Log in Article Discussion Edit History Go to the site toolbox

HTM ComDoc 10.

From HTMcommunityDB.org


This page has been updated. Go to HTM ComDoc 10

Alternate Maintenance Strategies and Maintenance Program Optimization

(This document was last revised on 8-7-18)

10.1 Introduction

In today's world we have become increasingly dependent on highly mechanized and automated businesses. We depend more and more on services such as an uninterrupted supply of electricity and society seems to have set high and perhaps unrealistic goals such as expecting all of our trains and other forms of public transportation to always run on time. Such expectations depend on the continued integrity and proper performance of all of the relevant assets (physical plant and equipment). In reality equipment failures have played a notable part in some of the worst accidents and environmental catastrophes in industrial history; so understanding the causes of equipment failures, what can be done to prevent those failures and how best to execute and manage the appropriate failure-preventing processes has become an important priority in the modern industrial world.

The same should be true in the world of healthcare. Modern medical equipment has become an important part of faster and more accurate diagnoses and more effective therapies - resulting in far better medical outcomes. So it is important to ensure that progress in the methods and techniques used in medical equipment maintenance is kept in line with the progress that is and has been made in the methods and techniques used to optimize equipment maintenance in the wider industrial world.

In this article we will describe the four different maintenance methods that are currently being used to reduce how often various types of medical equipment fail. In the current jargon these different methods are known as "maintenance strategies".

The following four maintenance strategies (methods) are used for virtually 100% of the medical equipment planned maintenance (PM) performed today.

  • Traditional fixed interval preventive maintenance. This method uses one or more device restoration tasks performed periodically to replace or refurbish one or more of the device's non-durable components (i.e. those components that are either not intended by the manufacturer to last the entire working lifetime of the device, or have been found not to do so in practice). In the modern world this strategy is often combined with periodic safety verification (see below).
  • Predictive maintenance (also known as condition-based preventive maintenance). This method uses some means of monitoring the wear or degradation of one or more of the device's non-durable components in order to trigger a just-in-time replacement or restoration of the failing component(s). The most common predictive maintenance method used for maintaining medical devices is visual inspection. While the inspections may be performed at regular, fixed intervals, any required restorations are performed at longer, more efficient, just-in-time intervals.
  • Periodic safety verification. This method uses one or more safety verification (SV) tasks performed periodically to determine whether or not the the device is still meeting all of its critical performance and/or safety specifications or whether it has developed one or more imperceptible ("hidden") failures
  • Light maintenance (also known as run-to-failure maintenance). This method makes no attempt to prevent any potential failures; it simply allows the device to run until it fails - at which time it is then restored to proper working order.

We will address below how to decide which of these four possible maintenance strategies is the most appropriate to use for different types of medical device.

10.2 Traditional fixed interval preventive maintenance

Probably because of our many years of experience with automobiles and domestic appliances such as washing machines, particularly when these predominantly mechanical machines were simpler, there is a long-held and widespread belief that all machines eventually simply wear out. We think of them as having a useful working life of several years, after which they enter a period when they are expected to break down more frequently. We expect the day-to-day stresses from things such as fatigue, corrosion and oxidation on the device's various components to result in an overall age-related degradation of the device that will cause the entire device to either stop working properly or break down completely.

The traditional remedy for this has been to "rejuvenate" the device through periodic overhauls. If the device has a way of logging the time that it has actually been in use, then these traditional overhauls were set for intervals such as every 5000 or 10,000 hours. This periodic refurbishment was the predominant maintenance strategy in the aircraft industry prior to the introduction of the jumbo jets in the 1960s. And until quite recently the traditional maintenance strategy in the automobile industry has been similar, with regular service intervals set according to mileage/time combinations such as every 5000 miles or 12 months.

In RCM these simple, time-based or usage-based maintenance tasks are called scheduled restoration tasks or scheduled discard tasks, depending on whether or not the component that is expected to deteriorate is to be refurbished or replaced. A very large number of the manufacturer's recommended PM procedures for medical devices currently consist of sets of device restoration tasks that are performed at regular intervals.

The all-important question with the fixed interval method is deciding what the maintenance interval should be. Logically, the target component needs to be restored at a point in time before it ages into the wear-out zone where it becomes much more likely to fail. The term used to describe this is the device's "useful life". At this point in time the only source for this information is likely to be the device manufacture - which is probably why for many years the regulators stipulated that all medical devices should be maintained according to the manufacturer's recommendations. If the device manufacturer does not have an adequate amount of data on the mean time between failures for any of the device's non-durable parts (which seems quite likely) then they are probably in the best position to set the interval using other information that they should have available. If the consequence of the device failing because of this component failure is potentially serious, the manufacturer is very likely to be conservative and choose a relatively short interval.

