Biomedical Engineering Reference
In-Depth Information
operational quantities such as a level that may be recognized as “detected”, one that
is expected to lead to detection, and one that has adequate measurement precision
for quantitative assessment.
A large effort, involving a number of agencies and many specialists, led to the
development and publication of
An American National Standard
—
Performance Cri-
teria for Radiobioassay
, approved in 1996 by the American National Standards Insti-
tute and published by the Health Physics Society as HPS N13.30-1996. The express
purposeof theStandard“...istoprovidecriteriaforqualityassurance,evaluation
of performance, and the accreditation of radiobioassay service laboratories. These
criteria include bias, precision, and determination of the MDA [
minimum detectable
amount
].”
Building on the original work of Currie and others, the N13.30 Standard presents
a protocol that defines a
decision level
and a
minimum detectable amount
(MDA) for
measurements of a radioactive analyte in a sample. These quantities play the same
roles as their counterparts in the last two sections. However, they differ from the
former in the way in which background is assessed. In our earlier treatment, the
decision level is applied to net counts above background, measured with a subject
under analysis. The background count is typically made with the subject replaced
by an
appropriate blank
—e.g., synthetic urine, radiometrically the same as the sub-
ject, but having no added radioactivity. In N13.30, an additional measurement is
made by the routine procedure, where the subject contains no analyte above that of
the appropriate blank. W
hen
systematic errors are negligible, the N13.30 decision
level replaces the term
√
2
n
b
in Eq. (11.72) above by the standard deviation
s
0
of the
net count
of a subject with no added analyte. It thus includes contributions of both
a subject with no added analyte under the routine procedure and the background
as measured with the appropriate blank. Instead of comparing the subject count
with the single background determination as before, the comparison is made with
the specified net count, having standard deviation
s
0
.
The MDA defined in N13.30 is similarly comparable to Eq. (11.82) with
√
2
n
b
replaced by
s
0
. However, an additional complication arises when dealing with very
low background rates. A well maintained alpha-particle counter might register a
single count on average over a long counting time with a blank having no activity.
Use of the formalism is based on having good estimates of the mean and standard
deviation of the background. A semi-empirical value of three counts is added to
k
β
s
0
in the formula for the MDA in order to render
β
≤
0.05 when the background
is very low.
The American National Standards Institute has played an important part in the
quality of radiation protection in the United States. Their work led to the Nuclear
Regulatory Commission's test program administered by the National Voluntary
Laboratory Accreditation Program (NVLAP) and to the Department of Energy's
Laboratory Accreditation Program (DOELAP). The Institute's activities extend into
other areas besides radiobioassay.
