Substance Abuse and Mental Health Services Administration
Center for Substance Abuse Prevention
Division of Workplace Programs
Subject: An Evaluation of Non-Instrumented Drug Test Devices
Date: January 29, 1999
Background
The Department of Health and Human Services (HHS) published in the Federal
Register on April 11, 1988, the Mandatory Guidelines for Federal
Workplace Drug Testing Programs. The Mandatory Guidelines detail comprehensive
standards for laboratory procedures, specified drugs for which Federal
employees can be tested, and established appropriate standards and procedures
for periodic review of laboratories and criteria for certification of
laboratories engaged in urine drug testing for Federal agencies. The Mandatory
Guidelines were revised on June 9, 1994, but the basic requirement for
laboratory testing was not changed. In addition to covering the testing of
Federal employees in the Executive Branch of the Federal government, the
Department of Transportation requires its regulated industries to use HHS
certified laboratories.
During the past few years, there has been a rapid expansion in the private
sector to test for the presence of illicit drugs in other biological specimens
(such as, sweat, oral fluid, and hair) as well as the use of on-site urine drug
test devices. Approximately 2 years ago, the Division of Workplace Programs
(DWP) began a review of the testing of alternative specimens and the use of
on-site test devices. As part of the review, DWP funded Duo Research,
Incorporated, to conduct a study to evaluate the performance of on-site drug
test devices currently being marketed. This study was actually a second study
of 15 new or modified on-site test devices that had been originally evaluated
for the Administrative Office of the U.S. Courts in 1996. As in the first
study, this study focused on the testing of specimens clustered above and below
the cutoffs, specimens that were clearly negative or positive, and known
quality control samples.
This study is not intended to make recommendations for specific products, but to
provide a general assessment of the performance of currently available
non-instrumented drug test devices.
Device Evaluation
All known non-instrumented drug test manufacturers and device distributors were
contacted to obtain devices. Of these, 15 devices were provided by the
manufacturers or distributors. A Behring Diagnostics’ ETS instrument using Emit
d.a.u. reagents served as a reference device. Other available instrumented
systems were not included in the study.
The study was designed to test each device with 90 selected clinical specimens
and 10 control samples for each drug. The specimens were selected from routine
specimens submitted by Federal Probation Offices to PharmChem Laboratories
under its contract with the Administrative Office of the U.S. Courts. They had
been tested on an Hitachi 747 analyzer using Diagnostic Reagents, Inc (DRI)
enzyme immunoassay test kits. The immunoassay rate data were compared to spiked
control values to classify specimens into four categories: negative, below
cutoff (ranging from about 25% below cutoff to cutoff), above cutoff (ranging
from cutoff to about 25% above the cutoff), and high (greater than 25% above
the cutoff). Approximately 60 specimens had responses in the below and above
cutoff categories. Specimens testing negative were selected and stored frozen
until needed. Positive specimens were selected from previously confirmed
positives that were ready for disposal. For phencyclidine, a sufficient number
of unique clinical specimens in the desired range was not available, so
dilutions (with negative urine) of positive clinical specimens were performed.
The accuracy of these dilutions was checked on the ETS prior to inclusion in
the study and by gas chromatography/mass spectrometry (GC/MS) following the
study.
Each day, 20 to 30 specimens for a given drug were thawed and tested on all the
study devices. Specimens were identified to the operators only by bar coding.
All devices in this study indicate a negative result by the appearance of a
colored line at the area designated for each drug. Positive results are
indicated by the absence of a line.
The performance of the devices was assessed in terms of their "Positive
Predictive Values" (PPV) and "Negative Predictive Values" (NPV) and percentages
of false positive and false negative results. These are standard analytical
measures of the certainty of obtaining a correct positive and negative result,
respectively. Thus, a high percentage or high PPV indicates that there is a
high certainty that a positive result from the device will be confirmed as
positive, or as negative for a high NPV.