For this reason it can be argued that this method is likely to be the most labor intensive way of maintaining a medical device, and it should be reserved for devices that have PM-preventable failures with outcomes that could have a serious, life-threatening outcome, and that are quite likely to occur. The Task Force has tentatively set the threshold for this likelihood of failure (quite likely to occur) at once every 75 years.

For devices classified as PM Priority 1, with parts that the manufacturer indicates need periodic restoration, the Task Force recommends that they continue to be subjected to the manufacturer-recommended PM (both interval and tasks) and to routinely monitor the level of patient safety being achieved (as described in Section 1.Z of HTM ComDoc 1). This should be continued until there is acceptable evidence in the HTM community database (Table 13 and Table 5) that there are other procedures with either more efficient tasks and/or a longer interval that are demonstrating the same or better levels of patient safety.

10.3 Predictive maintenance (also known as condition-based preventive maintenance)

Although the title of this type of maintenance methodology might create the impression that it involves consulting with some kind of mysterious oracle, it is instead based on identifying some kind of dependable indicator that a specific kind of device failure is in process. For example, the onset of a bearing failure in a motor is usually accompanied by detectable vibrations and by a detectable increase in the number of particles in the lubricant - both of which can be monitored and used to provide an early warning that the bearing is beginning to wear and may be about to fail. In practice there are many techniques that can be used to provide this type of advance warning that certain kinds of failures are in process.

In RCM, tasks using one of these techniques are called predictive tasks or on-condition tasks. This latter term is why this particular alternate maintenance strategy is sometimes called condition-based preventive maintenance. Of course, this method depends on being able to identify some kind of reliable indicator that can be monitored at a reasonable cost and which will provide an adequate period of forewarning. The limits of acceptability for both the cost and the length of forewarning will depend on the potential severity of the outcome and the economic impact of the targeted failure. Possible outcomes involving multiple deaths or a major environmental incident would obviously justify investments in this more elaborate methodology if they are technically feasible.

While methods such as infrared scanning for hot spots in thermal insulation and vibration monitoring of compressors and electrical generators have found favor in the physical plant area, the use of such techniques is less frequently used in the maintenance of medical devices. One important exception to this is the widespread use of generic visual inspection tasks to detect obvious-to-the human-eye indications that a component is beginning to fail. Many of the recommended PM procedures for both critical and non-critical medical devices include an initial task to "... inspect/ clean the exterior housing especially any moving parts, including any user-accessible areas that are under covers...." Although this kind of task is quite non-specific it is obviously quite prudent and inexpensive to perform.

While this generic visual inspection task is quite common in PM procedures for medical devices, there may be other, more specific visual inspection tasks. For example, the HTMC generic PM procedure for all line-powered devices includes a visual inspection task to "Check that the physical condition of the power cord and cap, including the strain relief, is OK. Check the auxiliary receptacles (if applicable). Check the circuit breaker/ fuse" and those with connections to the patient include a VI task to "Inspect all cables, electrodes and transducers to confirm their integrity and proper function". There are several other examples of these visual inspection tasks in the MPTF's collection of HTMC generic PM procedures listed in column C5 of Table 1

Maintaining a medical device using this "just-in-time" method will be more efficient than the fixed interval method because the restoration tasks are not performed until the condition of the part calls for it. However using just simple visual inspections to identify the level of deterioration, rather than a more reliable wear indicator, might be problematic because the accuracy of the wear detection is dependent on the skill and experience of the observer. Be aware that at least one RCM text (HTM ComRef 1.) warns that condition-based preventive maintenance can be a good way of predicting and managing failures "but it can also be an expensive waste of time".

For devices with parts that the manufacturer indicates need periodic restoration and that are not classified as PM Priority 1, substituting a visual inspection task at the same interval as the scheduled restoration or replacement of the non-durable part (i.e. changing to a predictive maintenance strategy) would improve maintenance efficiency. This might work well for those who are be reluctant to go all the way to changing to a run-to-failure (light maintenance) strategy.

10.4 Periodic safety verification (using performance verification and/or safety testing tasks)

A particularly troubling circumstance in medical device maintenance is the possible occurrence of "hidden" failures, especially when the undetected deterioration could result in a life-threatening outcome. A "hidden" failure is one that is unlikely to be noticed by the device user. In situations where the imperceptible or "hidden" failure might be life threatening it is very important to subject the device to periodic performance verification and safety testing tasks to provide assurance that the device is still performing safely.

