How Insurance Works

Quality Assurance -1

it is an German Dear Colleague first I’d
like to thank say organizer is meeting to give me the opportunity to talk about
Quality Assurance and external QC and the terrific question is need for
immunization the general background of this topic will be that the result of
the same tests between different clinical laboratory should be equivalent
but as you know in the survey you just don’t guidelines are based on test
result from specific laboratory but without considering the possibility of
difference between procedures and the absence of recognition of such results
are not aluminized may lead to erroneous clinical decision and the only way to
try to see the problem is to use external quality controls that allow the
state-of-the-art for the test comparisons and external quality in
medical laboratories should also include evaluation of method performance as we
see this morning training education and help as we also know that external
assessment of a laboratory identical performance that the proficiency testing
or external quality assessment is performance evaluation according to
predefined criteria this include qualitative and quantitative tests and
this should be and must be a continuous process for different cycle for example
ETA is military in France not only because we are in in charge of creation
labs but also since many years where this this is military in France and as
you know it care organiser could be medical laboratories of the of the type
of laboratory including government association professional organization
regional organization and so and it’s so important to know that it’s important
clue for aggradation bodies and in the field of a gradation
we all know that the norm izo one five one eight nine should say that the
laboratory shall participating in telugu rhetoric
comparison such as those organized by external
quality assessment shame and if not possible as stated in the other norm the
is in telemetry comparison is not available the laboratory my shoes way to
do that and to define the criteria of acceptability of procedure not as a word
evaluated in France QG standard recommendation since 1978
all labs male participate to a QA organised by originally by a French
government organization which is called as saps
but participation of each ISA veganized by regional agency or national or
regional agencies are highly recommended and with the back rotation all all
essays realized within a lab must have a correspondent in EQE what are those goals were actually first
to demonstrate the mastery of the global quality of result reveals the defect
elite laboratory implement corrective measures improved and directly of the
quality control there is an interaction between external QC and internal QC make
consistent the first step probably in telemetry comparison and results and
also wait to see QA is to eliminate all and poor method and promote new essays
but it’s a long way to go to these goals and the primary intention of hue in
elaborate medicine shall be to support in fact quality improvement of the
service provider participating laboratories for the benefit of patients
this is a typical shame for external Quality Survey this is organized by
analyzation body which could be as I said previously a government or a
regional accrediting body and send you a sample mix of samples to a participating
laboratory which back the results which must be evaluated of kokand ins to
expected result report expected target values
distribution of results evolution of acceptability and summarize all these
data furthermore the lab received the reports
implement corrective measures and continue to process for the entire
program cycle so the question of my topic is what could be allies among all
this data first step is immunization of measurement that is also called the test
it needs to precisely define the measurement for most of the test we
talked about this morning and this evening it’s rather easy for cholesterol
triglycerides etc but in France and in all our labs we have many many different
tests including immunologic test which deserves this assumption so it is easy
to test with a single reference procedure reference measurement but it’s
visible for single analysis not for global test such as the meta logic test
we’ll see we’ll see soon it’s difficult or rather impossible for intelligence
molecules complex molecular from etc for example to take the example of an
cardiac markers which has been naturally peptide we only knows from recently
which forms are circulating so how can we had pure BNP without knowing exactly
what is the circulating form of these molecules we must also start that
improvement and integral specificity of the measurement by defining for EMA
logic tests crucial or consensus epitopes this is a case for example is a
troponin where there is a global AFCC and international consensus about a
particular sequence within the troponin which must be recognized by all the
tests including non non official tests or written on reference test so we are
improving specificity by developing new si on the
basis of the new definition of measurement by epitope specificity we
must further improve compatibility of reference material not only using if
possible suitable material will which is soon but in order to quantify
or more exactly to minimize effect of the manipulation of sample matrix the
concept of Priscilla comment ability is accepted for little numbers of tests we
see that if the material is a comfortable reach there is a numeric
relationship between the two procedure the reference one and the test one and
in order to have a selection of QC samples behaving like patient samples
but most of the case there is no harmonization and reference material
behave differently from clinical sample then this is the case of BNP for example
as I say previously so one topic to be a menage the organization a target value
of he Kiki because the programmer neither must provide the target values