This differs from the percentage of true positives, called sensitivity, and
percentage of true negatives, called specificity. These measures indicate the
percentage of all specimens confirmed positive by GC/MS that were identified as
positive by the device, or of those confirmed as negative by GC/MS that were
negative by the device. Also, specimens confirmed as positive but identified as
negative by the device are false negative results, and for specimens confirmed
as negative but identified as positive by the device are false positive
results. The combination of all correct negative and positive results
represents an estimate of the overall accuracy for both positive and negative
results for each device.
The definitions used for this study are given below. Tables 1 through 5 show the
PPV, NPV, and accuracy data for each test device versus GC/MS at the HHS
cutoffs and with all borderline readings scored as negative (many devices give
equivocal results, which were scored initially as "borderline") for each drug.
Table 6 gives the combined results for all drugs for each test device. Figures
1 through 5 are the bar-graphs showing the sensitivity, false negatives, false
positives, and specificity as percentages for each drug class for each test
device ("PD" is used to represent the performance of a "Perfect Device").
Figure 6 gives the combined results for all drugs for each test device. It must
be noted that the results for opiates were based on the 300 ng/mL cutoffs. None
of the specimens included in the study would have been positive at 2,000 ng/mL.
Note: Codes have been used in the tables and figures contained in this report to
conceal the identities of the test devices. However, each manufacturer has
received a report that identifies the results for its test device.
The evaluation of the 15 devices was conducted with a majority of the specimens
grouped around the screening cutoffs. It was found, as expected, that many
devices gave a fair number of false positive and false negative results. It is
also expected, based on experiences in the field, that specimens encountered in
most Workplace testing situations will have fewer specimens with drug
concentrations near the cutoff. This means that a much higher percentage of
confirmed positive results and fewer false negative results should occur during
actual testing in the field.
The favorable performance of the devices was encouraging considering the
simplicity of their design and operational requirements. Some devices were able
to identify more positive specimens, but this was accompanied by a higher
percentage of false positive results. Other devices were more conservative,
giving few false positive results but missing many true positives.
Test Devices Evaluated
Dipro 10 Panel
Each device is packaged in a sealed pouch. The device consists of a flat plastic
card that has five dipsticks extending from one edge. Two of these are
connected back-to-back providing for ten separate test strips. The strips are
covered with a protective cap. Storage does not require refrigeration, but 36°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations
and areas for donor ID and date. The device includes a "control" or validity
check. The test requires the removal of the cap, insertion of the dipsticks
into the specimen for about 10 seconds. The cap can be replaced. The endpoint
for a positive results if the absence of a line at the test band windows.
Manufactured by American Biomedical, Inc.
| Distributor: |
Dipro Diagnostics
3415 Hycliffe Avenue
Louisville, KY 40207
Phone: 502-899-3108 |
Drug Check Cup
Each device is packaged in a sealed bag. Expiration dates are stamped on the
bags. The devices are well labeled, with clear result designations. The device
includes a "control" or validity check. The device is a self-contained
collection and storage cup, that has the drug test strips imbedded in its side.
The cup is closed with a screw-cap lid after specimen collection. The test
begins as soon as the urine is added to the cup. There is no other activation
step required. Results can be read within 5 to 9 minutes. The endpoint for a
positive result is the absence of a line at the test band.
| Distributor: |
Job Services, Inc.
32107 West Lindero Canyon Road
Westlake Village, CA 91361
Phone: 818-599-2512 |
Dtx 520
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not require refrigeration, but 36°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations.
The device includes a "control" or validity check. The test requires the
addition of 4 to 5 drops of urine to a well at one end of the device. Results
can be read within 3 to 8 minutes, but should be read within 8 minutes. The
endpoint for a positive results is the absence of a line at the test band.
Manufactured by Forefront Diagnostics, Inc.
| Distributor: |
Universal Drug Testing Company
467 Route 51
Large, PA 15025
Phone: 888-822-7120 |
Genie Cup
Each device is packaged in a sealed bag. Expiration dates are stamped on the
bags. The devices are well labeled, with clear result designations. The device
includes a "control" or validity check. The device is a self-contained
collection and storage cup, that has the drug test strips imbedded in its side.
