EBME Forums Forums Free discussion Free discussion [poll]What do YOU use to check the accuracy of infusion devices?

We use an IDA4, also a biurette (?? can't spell) and high accuracy weighing scales. DEspite certain infusion pump manufactureres saying the IDA4 isn't any good, we can test their pumps on all three of the above methods and get almost identical results.

Which pumps do you use? We use(d) infutest on Graseby 9000 series and Baxter Colleague - I wasn't convinced with the results.

KITT
My understanding is that the IDA from Ultramedics works via looking into a fine burette with a series of leds at fixed intervals up the burette.

As the water creeps up the burette the beam is refracted and picked up by a sensor opposite and the device increments its volume count. I don't know how it purges the burette between measurement cycles.

The main difference between the old IDA 2 Plus and the newer IDA 4 is that there are more leds closer spaced along the burette giving a higher resolution and more accurate performance.

The system is prone to slight optical deterioration from contaminants in the water.
My understanding is that the annual service of these devices includes the fitting of a new burette to keep the system clean.

Your design look superficially a bit like the Ultramedic system. Instead of using optical interference from refraction though you are using optical interference from an opaque solution.

The unit from Metron has a pump chamber within it and in response to the fluid input pressure allows a measuring chamber to fill, then purges it and restarts in the manner of a very tiny Imed 900 series piston pump, or perhaps in more contemporary terms like a Graseby 500 boxer mechanism. The units accuracy is principally therefore determined by the measuring chamber accuracy and its resolution is determined by the measuring chamber volume. The unit makes a rather alarming racket from its pump mechanism but is really quite excellent in use.

Your ideas are excellent but you need really to know how pump tester manufacturers test their pump testers. After all if you're going to develop a test method you will need to be able to test the methods you develop and compare their performance.

My guess is that they use a burette! If the volume measurement is accurate and the system clock is accurate then they can infer that the incremental volume measurements will in practice be accurate.

As regards currently used test methods no one has yet mentioned potentiometric measurement of syringe pump travel. Ultramedic market that as well. A large potentiometric device is fitted into the syringe drive and converts linear travel to resistance and this can then be converted and interfaced to a PC

Our trust has 3 service bays. One uses burette, one uses an IDA 2 Plus and one uses a digital scale. Each naturally prefers their own technique (and for very good reasons).

For something a little bit different why not think about using turbine rotation as a sensing method. I don't think anyones used that yet (probably for good reason too).

Marc

I dont think a turbine system would work as you have problems like dynamic friction, etc and maintainance would be much more than in a system without moving parts. Maybe a dynamic weiging system would work?

Hi Colleagues,

Infusion pump measurement is a very interesting discussion.

At North Middlesex Hospital we've used the two manual methods to test our Arcomedical Volumed uVP5005 pumps i.e the measuring cylinder (burrett? - I can't spell either!) and also very accurate weighing scales (e-Accura SC131), always using distilled water as the pumping medium to measure the total volume delivered. In addition, the total time of delivery is measured using an electronic bench stopwatch so that the rate of delivery can be checked. The method is to connect the stopwatch to the "nurse call" socket on the pump. Start the pump and the stopwatch together and when the pump finishes it's delivery it automatically stops the stopwatch. Hence, the total time that it takes to deliver the set volume can be read (without sitting watching the pump with a manual stopwatch!) Therefore, the average rate can be checked as total volume/time (ml/h etc)
Just reading the volume in the measuring cylinder isn't very accurate at all. Our 100 ml measuring cylinders have only 1 ml graduations and are only accurate to +/- 0.5 ml @ 20 deg C. Moreover, it is very difficult to read the level accurately from the cylinder by eye. Using the cylinder with the accuate scales is much better. Moreover you can "zero out" (tare) the weight of the cylinder when it's empty and hence get a direct reading from the scales of the weight of the volume of dist. H2O delivered.(This method is clearly a cheaper option than purchase of an IDA device.)
What I'm not clear about is, do we need to measure the instaneous rate as opposed to the average rate delivered as you would with an IDA-2 or IDA-4 device? The use of of this type of testing device is obviously much better, especially if your testing lots of pumps.
However, it seems to me that with all testing/checking we should only carry tests that the manufacture recommends to ensure that the equipment is working within its specification. Volumed do not specify that to test their pumps you must use an IDA device. In fact their engineer has only one rate check on his test sheet i.e. 100ml/h with VTBI = 50 ml. We check this with +/- 1% accuracy (49.50 - 50.50 ml with an expected period of 29.40 - 30.20 min). Possibly, too accurate a measurement?
Finally, it's obviously important to always use a new or hardly used giving set when testing pumps.
Having said all of the above I'm sure that our unit would benefit from having a new IDA device given the number of infusion pumps that we test. (We've got a very old IDA-2 sitting around gathering dust. I think I've seen it used more than a couple of times about 10 years ago as it always had problems!)
Ian Rez