for committable samples it rather simple in principle since mean or median value
of reference method by using certified reference material could be used but
it’s all it’s also limited to a strict number of tests and not all the tests
within a lab could be it could be commutable and the cutter the target
values or it depends on the purity of added material and once more the problem
is what added to my matrix effect in order to get to give different
concentration from a QC it’s also test the accuracy of device and the mean and
median could be to be shootin like target values for not commentable sample
the generally solution is to use mean or median values of the peer group of
inventories a group of methods for all laboratories but in this way you suppose
that we meet the same metrics bias is supposed and consequently the same
result are expressed within the lab or between the labs and the main number is
the main limit of this test as you say as usual as you see is a number of
participant because we have many technology for a same test and some time
is difficult to have pre-requisite numbers of statical
statistical pertinent number of providers what could it be I managed to
organize the same we could try to Amin eyes acceptance criteria for aq
interpretation since the program organiser may provide a limit value for
a range of acceptable value this value could be a damage for each test in term
of precision considering the state of the art for example 20 to 50 percent of
laboratory result such such a way is a Shuzhen by American or German
recommendation but Yuki seem limits of interpretation may be in fact or
regulatory but they are generally wider to identify to performing labs most of
and statistical but they rely on the assumption with that measurement
procedure is the acceptable but you know better
you cut your lab is not better than you another one but you don’t know which is
right our clinical but based on critical
difference that may affect clinical decision based on biological variation
but rather difficult to implement one way to solve the problem is to see the
imprecision esteem estimated by the TV are all biased estimate by the SDI’s
that sees for the CVR the ratio of the cv lab versus the civic group and you
can choose the cv according to the peer or according to the method in this case
of again phosphatase the result are rather good also bias could be estimated
by the difference of the you mean of the lab a versus mean of the group divided
by the standard deviation of the group ideally for the two is zero for SDI and
one for the CVR and corrective action could be done if you not fulfill this
criteria next point is to see what could be Amana is more but if we
seize that organization of participant participation could be interesting this
because there are several potential problems identified whenever sticking an
acceptable result and the LS SI guidelines give you some clues to make a
failure to examination of your behavior of the sample within lab you can make
clean see the clinical clerical errors for example we talked about few minutes
ago miss level samples for methodological problems problems
regarding region calibration etc equipment problem problem for example or
technical problems due to personal error and finally a quick EQ material problems
for example non-conservative problem of conservative the problem of sample
deterioration and the sample is a problem of matrix effects in fact with
this failure tree you can monitor trends in result of a time and different and
detection of bias there are some limitation of a key a from the from the
user and also from the manufacturers from the user lab must choose a provider
independent for in vitro diagnostic manufacturers this is not always
possible we must remember that we realize blind test but in fact we all
know that a corrective control is different from a patient sample even
with with the presentation of the kind of sample we must have an independent
treatment and no help or no comparison with your neighbor or your other labs et
cannot replace internal kharabeesh an entire quality control due to majority
of low frequency of those tests and delay for interpretation and finally a
key result is not recommended to modify calibration and a lab must provide exact
that in order to have exact peer sorry exact peer values must provide exact
data regarding region calibrator and equipment and from the manufacturer
point of view and which is different from the IVD diagnostic
we must provide an EQ provider must provide exact that about the composition
stability of the samples and the calculation of a target value acceptable
and the limits also statical calculation and furthermore EQ must be close as
possible from clinical sample what is the evolution of a QA a creation
procedure will offer a better comparison and guarantee for quality management but
the core element of a QA are validation of Hickey and statistical evaluation if
we see some example coming from outside biochemistry non biological samples
could be promoted for example for the realm of the idea or electronic response
hectoring pictures imitating microscopic view it could be
antagonistic was the clues that we must very near from human samples but in fact
it could be a solution in term of diffusion of information for example finally my conclusion the use of
external quality evaluation lo the simplest way to compare result in with
between laboratories all testing especially in the field of a creation
what must be realized within laboratory and should have in the QA is a Hickey
program must be realized in same manner as a clinical sample and ionization of
external QC is a part of a general process which begins this morning with a
different talk where you see providers and clinical laboratories act to suppose
the progress in quality of the analytical process thank you for

Leave a Reply

Your email address will not be published. Required fields are marked *