New Analyzer Verification of Accuracy and Spanning the AMR
24 followers
0 Likes
I have read various sources and asked CHATGPT and I am still unsure of the answer.
I have spoken to my lab director about this as well but there are conflicting opinion.
When doing a validation/verification for a new device it is preferable in an ideal world that you test patient samples that span the AMR of the device.
We do not live in an ideal world where we can get patients that span that range for the various analytes for example an analyzer like the GEM or the epoc.
Does running linearity material by itself and comparing it to the manufacturer range satisfy that we verified accuracy and reportable range if we combined it with the 25 patient sample's accuracy.
How do people normally span the range for analytes like pH, pO2, pCO2, k, cl, ica, na, ?
My lab director will not sign the validation document if it doesn't span the ranges with patient blood but that is unreasonable and unpractical.
What is a reasonable approach to this question?
I have spoken to my lab director about this as well but there are conflicting opinion.
When doing a validation/verification for a new device it is preferable in an ideal world that you test patient samples that span the AMR of the device.
We do not live in an ideal world where we can get patients that span that range for the various analytes for example an analyzer like the GEM or the epoc.
Does running linearity material by itself and comparing it to the manufacturer range satisfy that we verified accuracy and reportable range if we combined it with the 25 patient sample's accuracy.
How do people normally span the range for analytes like pH, pO2, pCO2, k, cl, ica, na, ?
My lab director will not sign the validation document if it doesn't span the ranges with patient blood but that is unreasonable and unpractical.
What is a reasonable approach to this question?
10 Replies
Hi Jo! We just received a proposed validation plan from Siemens and they included ways to manipulate the samples to obtain higher or lower ranges of certain analytes. I recommend reaching out to the vendors for their recommendations.
@Jenna, I have something for epoc but the document doesn't actually talk about all the analytes that you can expand, only some. It also included this quote "These are target distributions and will rarely be completely achieved, but efforts should be made to obtain as wide a clinically relevant range as possible" There is no real clear hard line.
That's unfortunate. We are preparing to make a switch to EPOC and will have the same questions soon. Hopefully someone else has some knowledge. I think ultimately it depends what your Medical Director is comfortable approving.
In trying to meet AMR, sometimes we need to get creative. I have used calibration verification materials, PT or Cross Check supplies orderable after submission dates. Those items will give you targeted acceptable values. If you have a device that you are comparing it to, I have used manipulated human samples. For gases, that might mean open to room air or diluting out. Looking forward to seeing the creativity of others!
@Erika, thank you for your input. My lab director is very particular in the sense that she says only blood samples/same matrix for patient testing can be used for an accuracy that spans the reportable range. your thoughts?
For blood gas and chemistries, I usually spike the blood with cal ver material, then run it on both devices. Eurotrol makes a hypoxic and hyperbaric control, to get high blood gas results. We use it on istat and Epoc. For clotting time, we pool patient samples and either add calcium chloride or heparin depending on what value we need. For our glucose device, I add dextrose to elevate the glucose level, you can also leave the sample sitting overnight or a couple of hours and that will change the value.
@Shaneece , I was thinking of doing the same but was concerned with altering the matrix. What is a good ratio of cal ver material to blood do you use? I'm actually doing that experiment today to figure out if it is feasible. Thank you Shaneece!
Good question lol. I don't have a specific ratio, I just eyeball it really. Once I know the original result, I add 5-6 drops of cal ver at a time and test. It can get tedious and messy. I pipette the blood into a small test tube and add the cal ver dropwise until I get the desired result.
We use Cal Ver, linearity, etc to span the reportable range. We do try to get patients to also challenge those, but that is not always feasible. If we can't prove the lower and high ends through any of these materials, then we change our reportable range. We just recently had to do this on AVOX, ISTAT and GEM for a couple of analytes. Periodically, I go through data and make sure we have the points to prove those low and high limits. Honestly, some of the ranges that the manufacturer's supply are not clinically relevant, so tightening ours to what we can actually prove has not been an issue.
An update. I spiked blood with cal ver material; 3 parts blood with 1 part cal ver and was able to get consistent results across the GEM and vitros chemistry analyzer. The matrix still looked like blood. There were no errors provided by both analyzers. I will try 2:1 next but it works to get a wider range.
Reply
Subgroup Membership is required to post Replies
Join POCT Listserv now
Suggested Posts
| Topic | Replies | Likes | Views | Participants | Last Reply |
|---|---|---|---|---|---|
| Breath Hydrogen/Methane Testing | 2 | 0 | 103 | ||
| Blood Gas/Chem POC devices | 1 | 0 | 444 | ||
| Epic-Beaker Venous Draw Requests for Unsolicited Orders | 3 | 0 | 377 |