Normal Ranges

[Gambit62]

These are my personal thoughts on the use of normal ranges and testing in general.

I would not be in favour of changing normal ranges - normal has a particular mathematical meaning and mucking around with that introduces a degree of uncertainty that I personally don't think would help.

What is really wrong with the system is a) a very poor understanding of statistics and how to apply them to interpreting results from the real world and b) a resultant tendency to think that the figures tell the whole story and are all that are needed. This is an issue for many conditions - not just B12 deficiency and PA as a particular cause.

Levels of clinical evaluation - looking at symptoms - have seriously declined and there is a growing tendency to forget about symptoms and their importance. GPs should be treating patients, not test results. Changing reference ranges would do nothing to address this problem and could result in patients receiving treatment that was unnecessary (and in some cases could be potentially dangerous - there was a recent case of a patient who was treated for low iron on the basis of ferritin results alone and went on to develop iron toxicity as a result).

The role of labs in testing isn't just running the tests and passing on the results, but also includes providing guidance on interpreting the output from tests. Labs could and should do much more to help GPs with the interpretation of results. Some labs seem to have started doing this by including specific ranges to indicate areas where results may or may not be indicative of a problem but there are no standards for doing this and the provision of useful explanation of ranges is patchy to say the least. Improving this would be much more likely to lead to GPs realising the limitations of the results they are looking at, and making them more aware that they need to look at the patient in front of them. It won't necessarily resolve problems with being totally unaware of the range of symptoms that can be caused by the condition, or a tendency to assume that if they can be explained by a pre-existing condition then there won't be anything else going on (eg B12 deficiency in patients being treated with metformin). However, it would be a starting point.

At the same time there needs to be much more recognition of the fact that GPs are generalists and it is unreasonable to expect that they would know everything about everything. That isn't just a statement about patients relating to patients. It is also a statement about GPs relating to and working with patients. GPs need the confidence to be able to own up to the fact that they don't necessarily know everything - unlikely to be a skill that they get any training in at all. GPs and patients need to recognise that their relationship isn't a parent-child relationship. It is a partnership. Partnerships are built on trust and one of the key inputs to building trust is honesty.

[wedgewood]

Dear Gambit , a really thoughtful and helpful message . Thank you !

[Snowflake16 in reply to wedgewood]

So did I. Very comprehensive and sensible thoughts. Thanks

[helvella]

I'd point out that in some tests there are many references to normal ranges - even when the distribution is definitely not normal (or Gaussian). One close to my heart is TSH (Thyroid Stimulating Hormone). Fully accepting your reluctance to call for wholesale change but that issue has not been satisfactorily dealt with.

We also have the issue that it is conventional to place the upper- and lower-bounds at 2.5%. Which rather falls apart when the population might contain 10% or more who should at least be looked at further.

[Gambit62 Administrator in reply to helvella]

totally agree which is why labs should be doing more to help GPs understand how to interpret and use the results

[Cankita]

Well I would agree about not messing with the normal ranges - if there was such thing as a 'normal' range.

In May 2015 I had B12 and Folate tests done locally (in the Western Isles) and the reference range from the lab was 145 - 637 for B12 and 4.6 - 18.7 for Folate. My B12 was 143.5 and was so marginally close to the lower range that 'it didn't warrant treatment'. Folate wasn't tested due to degradation of the sample. A year later I had the same tests in Glasgow and the reference range there was 200 - 900 for B12 and 3.1 - 20 for Folate. My B12 was 195 and although only 'just' below the lower range treatment with injections was agreed. I now feel a lot better. Folate came back at 1.7. Of course if the original range for B12 had been used my 195 level would have been 'completely normal' and no amount of argument would have got me any treatment.

I agree with all your other observations but what hope is there when 'normal' is not normal at all?

[Gambit62 Administrator in reply to Cankita]

There are two different sets of units that can be used for the serum B12 assay so you need to give not only the range but the units. The difference in the ranges you quote looks like the difference between using pmol/L and using ng/L(or pg/mL).

Ranges can also vary with assay methods where more than one assay method is being used.

[Cankita]

Okay, the 145 - 637 range is pg/mL and the 200 - 900 range is ng/L. Aren't they the same?

[Gambit62 Administrator in reply to Cankita]

the two ranges are equivalent ranges but using different units of measurement. - one is measuring using the weight of B12 in the sample and the other is measuring using the number of molecules of B12 in the sample.