Why do labs change reference ranges...? - Thyroid UK

Thyroid UK

112,361 members130,687 posts

Why do labs change reference ranges...?

PurpleCat71
PurpleCat71

I just compared lab results from this year and last year to previous years (going back to 2010 when I started using this particular lab).

I have noticed two changes in the past ten years:

the reference ranges for TSH have gone from 0.35-4.9 to 0.35-4.5 mUI/L...which I believe is good)

ref for FT4 unchanged (0.7-1.5 ng/mL)

the reference ranges for FT3 have gone from 1.7-3.7 pg/mL to 1.7-3.2 pg/mL...which is not so good for me as I feel best with FT3 levels around 3.5-3.6 which now would mean I am overmedicated whereas before they were in range. The best I´ve felt ever was with FT4 1.2, FT3 3.6 (then in range, now slightly out of range according to new ref ranges). That was on T4 + T3, dosing by symptom-relief only, but it´s pretty interesting that I had mid-range FT4 and top of range FT3 (suppressed TSH but I never cared about that)...something I would never, ever be able to achieve on levo only...and I´ve never felt better. So, a lab changing ref ranges, and a doctor obsessed with said ref ranges could mean huge potential problems.

Is there a reason why a lab would change its reference ranges for the TSH and FT3, but not the FT4...? Are there any international or European recommendations to lower narrow down FT3 reference ranges?

I know ref ranges vary from lab to lab, but does this mean it´s time to find a new lab as I find their current ref ranges for FT3 too narrow...?!

28 Replies
oldestnewest

It doesn't work like that.

The result is only correct in relation to the range of that test , done on that test machine,who's manufacturer sets a range for that machine , (which is then adjusted by the lab to account for local population)

So for example . if you took 2 identical samples and sent them to 2 different labs who use different machines one would come back as;

10 [5-15] and one would come back as

12 [6-18], both are 50% of way through range.

Annoying though isn't it. I've had about 6 different ranges for FT4 over the years. makes it really hard to figure out how to draw a graph.

Lab reference ranges which change this way are not the same thing as say an NHS policy change affecting treatment ranges.

Edit ; So most likely reason for a different Lab reference range from the same lab , is they bought a new machine.

Yes, annoying, and also confusing since it´s the same lab that has change the ref ranges for FT3 and TSH, but not FT4...I don´t care about the TSH, but the narrower FT3 ref range now means it´s easier to end up hyperthyroid (according to labs) than it used to be...

I too am dubious about narrower ranges. My FT4 range was changed fro [8-18] to [7.9-14] so much narrower, and i have had some problems with being much more % over range in the last few tests . However the first test done on this new range was more plausibly consistent with my previous history, so i cant say it's the range that's made the difference.

In theory , even though the range is narrower, your result should still come out at the same % through the range , with a narrower range your result would be a lower number , so still within range by the same %. .

In theory ! , but like i said , i am still a bit dubious.

The only way you'll know for sure is 2 identical samples , processed on the new range , and the old range , and see if the % is actually the same in practice.

I'll see if i can find the old post with an interesting chat about this very subject

healthunlocked.com/thyroidu...

see diogenes response to my question here.

Thanks!:-)

humanbean
humanbean in reply to tattybogle

Annoying though isn't it. I've had about 6 different ranges for FT4 over the years. makes it really hard to figure out how to draw a graph.

Work out the percentage through the range for each result using the range that came with each result.

Then graph the percentages rather than the raw data. It means you never have to worry about the range changing.

tattybogle
tattybogle in reply to humanbean

Thanks, humanbean and TSH110 . Yes that is what i did in the end . Mind you with FT4 ranges that have gone from [11-26] to [7.9-14] and everything in between over the yrs and my recent 240% result , my Graph now looks like the Cuillin Ridge!

(But the span of the new [7.9-14] range is only 6.1 , compared to the previous [8-18 ] span of 10......... so that skews things quite a lot, since a change of 1 pmol/L is a much bigger % of 6 than 10.)

Since the latest range change in 2017 there is also an inexplicable Mt Everest sitting at the end of the Cuillins;

14.6 [ 8 - 18] 66% .........Levo 125mcg

14.7 [7.9-14] 111%........

20.5 [7.9-14] 206%.......Levo then reduced to 112mcg

22.7 [7.9-14] 242%.........GP confused now !, me too. Levo then reduced to 100mcg.

19 [7.9-14] 181%.........

It would be easy to blame this these results on inconsistencies of timing of test/ last dose etc but i was very careful.... all exactly the same conditions. Hence my recent question;

healthunlocked.com/thyroidu...

TSH110
TSH110 in reply to tattybogle

It would seem that nobody knows! Do you think posting a shot of your intercontinental mountain range like graph might concentrate minds plus putting an @ diogenes ? He might be able to explain it.

tattybogle
tattybogle in reply to TSH110

Good idea , but am luddite without smartfone /photo device :)

could perhaps get passing child to do it for me sometime.

TSH110
TSH110 in reply to tattybogle

Could you do the axis as % through range, rather than numeric where the ranges are not consistent, for your graph?

humanbean
humanbean in reply to TSH110

It would be more helpful to visualise data by having the y-axis as % through the range and the x-axis would be time. So people see how the test result they are looking at varies over time.

Like looking at a cortisol graph for cortisol testing. The x-axis in that case is hours through the day and the actual level of cortisol is on the y-axis.

