EBME Forums Forums Management Medical Equipment Management Risk-Based PM

Dave’s basic principle: That no reduction of safety or reliability be done (much like the Hippocratic Oath really!) but that the requirement for assessment, testing and preventative maintenance of equipment be assessed on the actual, evidence based, and provable requirement for that testing/assessment/maintenance. That, the OEMs requirements are NOT the bottom line but, rather, the application of a full risk analysis can better provide an effective system of PM schedules.

I would hope that those who are going to enter into this conversation have a understanding of the principles of risk management and/or have read and understood the various standards that apply (here in Oz that’s AS/NZS 4360:2004 but in other parts of the world other, but similar will apply, me thinks.) The principle is the same no matter what.

I would have to say at this point that I am deeply indebted to the folk at The Ottawa Hospital for their generosity in sharing their work on risk based systems with me and, I hope they don’t mind, via this forum with the world.

First, I would draw your attention to the file attached (Leo-Risk Based PM Scheduling.pdf). It is through this that I first became aware of the work in Canada. This gives a good general impression of the work that was done there. Second, have a look at the detail of the Risk Form shown. This shows the basic format of the assessment as done. (assesment form.pdf).

If we run an example of this (and for this I will use a Datex Tuffsat Spo2 monitor) then the results pan out as..

Function rating: Patient monitoring device: 5
Consiquence rating: mistreatment 3
Maintenance rating: None 0
Protection rating: Yes (3), No (6) 6
Failure rating: none 0
Use & availability daily 6

Total score 20

This score would then place the Tuffsat into a once per year PM category.

Now, I have to say that I think that this is an incorrect assessment and as such I would (and intend to) modify the scoring system. My reasoning is that, although this is a patient monitoring device it is not used in high acuity monitoring but rather in low risk ward type environments. This should reduce the score under the “Function” section. Similarly, under the “Consequence” section, in theory a false (low) reading could lead to treatment of a condition that is not real but in the ward type environments that this devise is used is that actually the likely outcome? I think not, so again, perhaps some modification would be required to lower the risk score here.

I have intentionally used the Tuffsat as the example because my feeling is that such a device need never be tested (I can hear the gasps of disbelief from half a world away). It is a hand held, battery operated device. It has no calibration or adjustment. In short, it works or it does not. The only problems I have seen are of the probe or the complete shutdown of the unit. What do we achieve by doing an annual PM? Answer, nothing. I would put this type of device into a “check function only on repair” category.

“But what about performance verification?” I hear some cry, to which I would answer, why? what’s the point? A performance verification is valid only at the time it is done and becomes invalid as soon as the device is moved/used/etc. If we are to provide a true verification then it would need to be done before every use. To do so once per year is, I think, a complete waste of time.

At the other end of the scale of course are those critical devices such as ventilators, anaesthetic machines, defibs, etc. A quick check on an ICU vent, for me, scores a 46 (that’s 12,12,12,0,6,4 going down the check sheet). This quite rightly shows a requirement of 6 monthly PM. I have no argument with this requirement.

So, pros and cons……

Pro..

A risk based approach to PM can provide for continuing levels of equipment reliability and safety while reducing the resource needed to support this level of “up time”.

We are all aware of the increasing levels of equipment use within the healthcare sector but are also aware that there has not necessarily been an increase in the funding or staff levels in BME departments to match it. Risk based PM provides a way to use the available resources in a more efficient way, focusing the available resources where they are needed and away from areas they are not.

Con..

Risk based PM may actually reduce the resource requirement to such an extent that some staff are no longer required and budgets may be reduced or the level of income to the department will drop.

Well, yes, potentially, but is that a bad thing? We all, both in the public and private sectors, have a responsibility to those who fund us to provide best value. Risk based PM provides a channel by which this can be done. To take a Sir Humphrey Appleby like attitude of “the more staff we have the more important we are” is not reasonable in today’s environment. The tax payer (and that is you after all) does not have endlessly deep pockets. We all need to look for efficiencies in what we do.

Couple of other observations.

I took the form shown here down to one of our client hospitals to try them out and sat down with the nurse unit manager of ICU/CCU. I found that we had somewhat differing views in some areas. One of the examples used was the 12 lead ECG machine that lives in the unit. It was clear from the discussions that the NUM that she viewed the function of that ECG machine as absolutely critical to the running of the clinical unit. I did not share this view. This is not the only ECG machine available (there was another in a ward just round the corner) and the monitoring system used in the unit had 12 lead capabilities as well. To my way of thinking, these were viable alternatives in the event of a breakdown she however saw only that clinical practice required that a conventional ECG machine must be available in the unit at all times. My point here is two fold. First, some measure of the clinical practices should/needs to be taken into account when doing a risk analysis and secondly that there may well be a difference of view between biomedical and clinical groups. This is an issue that would need to be overcome for such a process to be put in place and accepted.

Some months back, I was talking to my manager about risk based PM and it turned out that he had also been looking at this as well but from a somewhat different angle. Being something of a wiz at getting info out of databases he had found that out of the thousands of PM tests done on infusion pumps over the last few years only three “failures” had been found as a result of that PM testing (I will try to get the exact figures from him again and post when I have them). This means that from a patient safety point of view a huge percentage of the tests done were, in effect, a waste of time and effort. Now it could be argued that the 3 devices that were picked up did not go on to cause a patient incident but given the numbers of pumps that do fail “in-use” and the comparatively rare incidence of serious incident arising from those failures I do not think this is a particularly strong argument.

Finally (Thank God I hear you cry), does PM really reduce breakdowns? In some device types I would certainly agree that it does and in these cases I would agree with the work being done but across the board, I think not. If we take the rational approach allowed by a risk based system we can keep our safety standards a high as ever but remove or at least reduce the amount of pointless effort that we seem to spend so much time on.