EBME Forums Forums Management Medical Equipment Management Risk-Based PM

Afwan. All part of the service!

And looking forward to some more posts on this interesting topic!

I'm preparing to present to the Environment of Care committee at WRMC, my suggestions of implementing a reliabiltiy-centered-maintenance strategy focused on select type of devices. We currently rely on a risk-based program derived from the original F&S model. Approval to conduct RCM and FMECA analysis will have to get by my VP, Risk Managment, and other non-technical upper management. Not having any luck finding other local institutions that have tried the RCM approach. There are several organizations here in the US that are actively recommending the clinical engineering community to engage the RCM in healthcare topic. Of course there are other types of tactics being discussed, evidence-based-maitnenance, gradient risk sampling. Its been easy to find articles and suggestions, not so easy to find any CE depts that have implemented. My instinct is that RCM would best be applied to high-volume, mobile assets e.g. IV pumps, vital-signs monitors, electric beds, stretchers, wheelchairs.

Hi Rick,

earlier thread I was searching a cheap/fee Risk Based PM.

https://www.ebme.co.uk/forums/ubbthreads.php/topics/41835/1

Lots of acronyms and abbreviations, there, Rick!

Here's another one:- HPM!*

What's that? Holistic PM. Isn't that what Risk-based PM really is (or perhaps that should be, er ... should be)? That is, a consolidation of all the earlier "PM's", taking the best (and most appropriate) aspects of each? Including (and again I stress the point) the ability to adjust PM intervals according to sensible and appropriate criteria.

I don't think that the "massive upheaval" model is the way to do the thing, especially (as you have indicated) that most likely means having to convince the dull and the ignorant before you can even take the first (hesitant?) steps. When you think about it, you can implement the tenets of the Risk-Based Model quietly, and in a time frame of your own choosing (or even, not at all). Simply overlay it on whichever system you are currently using.

I would suggest this approach. Implement it on an "as and when" basis. At first (at least) no special software is required. Just start off by thinking deeply about Risk Criteria, then apply some to a few types of equipment. Document what you are doing as you progress, note any changes you decide to make, and ... just take things from there. That is, add, build, develop, improve ... step by step, recognising from the onset that benefits will only accrue (if at all) over a number of PM cycles (meaning, of course, over a number of years). This is no "quick fix" ... and that's why it is unlikely to appeal overly much to the suits and bean-counters.

Lastly, by all means "stand on the shoulders of giants" (F&S, and all the others), but I would urge you also to sit down and objectively appraise the situation where you are. And then come up with your own Risk Factors (and Risk Scores). In my experience, your guess (and mine) is just as likely to as valid as anyone else's. You can always make adjustments later on, in light of continued experience. The criteria need not (should not) be set in stone! As long as you ground the thing in sound engineering, and common sense ... whilst disregarding any pressure to "save money" (because that's not what it's all about), you will find it to be a most worthwhile (dare I suggest enjoyable) exercise, no doubt.

* HPM (c)2009 me!

Appreciate the responses to my ?? regarding RCM strategy related to medical equipment. Judging by responses from the US based BMETS-ONLINE listerve it does seem to me that there is not a widely accepted alternative to numeric risk calculations and interval based tactics that I could propose to my EOC committee at this time. Due to citation from our accrediting authority
last October, Clin. Eng dept has been under the microscope so to speak and we can't implement any changes to risk classification or PM interval without the express approval of EOC, Risk Management, legal and compliance officers. Since Novermber 2008, we have been able to adequately demonstrate our ability to maintain a 100% completion rate on "life-support" and 90%(barely) on "non-life support" devices, but I'm not sure when dealing with shrinking resources this can be maintained indefinitely. I shall continue to monitor biomed listerve sites along with the ASHE maintenance practices task force web-site (http://www.ashe-mptf.org)and see if the "best practices" maintenance recommendations idea sharing proceeds.
Geoff, I think we shall go ahead as you suggest, do our internal data failure analysis, share with other service groups, and hold off the presentation to our Environment of Care (Safety)committee until a more receptive atmosphere exists.

Yeah, (to coin a phrase) "just do it"! You are the local expert, not "them". And by the time you are able to demonstrate results, "they" will have long gone*, no doubt about it!

