Mike,
I had previously thought from the "of course" response that you were disagreeing with the statement from Prof Ekins "but outliers could not generally be identified using duplicates, because it might not be evident which of the duplicates was the outlier."
I now understand you are actually repeating to check for and remove the data from the calculation of the reported result.
Thank you for clearing this up for me.
Kind regards
Chris
-----Original Message-----
From: Clinical biochemistry discussion list
[mailto:[log in to unmask]]On Behalf Of COLLINS MICHAEL
(RM1) Norfolk and Norwich University Hospital
Sent: 24 March 2010 11:03
To: [log in to unmask]
Subject: Re: Manually improving precision?
You identify the outlier by repeating the test. If there's not enough sample to repeat you report insufficient sample and ask for a repeat sample.
Mike Collins
BMS3
Biochemistry Automation
Norfolk & Norwich University Hospital
England
[log in to unmask]
http://www.nnuh.nhs.uk/
-----Original Message-----
From: Clinical biochemistry discussion list [mailto:[log in to unmask]] On Behalf Of SERGEANT Chris (RF4) BHR Hospitals
Sent: 24 March 2010 10:50
To: [log in to unmask]
Subject: Re: Manually improving precision?
Mike,
"Of course outliers can be identified using duplicates. If the duplicates do not agree closely you repeat the test."
Maybe I am a bear of little brain.
How do you identify the outlier at the time of initial duplication? If you do, would you not be able to report the non-outlier thus negating the requirement for a repeat?
-----Original Message-----
From: Clinical biochemistry discussion list
[mailto:[log in to unmask]]On Behalf Of COLLINS MICHAEL
(RM1) Norfolk and Norwich University Hospital
Sent: 24 March 2010 09:49
To: [log in to unmask]
Subject: Re: Manually improving precision?
Of course outliers can be identified using duplicates. If the duplicates do not agree closely you repeat the test.
Mike Collins
BMS3
Biochemistry Automation
Norfolk & Norwich University Hospital
England
[log in to unmask]
http://www.nnuh.nhs.uk/
-----Original Message-----
From: Clinical biochemistry discussion list [mailto:[log in to unmask]] On Behalf Of Roger Ekins
Sent: 23 March 2010 21:14
To: [log in to unmask]
Subject: Re: Manually improving precision?
Gordon. Andy, Johnathan, et al,
Duplicates certainly reduce the s.d. by a factor of sqrt 2, thereby
increasing intra-assay precision. Duplicates also enabled us to
identify a 'missed' tube, i.e. one in which a key reagent had been
omitted. (This was obviously in the days when everything was done by
hand, before automatic analysers became available.) But outliers
could not generally be identified using duplicates, because it might
not be evident which of the duplicates was the outlier. Also -
using duplicates - we could routinely calculate the intra-assay
precision/imprecision profile, which provided a form of internal
quality control.
But how, Gordon, used you to determine inter-assay imprecision? This
contains statistical errors resulting from intra-assay imprecision.
Of course one could calculate the inter-assay component, knowing the
intra-assay component, and I recall doing this. But whether one uses
singletons or duplicates in an assay will alter the overall
inter-assay precision.
On a different (semantic) point, I understand why purists prefer to
use the term 'imprecision' rather than ,'precision' in this context.
This seems to be because we generally represent precision/imprecision
by the s.d or c.v. of replicate measurements.
But the use of the term precision in this context has a long history;
statisticians concerned with bioassays would refer to the "index of
precision" rather than the "index of imprecision".
And I also wonder whether purists would refer to the "insensitivity"
of an assay because the detection limit is indicated the
'imprecision' with which we measure an analyte concentration of zero?
Roger
>Dear Andy and Jonathan
>Within-assay replication decreases the within-run component of
>variation by a factor related to the square root of the number of
>replicates. It does not affect the between-run and
>between-laboratory components of variation.
>It was widely used in bioassays and radioimmunoassay, but the
>primary use was to detect blunders and fliers; not to reduce
>imprecision.
>Best wishes
>Gordon Challand
>----- Original Message ----- From: "Jonathan Kay"
><[log in to unmask]>
>To: <[log in to unmask]>
>Sent: Tuesday, March 23, 2010 11:12 AM
>Subject: Re: Manually improving precision?
>
>
>Analysis in duplicate was common in immunoassay and some other
>methods until very recently. My feeling is that it was done more to
>catch blunders and flyers than to improve precision.
>
>Where are people currently using duplicate in clinical assays?
>
>Jonathan
>
>On 22 Mar 2010, at 15:47, Andy Minett wrote:
>
>>Dear All,
>>
>>I have recently been reading about the tightening of quality requirements in
>>proficiency testing for glycated haemoglobin in America
>>(http://james.westgard.com/the_westgard_rules/2010/02/gutcheck-
>>hba1c.html ). One of the issues raised here is that many platforms do not
>>deliver the necessary imprecision to meet current quality requirements, let
>>alone future improvements.
>>
>>In fear of making a rather banal suggestion, has anyone looked at "manually"
>>(elaboration below) reducing the imprecision of an analyser? In fact, I am so
>>sure this approach has been considered and rejected, the question I am really
>>asking is: what are the drawbacks of this approach, and can the mailbase
>>provide me with references that discuss it (my own searches have been
>>unfruitful).