Which tests should be performed? Typically, when the device is considered capable of developing a hidden failure that could be life threatening, the manufacturer's recommended procedure will specify appropriate tests. For example, the HTMC generic procedure for an anesthesia machine (ANES-01) lists nine safey verification tasks, seven of which are considered to be potentially critical. The HTMC generic procedure was developed by analyzing the PM procedures recommended by a cross section of anesthesia machine manufacturers.

Which SV tasks for which devices should be considered worth doing? Devices that have demonstrated that they are "quite likely" to experience a particular type of "hidden" failure - as evidenced by data showing that this particular type of "hidden" failure occurs with this type of device on the average more than once every 75 years should be considered worth doing. The relevant SV task is that recommended by the manufacturer for determining this particular type of "hidden" failure.

How often should the relevant SV test be performed? On first consideration, since hidden failures are usually caused by random failures of one of the device's components, it might seem to be prudent to test for such "hidden" failures as often as possible, particularly if the outcome of the hidden failure could be life threatening. However, as we outline in HTM ComDoc 6. "Choosing appropriate PM intervals", it is possible to come to a more nuanced conclusion.

In that article we show that the fraction of the SV testing interval during which the patient will be exposed to the "hidden" failure - which the MPTF calls the "average exposure (of the patient) to the hidden failure" or AEHF - and which is an excellent measure of the level of patient safety with respect to possible hidden failures - is related to the SV testing interval and the observed mean time between failures (MTBF) of the hidden failures - by the following equation:

AEHF = 0.5 x SV testing interval / MTBF

So, for an MTBF of 50 yrs and a testing interval of 6 months, the average exposure to the hidden failure (AEHF) is:

(0.5 x 0.5 / 50 =) 0.005 or 0.5%

And for an MTBF of 50 yrs and a testing interval of 12 months, the average exposure to the hidden failure (AEHF) is:

(0.5 x 1.0 / 50 =) 0.01 or 1%

With greater MTBFs of the "hidden" failures the AEHF becomes proportionately lower (better). For an MTBF of 250 years (which we consider a quite likely value) the AEHFs for testing intervals of 6 and 12 months are 0.1% and 0.2% respectively. We believe that the difference between the indicated levels of patient exposure to a possible hidden failure (the AEHF percentage) when doubling the testing interval from 6 to 12 months – whether it is from 0.5% to 1% and particularly when it is from 0.1% to 0.2% is not really a significant overall increase. The impact on the AEHF of using the longer testing interval is quite small, which makes it very reasonable (and very convenient) to conduct the SV testing at the same interval as might be required for any non-durable part restoration tasks. If the device has no non-durable parts and consequently no designated interval for NDP restoration, then it seems reasonable to choose a traditional (and entirely practical) interval in the range of 6 months to 4 -5 years, although the choice might be influenced by the actual MTBF of any recorded failures.

For devices classified as PM Priority 1, with no parts that the manufacturer indicates need periodic restoration, the Task Force recommends that, to be conservative, they continue to be subjected to the manufacturer-recommended PM (both interval and tasks) and to routinely monitor the level of patient safety being achieved (as described in Section 3.10 of HTM ComDoc 3). This should be continued until there is acceptable evidence in the HTM community database (Table 13 and Table 5) that there are other procedures with more efficient tasks and/or a longer interval that are demonstrating the same or better levels of patient safety. For devices not classified as PM Priority 1 devices, the arguments discussed immediately above indicate that extending the testing interval would result in minimal additional risk to the patient.

10.5 Light maintenance (also known as run-to-failure maintenance)

The RCM term for "light maintenance" is "no scheduled maintenance" and as the name implies this method entails making no effort to anticipate or prevent any PM-related failures to which the device might be vulnerable, and so those failures are simply allowed to occur - at which time they are then repaired. This method is also known as "run-to-failure". Clearly it is the most efficient approach to maintenance and is the recommended approach for all except devices classified as a PM Priority 1 device. The acceptability of the level of safety resulting from this approach for all except devices classified as PM Priority 1 will determine whether or not this is the most cost effective approach for each the entire program. If the overall level of patient safety using this approach is less than acceptable then optimization of the maintenance program will require a more careful selection of the most cost-effective method for each individual device or group of devices.

Site Toolbox:

Personal tools
This page was last modified 21:54, 1 October 2018. - This page has been accessed 8,620 times. - Disclaimers - About HTMcommunityDB.org