment placed on your AEM program. (Yes, the guideline specifically states, “…there must be documentation indicating…”) For the most part, these requirements are fairly intuitive: knowing the risk to the patient, rationale for the new PM activity or frequency, the dates when maintenance activities were performed, etc. However, one of them bears further discussion: “Documentation of any equipment failures (not including failures due to operator error), including whether there was resulting harm to an individual.”
This statement sounds like standard work order documentation, but consider what CMS is trying to accomplish here. You have taken a device and, based on data, changed what the manufacturer has suggested be done during a PM in favor of your own recommendations. How do you know you’re right? Unfortunately, we can never be 100% sure, but we hope that through continuous monitoring we will make the proper adjustments to ensure patient and staff safety.
Work order documentation can be a tricky thing. In many cases, we are not able to generate a report based on the text of the repair, so trending data becomes difficult. Standardizing the text your technical staff uses would also be difficult. Let’s say you have a particular device on the AEM program whose sensor happens to be subject to a high rate of failure, and you just happened to omit that part replacement from the manufacturer’s PM. How can you identify if you need to add that step back into the PM process? It is unrealistic to create work order action codes for every possible failure type for every device in your inventory. For the most part, maintenance management systems can be difficult, if not impossible, to configure to report specific events like this one.
The solution we have found most effective is incorporating user-defined fields for any failure-related work order within our maintenance management system. By attaching three simple, mandatory questions, we can gather enough data to at least flag whether we have an event that may jeopardize the integrity of a device on the AEM program. The work order cannot be closed until these questions are answered with a yes/no drop-down menu.
The questions are quite simple for your technical staff to answer with a little training:
- Was the failure predictable? For example, was it caused by PM-related calibration drift, PM-related battery failure, or PM-related component failure (excluding user-replaceable accessories)?
- Would a change in the PM frequency or an addition to the PM schedule have prevented this failure?
- Would a change to the PM procedure or the addition of a procedural step have prevented this failure?
By answering these questions (which takes less than 10 seconds) the technical staff is giving us a searchable database with details specific to PM-related failures. Any “yes” answer finds its way onto an automated report that details the make, model, control number, date, and work order text surrounding the failure. This report is then reviewed by leadership staff and compared to trended data to determine the need for a PM procedural or frequency change, and is reported to the Environment of Care as quality data. One interesting aspect of this approach is that it’s not just for equipment on the AEM program. We report on all repairs, which also serves to evaluate the effectiveness of manufacturers’ recommended PM processes.
It is understood that not all maintenance management systems are created equal. You may not have the capabilities described above, but hopefully you can overcome and adapt your current workflow to attain similar results. This is a snapshot of the program that we have implemented, and it must be noted that the success of this program can be directly attributed to the collaboration and support of every member of our clinical engineering team. The development, evaluation, and revision of numerous policies and processes that make up the AEM program would not have been possible without their valuable knowledge, flexibility, and tenacity. For that, the credit goes to them.
An AEM program may not be for everyone, and it will involve a little work to implement. But the long-term benefits gained in flexibility and efficiency far outweigh the initial investment in time. Those who choose to institute an AEM program may also benefit by taking the opportunity to review and revitalize departmental policies and procedures that have been neglected over the years.
Good luck, and remember: If it’s not written down, it didn’t happen. More importantly, from a process improvement perspective, if it’s not written down, you won’t know what happened or how to fix it.
Dallas T. Sutton, Jr, CRES, is Supervisor, Imaging Engineering at WakeMed Health and Hospitals in Raleigh, NC. For more information, contact chief editor Jenny Lower at [email protected].
Photo credit: Copyright Michael Beer | Dreamstime.com
As A Biomed of 35 years experience, this “new” way of looking at PM raises a lot of “red flags”. there is something missing from the big picture here. we need a CMMS/TJC set of standards for manufacturers to follow when they write their PM procedures. The manufacturer needs be held accountable to write procedures that meet our(Biomed/HTM) professional standards. for example, I recently experienced a that company that makes a large surgical aspirator, used in wet OR locations, filled with conductive fluids, that functions close to the OR table, Does not only NOT specify electrical safety testing as part of their annual PM, but the factory techs they send to PM these devices don’t know what an electrical safety test is. The ECRI procedure (a professional standard)for Surgical aspirators specifies electrical safety testing. morally ethically and legally – who is right here? is this manufacturer legally negligent for writing a PM procedure that does not meet minimal professional standards ? Is a Biomed morally and ethically obligated to uphold the higher of the two standards. Or does a Biomed have legal loophole to not meet the higher professional standard if the manufacturer’s procedure is “less rigorous” than prevailing professional practice would dictate? If a CMMS or TJC or ECRI or AAMI employee reads this Post, please be sure to post a comment to the points made here.
ps- as an old school Biomed, I can see certain equipment types for which the EST may well be obsolete. But a device that lives in the patient zone, in a wet location and is full of conductive fluids 24×7 is one of the few things that should always have an EST as part of its PM.