>>
>>By "manually reducing imprecision" what I mean is taking multiple
>>simultaneous
>>readings and reporting the average of those as the test result. In
>>an unrelated
>>(i.e. not Biochemistry related) mathematical project, I needed to
>>find a way of
>>minimising the scatter of data from a Normal Random number generator with
>>fixed Mean and CV. Through mass data testing I calculated the reduction in
>>scatter (CV) to be closely estimated by:
>>
>>% reduction in CV = (1-1/(n^0.5)) * 100
>>
>>(where n = number of replicate measurements averaged)
>>
>>Using this in my project, I was able to reduce the CV of the number generator
>>by 80% by averaging 25 results from it.
>>
>>Imagining a hypothetical glycated haemoglobin analyser that has a stable CV
>>of 3% say, taking the average of 4 simultaneous analyses of each sample
>>would reduce the CV of the method by 50%; to 1.5%.
>>
>>What are the drawbacks to using this approach in blood analysis? Other than
>>the approach being n times as costly as single-result analysis.
>>
>>Best regards,
>>Andy Minett
>>
>>------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
>>Green Laboratories Work
>>http://www.laboratorymedicine.nhs.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/
>
>------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
>Green Laboratories Work
>http://www.laboratorymedicine.nhs.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/
>
>------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
>Green Laboratories Work
>http://www.laboratorymedicine.nhs.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/
--
Prof Roger Ekins, PhD DSc FRS
Windeyer Institute of Medical Science
University College London
London W1T 4JF
Phone +44 20 7679 9410
Fax +44 20 7679 9407
------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
Green Laboratories Work
http://www.laboratorymedicine.nhs.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/
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom
they are addressed.
If you have received this email in error please notify the originator of the message. This footer also confirms that this
email message has been scanned for the presence of computer viruses but this should not be relied upon as a guarantee that the contents are virus free.
Any views expressed in this message are those of the individual sender, except where the sender specifies and with authority,
states them to be the views of the Norfolk and Norwich University Hospitals NHS Foundation Trust. The information contained in this e-mail may be subject to public disclosure
under the Freedom of Information Act 2000. Unless the information is legally exempt from disclosure, the confidentiality of this e-mail and your reply cannot be guaranteed
------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
Green Laboratories Work
http://www.laboratorymedicine.nhs.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/
Barking, Havering & Redbridge University Hospitals NHS Trust - Most Innovative Trust in London 2009
This communication, together with any files or attachments transmitted with
it contains information which is confidential and may be subject to legal
privilege and is intended solely for the use by the person(s) or entity to
whom it is addressed. If you are not the intended recipient you must not
copy, distribute, publish or take any action in reliance on it. If you have
received this communication in error, please notify
[log in to unmask] and delete it from your computer systems.
Barking, Havering & Redbridge University Hospitals NHS Trust reserves the
right to monitor all communications for lawful purposes. Receipt of this email
does not imply consent to use or provide this email address, or any others
contained therein, to any third party for any purposes. The contents of this
email are protected under international copyright law.
To save energy and paper please print this email only if you really need to.
_____________________________________________________________________________
This email has been scanned for viruses using antivirus software by
Barking, Havering & Redbridge University Hospitals NHS Trust. Whilst every
care is taken to avoid the transmission of virus code; please ensure that you
have up to date virus detection software before opening any E-mail
messages or attachments
______________________________________________________________________________
------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
Green Laboratories Work
http://www.laboratorymedicine.nhs.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/
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom
they are addressed.
If you have received this email in error please notify the originator of the message. This footer also confirms that this
email message has been scanned for the presence of computer viruses but this should not be relied upon as a guarantee that the contents are virus free.
Any views expressed in this message are those of the individual sender, except where the sender specifies and with authority,
states them to be the views of the Norfolk and Norwich University Hospitals NHS Foundation Trust. The information contained in this e-mail may be subject to public disclosure
under the Freedom of Information Act 2000. Unless the information is legally exempt from disclosure, the confidentiality of this e-mail and your reply cannot be guaranteed
------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
Green Laboratories Work
http://www.laboratorymedicine.nhs.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/
Barking, Havering & Redbridge University Hospitals NHS Trust – Most Innovative Trust in London 2009
This communication, together with any files or attachments transmitted with
it contains information which is confidential and may be subject to legal
privilege and is intended solely for the use by the person(s) or entity to
whom it is addressed. If you are not the intended recipient you must not
copy, distribute, publish or take any action in reliance on it. If you have
received this communication in error, please notify
[log in to unmask] and delete it from your computer systems.
Barking, Havering & Redbridge University Hospitals NHS Trust reserves the
right to monitor all communications for lawful purposes. Receipt of this email
does not imply consent to use or provide this email address, or any others
contained therein, to any third party for any purposes. The contents of this
email are protected under international copyright law.
To save energy and paper please print this email only if you really need to.
_____________________________________________________________________________
This email has been scanned for viruses using antivirus software by
Barking, Havering & Redbridge University Hospitals NHS Trust. Whilst every
care is taken to avoid the transmission of virus code; please ensure that you
have up to date virus detection software before opening any E-mail
messages or attachments
______________________________________________________________________________
------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
Green Laboratories Work
http://www.laboratorymedicine.nhs.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/
|
