Hi

I thought I would step in and give some general information about oestradiol methods which emerge from the UK NEQAS service.  I am not going to comment specifically on the detail of the postings; in my view, all patients' specimens are 'unique' and all/most methods are capable of anomalous behaviour with the odd sample from time to time.

Oestradiol is a difficult steroid to measure - we all know that, and the immunoassay industry has struggled to cope with the severe demands of achieving adequate working range on the one hand and sufficient analytical sensitivity on the other.  Different manufacturers have taken different approaches to optimise these conflicting requirements according to their perception of the 'market' and the perceived expectations of users, and all builders of major analytical platforms have to make compromises due to the fact that they cannot optimise all the elements of the assay for each analyte individually.  Since everything you do to an immunoassay affects antibody specificity, and you can't properly standardise an assay which is not adequately specific, it is not surprising that there are huge differences in the results that emerge from oestradiol assays.  It is common for there to be a 10-fold range in results across all laboratories for a single UK NEQAS sample and a four to five fold range within a single method group at low concentrations (see annual review low level study in link below).

Without going into a full rant about standardisation and traceability (most of you have heard this before anyway), I will just say that there should be absolutely no problem at all with standardisation and calibration of oestradiol assays.  All the elements of a full reference measurement system (materials and methods) have been in existence for steroids for decades, and I (and Graham Groom before me) have invested huge amounts of time and money in doing regular recovery and GCMS exercises in the UK NEQAS service over 30 years to show where the problems lie.  Between-method comparability, especially at low levels, is worse now than it ever was.  The laboratory medicine community has allowed pressure to produce any acceptable 'number' to take precedence over analytical validity, trueness and comparability.  This is wrong - it will come back to bite you some time soon!  Trust me, I'm a scientist.

If you go to the steroid hormones page of the UK NEQAS website at http://www.ukneqas.org.uk/Directory/CC/steroid.htm you will find three new links just under the banner which take you to pdfs of the latest work that I have undertaken.  My regular exercises are always available freely to UK NEQAS participants online.  If you don't participate - you are getting a freebie and can send a donation to a charity of your choice for the privilege!

The GCMS exercise gives you a feel for those methods which are properly calibrated (ie have a slope approaching 1.0), good baseline security (insignificant intercept), and tight within-method comparability across the concentration range examined.  (Note that at the point this work was done, the Architect method was on the change from the old to the new method - hence there were some very positively biased results in the data set.)  

The recovery exercise reveals how manufacturers have 'adjusted' the calibration curves of assays so as to try to overcome the low concentration uncertainty problem and give broadly 'true' results in the mid range (many assays have positive bias to GCMS at low levels and negative bias at higher levels.

The draft of my 2005 Annual Review for oestradiol shows the state-of-the art for this analyte.  I would just point you towards the 'penalty box plots' (B score vs C score) for each method group which shows the scatter for users, and the method-related low level oestradiol statistical analysis at the end of the file.

John Kane and I are currently writing an article for the Annals about the oestradiol 'fitness for purpose' problem which should emerge during 2006, which will go into more detail about all of this and what we do about it.

Hope this helps - please email me IN THE NEW YEAR if you want any further info.

Festive stuff as well of course!

J





Dr Jonathan Middle
Deputy Director, UK NEQAS Birmingham
0121 414 7300, fax 0121 414 1179
-----------------------------------------------
Please use [log in to unmask] for PERSONAL work-related email
Please use [log in to unmask] for UK NEQAS service-related email
For work-unrelated personal email please ask for my private Gmail address
-----------------------------------------------
All opinions expressed in this email are mine alone and are not necessarily representative of the views of the UK NEQAS organisation, UK NEQAS Birmingham (Wolfson EQA Laboratory), University Hospital Birmingham NHS Foundation Trust or University of Birmingham.
------------------------------------------------
The content of this message may be confidential and legally priviledged.  If you receive it in error please delete it immediately from your system.  Thank you.
------------------------------------------------

------ACB discussion List Information--------
This is an open discussion list for the academic and clinical
community working in clinical biochemistry.
Please note, archived messages are public and can be viewed
via the internet. Views expressed are those of the individual and
they are responsible for all message content.

ACB Web Site
http://www.acb.org.uk
List Archives
http://www.jiscmail.ac.uk/lists/ACB-CLIN-CHEM-GEN.html
List Instructions (How to leave etc.)
http://www.jiscmail.ac.uk/