Mr. Mackeil, I appreciate your comments (I’m ‘old school’ too) and I completely feel your pain. This article was written more to address a new approach rather than the practice of PM’s. While I agree that some manufacturer’s should have a more active role writing PM procedures that more closely align with the real world risks, we all have variations in application of the same device, and therefore what might seem as absolutely required for one organization, may be overkill for another. This is where I find so much value in the AEM program as proposed by CMS – it gives each of us the flexibility to make changes to our program based on the specific needs of our equipment and ultimately our customers. Thanks again for the comments.
I am happy to see someone sharing the experience of establishing an AEM program as mandated by CMS in Dec. 2013. As the readers can see, it is a lot more elaborate than the traditional “risk-based criteria” allowed by TJC, even though there is no evidence that maintenance omissions have caused a significant amount of patient harm (see the analysis of 8 years of TJC Sentinel Event data published in Biomed. Instr. & Tech, June 2014).
Conceptually, AEM is consistent with Evidence-Based Maintenance – EBM that I wrote about in the Soapbox of April 2007 issue of 24×7. In principle, AEM is superior to traditional methods because it requires not only careful planning and implementation, but also evaluation of the outcomes and revisions, if needed. Unfortunately, AEM implementation is not trivial, particularly the safety and effective evaluation required by CMS. As the WakeMed team realized, it is necessary to determine the cause(s) of the failures and discern whether at least one of the causes is a deviation from the manufacturers’ recommendations. The three questions adopted by WakeMed are essentially what has been called Predictable & Preventable Failure (PPF) in the four EBM article series that my former colleagues and I published in the J. Clin. Eng. between 2011 and 2013. The results reported in those articles showed that PPFs rarely exceeded 5% of all service records, which means it take a long time to accumulate enough data for statistically valid analyses. For example, if you have 100 units of identical devices used in the same environment, you can expect no more than 10 PPFs in two years. Such a small amount of records is typically not enough to conclusively prove that the AEM is as safe and effective as if the manufacturers’ recommendations were followed, especially when these PPFs have different root causes.
Unless the clinical engineering (CE) team is supporting a very large hospital or health system, or is part of a large independent service organization, it is very challenging to gather enough data to prove the safety and effectiveness of an AEM, regardless of how well it is structured and implemented. Obviously, the ideal solution is for many hospitals and systems to pull their data together to allow statistically valid analyses. This dream is, however, unlikely to happen anytime soon due to the different software systems used, inconsistencies in inventory nomenclature, diverse failure cause coding and, above all, the lack of cohesiveness among CE professionals.
Dr. Wang, I appreciate your comments as well, and while they are statistically accurate, they essentially leave individual hospital systems with little alternative other than to follow manufacturer’s recommendations on all inventoried devices. It would be interesting to see the actual data on the number of organizations out there that claim to be following manufacturer’s recommendations vs. those that actually are. In most cases this is impractical, unnecessary and cost prohibitive. Whereas a pool of data contributed by all clinical engineering organizations within a given group (region, country, etc.) would be ideal, this also fails to consider the individual application and environmental differences experienced at different organizations for specific devices. I believe the intent of the AEM Program is to allow for flexibility in program management while requiring organizations to monitor risks with due diligence. Without an AEM program, we lose the ability to modify our programs to meet the specific needs of individual equipment items based on their use and location, not to mention historical reliability. I believe it is also safe to say most organizations, having inventory items that meet a predetermined risk threshold, would be maintained according to manufacturer’s recommendation anyway. This infers that those items placed into an AEM program are generally of lower overall risk to the patient (and organization) based on the localized data, localized experience and the risk threshold of the owner organization. Once again, thank you for your comments and contributions to the career field.
Is there an AEM specification for test equipment?