The cup is closed with a screw-cap lid after specimen collection. At the time
of testing, the lid is turned to its fully closed position. This depresses a
plunger to trap a portion of the specimen in the test chamber. Results can be
read as soon as a line appears in the "test valid" area. The endpoint for a
positive result is the absence of a line at the test band. Timing is said not
to be important. Manufactured by American Biomedical, Inc.
| Distributor: |
Point of Care Technologies
6 Taft Court, Suite 150
Rockville, MD 20850
Phone: 888-713-8700 |
InstaCheck
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not require refrigeration, but 36°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations.
The device includes a "control" or validity check. The test requires the
addition of 4 to 5 drops of urine to a well at one end of the device. Results
can be read within 3 to 8 minutes, but should be read within 8 minutes. The
endpoint for a positive result is the absence of a line at the test band.
Manufactured by Forefront Diagnostics, Inc.
| Distributor: |
Forefront Diagnostics, Inc.
23561 Ridge Route Drive, Suite D
Laguna Hills, CA 92653
Phone: 949-595-0673 |
PharmScreen Drug Screen Card
Each device is packaged in a sealed pouch. The device consists of a flat card
that has five dipsticks extending from one edge. The strips are covered with a
protective cap. Storage does not require refrigeration, but 36°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear designations and areas
for donor ID and date. The device includes a "control" or validity check. The
test requires the removal of the cap, insertion of the dipsticks into the
specimen for about 10 seconds. The cap can be replaced. The endpoint for a
positive result is the absence of a line at the test band windows. Manufactured
by American Biomedical, Inc.
| Distributor: |
PharmChem Laboratories, Inc.
1505A O’Brien Drive
Menlo Park, CA 94025
Phone: 800-446-5177 |
PharmScreen Drug Screen Multi
Each cassette-style device is packaged in a sealed pouch. Storage does not
require refrigeration, but 36° to 86°F
is recommended. Expiration dates are stamped on the pouches. The devices are
well labeled, with clear result designations and areas for donor ID and date.
The device includes a "control" or validity check. The test requires the
addition of 4 to 5 drops of urine to a well at one end of the device. Results
can be read within 3 to 8 minutes, but should be read within 8 minutes. The
endpoint for a positive result is the absence of a line at the test band
windows. Manufactured by American Biomedical, Inc.
| Distributor: |
PharmChem Laboratories, Inc.
1505A O’Brien Drive
Menlo Park, CA 94025
Phone: 800-446-5177 |
Rapid Drug Screen
The device is a cup and a separate card containing the individual test strips.
It is packaged in a sealed pouch. Expiration dates are stamped on the pouches.
The devices are well labeled, with clear result designations and areas for
donor ID and date. The device includes a "control" or validity check. The test
requires the insertion of the test card through a slit in the lid and into the
specimen. Results can be read within 3 minutes, but should be read within 8
minutes. The endpoint for a positive result is the absence of a line at the
test band windows. Manufactured by American Bio Medica Corporation.
| Distributor: |
Integrated Corporate Solutions, Inc.
3121 Sunnybrook Road
Mogadore, OH 44260
Phone: 330-677-2441 |
Status DS-5
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not require refrigeration, but 35°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations.
The device includes a "control" or validity check. The test requires the
addition of 3 drops of urine to a well at one end of the device. Results can be
read within 3 to 5 minutes, but should be read within 10 minutes. The endpoint
for a positive result is the absence of a line at the test band. Manufactured
by Princeton BioMeditech.
| Distributor: |
Orion Diagnostica, Inc.
71 Veronica Avenue
Somerset, NJ 08873
Phone: 800-526-2125 |
Syva Rapid Test
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not require refrigeration, but 35°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations.
The device includes a "control" or validity check. The test requires the
addition of 3 drops of urine to a well at one end of the device. Results can be
read within 3 to 5 minutes, but should be read within 10 minutes. The endpoint
for a positive result is the absence of a line at the test band. Manufactured
by Princeton BioMeditech.
| Distributor: |
Dade Behring, Inc.