But for results like a TSH or a FT4, say, it could take people quite a long time to generate enough data to make a graph worth doing. It would really depend on how often people tested. The NHS tests as infrequently as they can get away with. So for most people they would need to be paying for testing themselves, and as we all know that costs a lot of money.

Edit : I misread the post from TSH110 when I wrote this. But I'll leave this post here anyway.

TSH110
TSH110 in reply to humanbean

That’s exactly the graph I had in mind!

humanbean
humanbean in reply to TSH110

Yes, sorry. I misread/misunderstood what you wrote!

TSH110
TSH110 in reply to humanbean

But you described it so clearly and exactly 😉 I now wonder what you thought I meant.....this could tip over into paranoia 😂🤣😂

humanbean
humanbean in reply to TSH110

Could you do the axis as % through range

I read this as

Could you do the x axis as % through range

when I thought it made more sense to put percentage through the range on the Y-axis, so I went into one of my waffles in response. :D

Hope that puts your mind at rest. ;)

TSH110
TSH110 in reply to humanbean

Well I totally agree with you, so how could that provoke a paranoid episode I don’t know....unless I was Stalin who said he trusted no one not even himself....for good reason!

humanbean
humanbean in reply to TSH110

Yes, I know we agree with each other. Which is why my reply (which started this whole string of posts) was superfluous. :)

tattybogle
tattybogle in reply to TSH110

I too was initially confused by your description , but that's because i'd forgotten that Axis is the correct word for my 'bit up the side' or 'bit along the bottom'

Using %'s also means i could add my precious FT3 results to same graph (all 5 of them ! in 17 yrs)This now makes a nice little 'stream' running along beneath the mountains.........too far beneath the mountains probably.

One day~ when i can find the tracing paper , i'll overlay the TSH on top. Which is all very interesting.... but still explains nothing unfortunately. I was hoping for Enlightenment ... oh well .at least i've got an impressive picture to wave at doctors.

tattybogle
tattybogle in reply to humanbean

Yes , you need lots of results over a long time.I have NHS results fro 2003 to present..... hence my pretty picture of the Cuillins next to Everest

TSH110
TSH110 in reply to tattybogle

⛰ 😂🤣😂 🏔

helvella
helvellaAdministrator

Many reasons.

They might have changed the analyser machine and technology used.

They might have re-calibrated and found that they needed to adjust the reference interval.

A number of years ago, many TSH ranges had the top end reduced because the original (or earlier) calibration exercises had not rigorously excluded all patients who had any form of existing thyroid disorder. (Some antibodies affect test results directly.)

If you had a result of 3.5/3.6 on an FT3 test some time ago, then they change the test, a re-test even with you having the same good level might have a different result. Without knowing much more we cannot know if that would be the case - or not.

PurpleCat71
PurpleCat71 in reply to helvella

Interesting...but it does not make it easier to stick to the same lab for consistency, does it?!

helvella
helvellaAdministrator in reply to PurpleCat71

Some lab tests are standardised - around the world. And that does help everyone.

Afraid that for those which are not standardised, we have to live with how they are. However difficult that is.

Instead of being concerned about different lab ranges it might be more useful to calculate how far through each given range your results sit.

chorobytarczycy.eu/kalkulator

Ideally FT4 and FT3 should be close to 75% through their respective ranges

Comparison is then straightforward whatever the lab range.

Thanks!

A bit OT but still related to this subject: I could never get my FT3 above 50% through range (and that was on 200 mcg of levo daily which most doctors consider overdosing), even with top of range FT4 levels and a lot of remaining hyper symptoms...which is why I started self-treating with T3. That of course made my TSH drop very low (it´s now fully suppressed at <0.01) and every doctor I see for something completely unrelated tells me I risk everything from osteoporosis to heart disease to premature death...they don´t even look at the free Ts:-( One doctor even told me that "my TSH showed that my thyroid gland was overproducing hormones" which would be medically impossible since I was diagnosed with Hashimoto´s over 20 years ago....

Most doctors are completely useless when it comes to thyroid disease, that is at least my experience. I have had to read a lot, almost to the point of feeling I´ve gone to med school, and then try to convince doctors to change my treatment...when they refused, I chose to self-medicate. I now have stopped telling doctors I´m on thyroid meds as they will inevitably want to order a TSH test and then they will start lecturing me...once, when a doctor ordered thyroid labs without telling me, I simply told the lab I did not want them done. Of course you have a right to refuse testing just as much as you have a right to refuse treatment. The doctor got very upset so I dropped her. I now try to avoid doctors as much as I possibly can.

Following from the comment by tattybogle about reference ranges being adjusted for local population, I found this article as below. (which is then adjusted by the lab to account for local population). I find this fascinating because if you live in an area where more people have thyroid issues then the reference ranges will take that into account. I asked my GP several years ago why the T4 reference had changed from 9-21 to 12-21 and he said because too many people were falling into low thyroid function and they were needing to treat them! ‘labtestsonline.org.uk/artic...

PurpleCat71
PurpleCat71 in reply to Aunds

It´s good when they things change in that direction...don´t like the narrower range for FT3 though as I need to be at the top of the former rather than the new range...as I imagine is the case for many people taking T3.

I do believe that changes in reference ranges have to do with the fact that average human hormone levels have been dropping for decades. In other words, they tend not to reflect the way things were 75 years ago, when we had not yet polluted ourselves (and most other animals) almost to a state of invalidism.

You may also like...