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

It is possible with a stopwatch and burettes/dial guage to determine short and medium term variations in flow rate (and I have detected bent lead screws using this technique) but it requires monitoring the delivery test. The advantages of using an IDA are, I think, that it will give you a printed record of the result, important in some situations, and you do not have to monitor the test, you can get on with something else.

Not exactly the basis for a new thread of research though - and not particularly interesting if its been done before a hundred times.

Pump Testing

Gentlemen, Are we getting a bit too pedantic here?
Surely with any infusion device, in particular syringe drivers, all we need to measure is the accuracy of delivery (volume) over a range of therapeutic time periods (1 to 24 hours) taking into account different sizes and types of syringes.
You could also add into the equation different viscosities of fluid, again representing the different drugs that are used, as well as time to reach specified delivery rate (trumpet curve).
The IDA is probably the most widely used device for measuring most of the above, but for more info on testing equipment talk to Teresa Dunn at the Bath Institute.

Unfortunately research is a bit pedantic I think that's the objective.

Personally, for bench work, I'm happy with a burette that measures volume very accurately over a period I can measure pretty accurately. E.g. uncertainty in measurement on a 50ml f.s. burette of 0.1ml resolution is 0.lml hence +/-0.2% accuracy (it all depends on the raio of total volume measured to resolution of course but at least this is within our control). Add this to the system accuracy of a syringe driver and the is +/-2.2% the acceptable tolerances for volumetric delivery for a syringe driver with a 2% volumetric accuracy specified.

Errors in volumetric measurement using a burette can be made negligible for benchwork really. A range of burettes of different volumes allows pedants to maintain volumetric accuracy for delivery of a range of volumes without thinking about accuracy when doing the work at the bench.

Do the volumetric measurement with an IDA with uncertainty in measurements of +/-2% accuracy add this to the +/-2% specified system volumetric accuracy of a syringe driver and the potential error in delivery that is indicated can be +/-4%.

For a random pump and your IDA you tell me where the biggest error lies if you get a reading of -4%. Is it in the IDA or syringe driver? Where the IDA wins is in providing an indication of rate-delivery dynamically.

Gravimetric is the best of both worlds in my opinion: probably the best absolute volumetric accuracy and the capability to provide dynamic measurement of rate-delivery is very useful to indicate the delivery profile of pumps i.e. delivery rates over specified observation windows.

Actually the system accuracy should take all this into account since the fluid, set, pump, etc should be specified with this for repeatability of testing. Pure water at a specified temperature with calibrated scales interfaced to a PC is what's used for type-testing I think - that says it all.

But do you want to know how accurate the pump is under controlled laboratory conditions or the accuracy when delivering different drugs with different types of syringes over different periods of time in the clinical situation?

I want to have enough knowledge to indentify the pros and cons of each measurement method and choose the appropriate one - whatever that is. Working on the bench I would prefer a zero-error measuring instrument, whatever the device, anyday, then I wouldn't have to consider uncertainties in the measurement or the methods used to make the measurements repeatable and meaningful for the next pedant that comes along.

If the error in the IDA were 20% and the system error of a pump were 2% would the measurement be of any value? What if it were 10%, 5%, 2% where's your subjective threshold for the accuracy of an instrument used to give you confidence that a syringe driver is delivering within specification?

My view is that,for volumetric measurements, 2% uncertainty is not good enough when I've a burette that makes measurement uncertainties negligible. However if I needed an indication of the consistency of delivery then the IDA would win hands-down. If I was type-testing then I'd probably be inclined to use published methods in standards i.e. gravimetric checks.

Is there any specific reason that the IDA is inaccurate for measurng the volume of liquid infused by perilstaltic devices?

I think devices like the IDA work by timing how long a small internal volume takes to fill i.e. the IDA integrates lots of small volumes to measure volumetric accuracy. During the period this small volume is filling I guess if the rate fluctuates on the pump then the time taken to fill the volume varies i.e. ml/s changes but the IDA can't detect it. If the IDA can't resolve this fluctuation then this is an error I think -it will probably be cumulative so may become significant over time and at different rates so I guess the error will fluctuate depending upon rate, the total volume delivered and how pulsatile the pump delivery is.

Anyone know how an IDA really works?

What do you think on this prototype of a new kind of IDA?


CLICK HERE TO VIEW
Admin edit- Image removed as it was too big


I think the pic is self explanatory, although im not sure if it would be financialy practical to produce.

Im not sure how accurate this would be either, but producing a simple model (not as accurate as the finished product and on a smaller scale) to use in my presentation would be practical enough.

I just want some comments on my idea. Is this idea feasable, or a load of rubbish?

I have posted a private message to you.