3403 Yerba Buena Road
San Jose, CA 95135
Phone: 800-729-7982 |
TesTcup 5
Each device is packaged in a sealed bag. Storage does not require refrigeration,
but 65° to 85°F is
recommended. Expiration dates are stamped on the bags. The devices are well
labeled, with clear result designations. The device includes a "control" or
validity check. The device is a self-contained collection and storage cup, that
has the drug test strips imbedded in its side. The cup is closed with a
screw-cap lid after specimen collection. At the time of testing, the lid is
turned to its "test"position, the cup tilted until the urine covers 1/2 to 3/4
of the lid (do not invert fully). It is held in this position for 10 seconds,
then returned to its upright position. Results can be read as soon as the "TEST
VALID" window develops a blue color, usually within 5 minutes. The endpoint for
a positive result is the absence of a line at the test band. Timing is said not
to be important. An adhesive covering strip is removed from the test display
windows and placed on a small "breather" hole on the back of the cup. The lid
should then be returned to the sealed position if the specimen is to be stored
or sent for confirmation. Manufactured by Roche Diagnostic Systems, Inc.
| Distributor: |
Roche Diagnostic Systems, Inc.
1080 U.S. Highway 202
Somerville, NJ 08876-3771
Phone: 800-526-1247 |
Accutest
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not required refrigeration, but 36°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled, with clear result designations.
The device includes a "control" or validity check. The test requires the
addition of urine from a pipette marked for about 0.2 mL to a well at one end
of the device. Results can be read within 3 to 8 minutes, but should be read
within 8 minutes. Manufactured by Jant Pharmacal Corporation.
| Distributor: |
Jant Pharmacal Corporation
16255 Ventura Boulevard, Suite 505
Encino, CA 91436
Phone: 818-986-8530 |
One Step
Each device is packaged in a sealed pouch with an attached pipette. Storage does
not require refrigeration, but can be if desired, or a room temperature, 65°
to 85°F. Expiration dates are stamped on the
pouches. The devices are well labeled. The device includes a "control" or
validity check. The test is in the form of a dipstick, requiring the dipping of
the strip into the urine, preferably a small aliquot in a test tube. Results
can be read within 3 minutes, but must be read within 5 minutes. The endpoint
for a positive result is the absence of a line at the test band. Manufactured
by Technical Chemicals & Products, Inc.
| Distributor: |
Technical Chemicals & Products, Inc.
P.O. Box 9748
Ft. Lauderdale, FL 33310
Phone: 954-979-0400 |
QuickScreen
Each cassette-style device is packaged in a sealed pouch with an attached
pipette. Storage does not required refrigeration, but 40°
to 86°F is recommended. Expiration dates are stamped
on the pouches. The devices are well labeled. The device includes a "control"
or validity check. The test requires the slow addition of 4 drops of urine to a
well at one end of the device. Results can be read within 10 minutes. The
endpoint for a positive result is the absence of a line at the test band.
Manufactured by PhamaTech.
| Distributor: |
PhamaTech
9265 Activities Road
San Diego, CA 92126
Phone: 619-635-5840 |
TesTstik
Each device is packaged in a sealed pouch with an attached pipette. Storage does
not require refrigeration, but 36° to 86°F
is recommended. Expiration dates are stamped on the pouches. The devices are
well labeled. The device includes a "control" or validity check. The test
requires the dipping of the device into a urine samples up to a mark on the
device for 5 to 7 seconds. Results can be read within 5 minutes, but should not
be read after 30 minutes. The endpoint for a positive results is the absence of
a line at the test band. Manufactured by Roche Diagnostic Systems, Inc.
| Distributor: |
Roche Diagnostic Systems, Inc.
1080 U.S. Highway 202
Somerville, NJ 08876-3771
Phone: 800-526-1247 |
Definitions
True Positive (TP)
A positive result by the test device and a positive result by GC/MS (the
reference method).
False Negative (FN)
A negative result by the test device and a positive result by GC/MS.