Just select a few types of equipment for a pilot scheme, and take it from there. Infusion pumps might be a good place to start, I would have thought.

Thanks for that link. It's an interesting resource.

* Only to be replaced with more of the same, unfortunately.

I came across this resource (by accident, as usual)! It looks like there's a wealth of good stuff there.

Earlier I had been listening to an interview on the radio about "Predictive Policing", and that set me thinking about predictive this, and predictive that (medicine, maintenance etc.).

Proactive? Yes, that's how it has to be. I'm not yet sure about Predictive Maintenance (in our context, that is), but believe I could be a champion for Predictive Service per se.

By that I mean my usual mantra of "getting out of the workshop* and onto the wards" with the aim of nipping problems in the bud, anticipating problems down the line, spreading the gospel of good equipment husbandry, listening to users, and other good stuff like that! This is nothing new, of course, but something that many of us have been doing for years.

I seem to recall hearing about a strategy once called "Management by Walking About"!

* Despite the view that Dave in Oz seems to taken of yours truly!

Useful Site Geoff loads of stuff there...nice discovery

As mentioned earlier one of the key elements of the Risk-Based PM scheme is an ability (an option) to automatically increase or decrease PM intervals according to the condition of equipment as encountered at PM visits. To my mind it is always the actual condition of equipment that really matters (that may sound obvious, trite even, but I believe it is sometimes lost sight of). So the ability to adjust intervals based upon condition makes sense to me!

Don't forget that we have already established our initial intervals and priorities for PM based upon risk levels.

With automatic interval adjustment, as the PM scheme continues to take effect (that is, as time marches on), PM intervals will eventually break away from those that were initially set by the risk level.

Here is an example (using typical criteria). PM intervals may be shortened or lengthened based on two basic conditions:-

Firstly the condition must be met that the three previous PM intervals must be approximately equal (that is, for any further question to be based on a "fair" sampling).

Secondly, if a repair was needed at all of the previous PM's then an interval reduction is triggered. In an effort to prevent potential failures, it looks like we need to carry out PM more often. So let's reduce the interval by 50% and see how we get on with the next three visits.

However, if no repair was needed at all of the previous PM's then an interval extension is triggered. OK, let's increase the interval by 50%.

If one or two repairs were carried out, we leave the interval as it is. It has to be a "straight flush", as it were, before any change is triggered.

So it can be seen that we need to be careful to record whether a repair was needed at each PM we carry out. This is a crucial point. Naturally, we may also need to be clear in our minds what we actually mean by "repair" in this context.

Now we need to look a bit more closely at the criteria involved.

In the example above we used:-

1) An interrogation of the previous three PM's. So there's our first variable. How many PM's to look back through (to see if repairs were carried out). We could call this (key) variable the "Recent PM's repair check".

2) Three PM intervals of "equal" length were required. We need to allow a bit of slack here as to what we mean by "equal". Plus or minus ten per cent sounds reasonable (eg, for a nominal 180 period, we could allow between 162 and 198 days). Anyway, let's call this variable the "Equal interval bracketing" (see how easily these trip off the tongue)!

3) We used an interval adjustment of 50%. So, for an interval starting off at 180 days (for example) we would increase to 270. Or decrease to 90 days. If the pattern of repairs (or none) was repeated following the next three PM cycle, these would become 405 and 45 days respectively. Obviously if we had actually encountered a run of six repairs needed at PM in a row we would (hopefully) have done something else rather than simply continue to attempt PM! This one can be known as the "Interval adjustment" (fair enough).

Here are some suggested figures to play around with:-

1) Recent PM's repair check:- min 1, max 9, default 3
2) Equal interval bracketing:- 5 to 50%, default 10%
3) Interval adjustment:- 10 to 90%, default 50%

... all arguable, of course (just like the risk levels)!

But (as I say) I would regard the "Recent PM's repair check" as the key setting. If you wanted to see quick results, this would need to be set low. With it set at 1, for example, the system would very quickly break away from the initial intervals, as it would look back at only the last PM to see whether a repair was needed. It is debatable whether a single interrogation would give sufficient justification to trigger an interval change ... but I'll leave that for you all to ponder.