Ranges

[Obsdian]

I see on the forum that it's always important to post ranges. I initially assumed there was one standard for all.

What makes the ranges different?

[TiggerMe Ambassador]

It varies due to the test equipment and how it is set up and calibrated this is why you'll see that we swap results into percentages which makes it easier to compare them

[Agitator23]

Good question - and what's really bad is that doctors don't use percentage through range when they're discussing present and past blood tests. I pulled up one doctor recently who said my T4 hadn't fallen that much quoting figures instead of percentages. However, when I pointed out the different ranges in the tests it showed it had fallen from 40% to 9%

The blood testing system is really a mess!

[FallingInReverse]

units

There’s also one other huge reason in many cases, which when I learned about it made so much sense…

When you look at your next set of blood tests - look at the labels themselves. They are different units… kind of like meters and feet can be converted. So the numbers look wildly different but are actually exactly the same!

So 1 meter = 3.2 feet = about 36 inches = exactly the same distance

I clipped the attached from an article - I could never remember the details but if you read it, you can see how sometimes it’s as simple as the difference between meters and feet.

population variation

The one thing I don’t understand is the variation that comes from the population group the numbers come from in the first place. We often use the shoe size analogy- how shoes range from let’s say 5 to 13 in adult US sizes. I think sometimes ranges can vary based on the group of people who were included in the population of “normal,” by age, not including sick people, etc. When and how often did that exercise happen?!

lab variance

Then on top of that, TiggerMe’s comment points to actual measurement variance at the test level. As different labs use different methods and calibration - points to when blood test results actually vary from lab to lab.

Although there is still some consistency because of because of molecular weights and also standards by country. So it’s not entirely wacky but it is complex.

Conversion

[tattybogle]

eg: the different ranges used from 2003 to 2020; ( all using Royal Lancaster Infirmary NHS).

TSH.

2003 [0.36-4.1].

(i assume they got a different testing machine here )

from 2005 [0.02- 6].

from 2015 [0.05- 6].

from 2016 [0.34-5.6].

(I assume they got another different testing machine here, probably the one made by Beckman Coulter/ or possibly Abbott )

from 2018 [0.57-3.6].

FT4.

2005 [13-26].

from 2007 [11.2-26].

from 2009 [11-24].

from 2011 [9.5-20].

from 2012 [8-18].

(i assume they got a different testing machine here, probably Beckmann Coulter)

from 2018 [7.9-14].

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

in case anyone wants to get nerdy : healthunlocked.com/thyroidu.... difference-in-reference-ranges-reading-matter-for-a-rainy-day-.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

and just to add to the complexity ... personally i suspect the lower and narrower [7.9-14] fT4 test range is not giving a good match for the [12-22] range ... because when converting my fT4 results to '% through range' , the [7.9-14} range is giving a much higher % .

(my suspicion is that this difference has been consistent .. but i don't have enough date to prove it yet).

eg:

15/1/24 11am (GP test /NHS Lancaster) fT4 15.7 [7.9-14] 128 %

same day ... 11.45am (monitor-my-health/ NHS Exeter) fT4 22.1 [12-22] 101 %

So i can be pretty sure i have 'loads' of T4 as both are definitely showing 'highish' ... but the difference in results is the difference between the GP saying 'satisfactory' .. or having a panic attack.

ok... one was blood draw at GP , one was fingerprick/ posted...and obviously one was 45mins later... so some difference in fT4 result is to be expected ... and yes, labs are allowed a very small % difference in results from same sample / same machine. But even taking these things into account i still think this difference is big enough to be significant and shouldn't be happening.

One day when i can be arsed to try and have an intelligent conversation with a GP , i'll try to get them to organise a proper 'like for like' comparison.. ie ....2 samples from same blood draw, test one on [7.9-14]machine ,test another at different lab who have [12-22] machine .

[helvella Administrator in reply to tattybogle]

Around 20 or so years ago, there was quite a lot of discussion about TSH reference intervals.

One prime issue was that labs had not rigorously excluded anyone with any known thyroid issues when establishing the RI. This was picked up and we saw a wave of TSH RIs changing - in the USA and the UK. But it took years.

Similarly, some tests are sensitive to macro-TSH and appear high in those who have any TSH antibodies. But some assays were immune to interference of that sort. Again, even recognition of an issue took years and I think some assays are still sensitive to interference by macro-TSH. That could see differences between labs for the same patient.

[humanbean in reply to tattybogle]

One thing that doesn't get mentioned very often is that reference ranges are sometimes altered to cut costs i.e. increase the "normal" range and fewer people are diagnosed as out of range and therefore fewer need treatment.

The opposite also happens - the range is reduced to sell more drugs.

I have some blood test results for cholesterol from the 1990s. I noticed that the results from earlier in the decade had a reference range for Total Cholesterol which had an upper level of 7.5. Later in the decade the top of the reference range had been reduced to 5. I vaguely remember a discussion about this, possibly on TV or in a newspaper, that pointed out that the change in reference range had meant that well over half of the UK population were now going to be told they had high cholesterol and must start taking statins if they didn't want to die of a heart attack or stroke. I also remember that early versions of the Patient Information Leaflets (PILs) for statins used to say that statins wouldn't prevent heart attacks or strokes, but that now seems to have been removed.

It's worth knowing that PILs have to tell the truth about the drugs they apply to. I'm not sure if they are allowed to mislead or lie by omission. It is always worth reading the PIL for any new drug you start taking. PILs for many drugs can be found on the emc website (it isn't a complete list of all drugs supplied in the UK).

emc = electronic medicines compendium

medicines.org.uk/emc/

[helvella Administrator in reply to humanbean]

The EMC/medicines.org database of documents is easy to use. But, as you say, is definitely not complete. Worse, there are numerous references to EMC which actually say that it contains details of ALL medicines.

The database that does have all PILs (and SPCs and more recent Public Assessment Reports (PARs)) is the MHRA database:

products.mhra.gov.uk/

But it is horrible to use for many reasons. And, while you might find what you were looking for, the link might change five minutes later. Every update to each document means a new link. You CAN get round that by using a search for the product, but that is very awkward and has its own set of problems!

So, try EMC first and only switch to the MHRA products database if you have to!

[Obsdian]

Wow, that makes it more clear but is also a bit nuts as well.