I think for practical purposes, i.e. outside the physics lab, it's generally accepted in engineering circles that the uncertainty in a measurement should, ideally, be about 1/10th of the stated accuracy in whatever parameter you're measuring. Then you can avoid, i.e. not have to get involved in, discussions about whether accuracy is an issue

Bravo Ian,

Do we need another thread to discuss occlusion pressure testing – or should it come under this one also?

The “North Mid model” is a sound one, based as it is on physical principles. However, as described it does not “test across the whole range”, as it were.

But what I like about the IDA’s, Infutest’s or what-have-you is that you can obtain a quick check as to whether a pump is generally working OK (or not). You can quickly test it over a range of flow-rates, and also check the occlusion pressure. And you get the all-important printout for your equipment records files!

Incidentally, in my experience, most IDA problems have been due to impurities in the fluid (water) being used (ie, not being flushed out from the sensing mechanism). Follow the manual and all is usually well!

I thought the context of the thread was aimed at a "novice" who wanted some background advice. All this is old hat and a bit tedius for me as well but if people are not prepared to occasionally state the obvious to "newbies" then there's no point in having the forum.

Personally speaking my trivial contributions are just to highlight what should be well-known issues RE: uncertainties in measurements and the problems that have been experienced with accuracy, in the past, with IDA-like analysers - particularly when used to measure pulsatile-delivery. Things that are not necessarily obvious to "schoolies" looking for a project perhaps.

In general, syringe pumps are used to deliver drugs and IV pumps are used to deliver fluid volume.
I think you do need to have a reading of short term accuracy for syringe pumps. not just whether it can deliver the syringe full in a given time. If you are giving a drug with a short half life such as inotropes, fluctuations in the flow rate will cause physiological changes. These drugs have half lives measured in a few minutes so you need to be able to measure for consistancy of flow rate over similar timescales.
I have seen a patients BP fluctuating because their inotrope was given with a pump with a bent lead screw. But on average the pump was accurate for the whole syringe.

If you are giving plain fluid through and IV pump it does not really matter if a patient gets 90mls one hour and 110mls the next if it is set to 100mls/hr.

Should we be adopting two flow measuing systems depending on the pump and its clinical use?

For those with the luxury of an IDA4 with hydrograph or a similar setup I would suggest you have that facility.

As to the initial question as to what is used, here are the many answers. As to what is required here are many suggestions.

What is needed is an easy to use system that gives the information required in an easy to analyse format with little operator input. You need to be able to set the system up to run for several hours then return and read the results quickly and easily. We do not have time to baby sit it, timimg and counting.
If you can fulfil our needs with a system that does it all well done and thank you.
Robert

We test all infusion device on an IDA-4 Plus, with the hydrograph software. The two reason for this.

1: It gives you nice written report of how the infusion device has run, plus a nice software archive of all the test you have done.
2: We also use a mains timer, which cuts the power to the infusion device in the middle of the night for three hours. Good check of the battery life.

If the pump fails any of these tests you can fall back to the manufactory recommended instruction. Be it glass burettes or dial test indicator. The other thing we do is after all the test are complete is to do a data download from the pump, so any one can see what test were run and what alarms have shown.

Good luck with the project.

A.M