False Positive (FP)
A positive result by the test device and a negative result by GC/MS. This
includes both "unconfirmed" positives, i.e., samples with drugs present below
the cutoff and samples with no drugs detected.
True Negative (TN)
A sample negative by the test device and negative by GC/MS.
Sensitivity
The number of True Positive results for a test device out of all GC/MS positives
in the study expressed as a percentage (TP x 100/(TP + FN)).
Specificity
The number of True Negative results for a test device out of all GC/MS negatives
in the study expressed as a percentage (TN x 100/(TN +FP).
Prevalence
The percentage of positive specimens in a given population of specimens.
Positive Predictive Value (PPV)
The probability that a positive result for a test device will be a True Positive
in a population with a known or estimated prevalence of positive specimens
(that is, a value calculated for a device as to its ability to produce correct
positive results, which is dependent upon the prevalence of positive
specimens).
PPV = [Sensitivity x Prevalence] / [(Sensitivity x Prevalence) + ((1 -
Prevalence) x (1 -Specificity)]
Negative Predictive Value (NPV)
The probability that a negative result for a test device will be a True Negative
in a population with a known or estimated prevalence of positive specimens
(that is, a value calculated for a device as to its ability to produce correct
negative results which is dependent upon the prevalence of positive specimens).
NPV = [Specificity x (1 - Prevalence)] / [((1 - Sensitivity) x Prevalence) +
(Specificity x (1 - Prevalence))]
Table 1. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
Amphetamines
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
Ja
|
0.818
|
0.873
|
0.867
|
|
Ga
|
0.786
|
0.895
|
0.878
|
|
Da
|
0.714
|
0.942
|
0.889
|
|
La
|
0.556
|
0.827
|
0.800
|
|
Ea
|
0.545
|
0.896
|
0.809
|
|
Aa
|
0.500
|
0.981
|
0.789
|
|
Ha
|
0.476
|
0.870
|
0.778
|
|
Nm
|
0.467
|
0.840
|
0.778
|
|
Ca
|
0.432
|
0.943
|
0.733
|
|
Nm
|
0.400
|
0.843
|
0.744
|
|
Km
|
0.360
|
0.846
|
0.711
|
|
Bam
|
0.327
|
0.947
|
0.589
|
|
Pm
|
0.313
|
0.811
|
0.722
|
|
Lm
|
0.281
|
0.828
|
0.633
|
|
Im
|
0.278
|
0.833
|
0.611
|
|
Fam
|
0.267
|
0.841
|
0.551
|
|
Dm
|
0.258
|
0.814
|
0.622
|
|
Cm
|
0.213
|
0.793
|
0.400
|
|
Qm
|
0.091
|
0.772
|
0.689
|
Note: "a" is amphetamine specific, "m" is methamphetamine specific, "am" is
sensitive to both amphetamines
Table 2. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
Cannabinoids
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
P
|
1.000
|
0.474
|
0.556
|
|
H
|
0.935
|
0.576
|
0.700
|
|
B
|
0.933
|
0.567
|
0.689
|
|
E
|
0.848
|
0.544
|
0.656
|
|
M
|
0.833
|
0.556
|
0.667
|
|
D
|
0.810
|
0.583
|
0.689
|
|
N
|
0.784
|
0.641
|
0.722
|
|
J
