Comparing Results With Difference Reference Ranges

[greenbeanie]

I have a general question about how reference ranges are calculated for tTG-IgA. Is the difference just that some labs are more conservative than others in terms of what they consider normal? Or are test results from different labs not comparable at all because they use different processes or measurement techniques? Do different results from different labs correspond to actual differences in the amount of the antibodies in the body?

 

For example, suppose Lab A has a normal range of 0-3, with any values of 4 or higher considered positive. Lab B has a normal range of 0-19, with any values of 20 or higher considered positive. If a single tube of blood was tested and found to have a level of 8 at Lab A, would you expect the same blood tested at Lab B to also have a value of 8, or would you expect it to have a value of 40 (i.e., twice the lowest positive value at that lab)?  Is the only difference that Lab A thinks that a level of 8 is a cause for concern, whereas Lab B doesn't? 

 

At first I was just mildly curious about this, but the more I think about it, the more it really matters. A level of 60 at Lab A would be 15x normal, whereas a level of 60 at Lab B would only be 3x normal. This might make a big difference in how fast the level might be expected to come down on a gluten-free diet. It would also matter in terms of tracking levels over time if follow-ups were done at different labs. 

[brileighlevi]

Interesting question. I haven't even thought about that. I'd love to see what ideas everyone has about this. I'm very new to the celiac testing and have no real clue what is going on. It would make more sense if the labs did different tests or just read them at different levels but that seems unlikely. I wonder what difference it does make. My daughter had labs done and they came back negative but she's having more tests run at a different lab so I wonder if it'll make a difference. 

[RMJ]

I've been lurking here for a while, but finally registered because I believe I can answer this question by explaining how such tests are developed. You may notice that the results are usually in U/mL, meaning units per milliliter. There are no official standards for the tTG or DGP IgA or IgG antibodies, so each lab prepares their own and assigns their own arbitrary units. When a laboratory develops tests like these they will have a collection of blood samples from people known to have celiac and people who probably don't have celiac. They run the samples. Then they choose a cut off value so that most of the samples from people with celiac give results above that value, and most of the samples from people without celiac give a result below that value. If you took a sample that was right at the cutoff at one lab and tested it at another lab it would probably be near the 2nd lab's cutoff but not right on, since the cutoffs depend on the sample sets each lab used to establish them and the specificity of their test reagents

Total IgA is different. There is an established standard and it is reported in units of weight per volume, such as mg/dL.

To try to answer some of your specific questions:

No, one lab is not being more conservative than another.

Test result numbers are not comparable between labs, different numbers from different labs don't necessarily mean different amounts of antibody.

If a sample from your lab A gave a result of 8, I'd expect the same sample tested at lab B might give a result in the range of 30 to 50. I gave it a big range to allow for variability in setting the cutoff, and normal variability in results. If you test the same sample three times at the same lab it will probably give a slightly different result each time.

Within one lab, a result of X would mean the sample had approximately half the antibodies as a sample with a result of 2X, but this might not hold at the high and low values.

Hope this helps!

[greenbeanie]

Thank you so much for the exceptionally helpful and thorough explanation! That makes total sense. Welcome to the forum, and thank you for registering in order to post this!