|
0.774
|
0.649
|
0.722
|
|
I
|
0.757
|
0.509
|
0.611
|
|
G
|
0.721
|
0.773
|
0.733
|
|
Q
|
0.705
|
0.500
|
0.600
|
|
L
|
0.697
|
0.929
|
0.733
|
|
K
|
0.697
|
0.929
|
0.733
|
|
C
|
0.693
|
0.867
|
0.722
|
|
F
|
0.667
|
0.619
|
0.656
|
|
A
|
0.628
|
0.426
|
0.522
|
Table 3. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
Cocaine
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
C
|
1.000
|
0.667
|
0.856
|
|
H
|
1.000
|
0.591
|
0.800
|
|
B
|
1.000
|
0.578
|
0.789
|
|
M
|
0.966
|
0.774
|
0.900
|
|
A
|
0.963
|
0.667
|
0.844
|
|
I
|
0.953
|
0.511
|
0.722
|
|
Q
|
0.953
|
0.511
|
0.722
|
|
P
|
0.933
|
0.733
|
0.867
|
|
G
|
0.933
|
0.511
|
0.722
|
|
K
|
0.932
|
0.710
|
0.856
|
|
J
|
0.932
|
0.710
|
0.856
|
|
E
|
0.925
|
0.913
|
0.922
|
|
N
|
0.906
|
0.769
|
0.867
|
|
D
|
0.892
|
0.760
|
0.856
|
|
L
|
0.886
|
0.900
|
0.889
|
|
F
|
0.813
|
0.800
|
0.811
|
Table 4. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
Opiates
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
B
|
0.481
|
0.968
|
0.820
|
|
J
|
0.464
|
0.968
|
0.811
|
|
E
|
0.341
|
0.980
|
0.689
|
|
I
|
0.325
|
0.960
|
0.678
|
|
Q
|
0.319
|
1.000
|
0.644
|
|
A
|
0.311
|
0.978
|
0.644
|
|
G
|
0.308
|
0.941
|
0.667
|
|
P
|
0.304
|
0.977
|
0.633
|
|
H
|
0.286
|
0.976
|
0.600
|
|
L
|
0.246
|
1.000
|
0.489
|
|
C
|
0.234
|
1.000
|
0.456
|
|
M
|
0.231
|
1.000
|
0.444
|
|
D
|
0.227
|
1.000
|
0.433
|
|
N
|
0.221
|
1.000
|
0.411
|
|
F
|
0.220
|
0.935
|
0.467
|
|
K
|
0.208
|
1.000
|
0.367
|
Table 5. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
Phencyclidine
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
B
|
0.844
|
0.931
|
0.900
|
|
G
|
0.611
|
0.833
|
0.744
|
|
N
|
0.543
|
0.782
|
0.689
|
|
H
|
0.509
|
0.914
|
0.667
|
|
C
|
0.500
|
0.912
|
0.656
|
|
A
|
0.492
|
0.935
|
0.644
|
|
J
|
0.475
|
0.760
|
0.633
|
|
E
|
0.460
|
0.926
|
0.600
|
|
Q
|
0.437
|
1.000
|
0.556
|
|
F
|
0.429
|
0.794
|
0.567
|
|
K
|
0.424
|
0.875
|
0.544
|
|
L
|
0.423
|
0.947
|
0.533
|
|
D
|
0.419
|
1.000
|
0.522
|
|
I
|
0.413
|
1.000
|
0.511
|
|
P
|
0.411
|
0.941
|
0.511
|
|
M
|
N/A
|
N/A
|
N/A
|
Table 6. PPV, NPV, and Accuracy Values for Test Results versus GC/MS using HHS
Cutoffs (All borderline results as negative)
All Drugs
| Device |
PPV |
NPV |
Accuracy |
|
Perfect Device
|
1.000
|
1.000
|
1.000
|
|
J
|
0.717
|
0.822
|
0.778
|
|
B
|
0.699
|
0.798
|
0.757
|
|
G
|
0.673
|
0.810
|
0.749
|
|
E
|
0.642
|
0.830
|
0.735
|
|
H
|
0.629
|
0.774
|
0.709
|
|
M
|
0.611
|
0.767
|
0.689
|
|
N
|
0.597
|
0.797
|
0.693
|
|
A
|
0.591
|
0.798
|
0.689
|
|
P
|
0.569
|
0.734
|
0.658
|
|
Q
|
0.551
|
0.726
|
0.642
|
|
L
|
0.545
|
0.873
|
0.680
|
|
D
|
0.538
|
0.830
|
0.669
|
|
K
|
0.537
|
0.849
|
0.642
|
|
I
|
0.532
|
0.726
|
0.627
|
|
F
|
0.513
|
0.814
|
0.610
|
|
C
|
0.509
|
0.862
|
0.637
|
|
