A Case Series of Discordant Laboratory Results with Statewide Rapid HIV
Testing in New Jersey
Eugene G Martin, PhD1, Gratian Salaru , MD1, Sindy M Paul, MD, MPH2, Rose Marie Martin, MPH2, Maureen Wolski,BA2, Linda Berezny, BSN2, Evan M Cadoff1, MD
1UMDNJ
NEW JERSEY DISCORDANTS - 2006
Abstract
554
PRELIM POS
8
DISCORDANTS
6
4
2
Jan-06
Feb-06
Mar-06
Apr-06
May-06
Feb
Mar
Apr
Jun-06
Jul-06
Aug-06
Figure 4
Sep-06
60
NEGATIVE
Negative
Negative
Discordant
Discordant
TOTAL TESTS 11/1/03 – 10/5/05
WB
WB Pos
Pos
HIV PRELIM POSITIVE
32,463
Specimen Type
50
554 (1.7 %)
WESTERN BLOT – Discordant
20 (3.6 % of Prelim Pos. –
0.06% Overall)
40
30
20
10
0
Jan
May
Jun
Jul
Aug
Sep
Figure 2
ORAL
Correlations with Disease States – NJ (2003-5)
FINGERSTICK
Venipuncture
YTD
Total
Figure 5
60
50
NJHIV Discordant Series
11/2003 --> 10/2005
40
QC Usage - Year One
Ctrl 1 First Day of the Week
Ctrl 3 Annonymous Testing
Figure 1
Ctrl 2 Temperature Out of Range
Ctrl 4 New Shipment
Ctrl 5 Invalid Result
Ctrl 6 QC Failure
Ctrl 7 Coordinator Assigned
Ctrl 8 - Other
METHODS:
– 4-6 weeks Independent confirmation of the original negative Western
blot,
• Demographic data were collected using the standard Centers for
Disease Control and Prevention counseling and testing form
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
10
False Pos
CONCLUSIONS:
•
•
•
•
Ac
ute
He
p. B
HC
VI
nf e
Rh
ct io
eum
n
at o
id F
D is
act
t an
or
tE
BV
I nfe
Ora
c tio
qui
n
ck
Sp
eci
f ici
ty
Qualitative HIV PCR
Ultrasenstive Quantitative RNA PCR
Hepatitis B surf. Ag, Ab, HBcAb
EBV, RF,
20
Jan-06 Feb-06 Mar-06 Apr-06 May-06 Jun-06
nt A
bs
– Repeat HIV by Oraquick
– Venipuncture. Collect SST, white top, & purple top tube. Centrifuge and
prepare ASAP (< 4 hrs).
– Ship to reference laboratory (ARUP)
30
0
val
e
• The NJ Rapid HIV Program is a large, centralized QA effort (Fig. 1)
• After an initial positive Oraquick result and negative
confirmationoccurs at any site NJ HIV technical support is contacted
(http://www.njhiv.org)
• At 4-6 WEEKS:
Repeat HIV1/2 EIA
HIV1 Western blot
HAV (Polyvalent)
HCV Ab
10
15,570
32,463
6/5/2006
•
•
•
•
12
0
Outcome
AIDS Coalition of Southern New Jersey
Atlantic City Health Department
Bergen County Health Department
Burlington County Health Department
Camden AHEC
Camden County Health Department
Check-Mate
East Orange Health Department
Eric B. Chandler Health Center
FamCare
Henry J. Austin Health Center
Hope House
Horizon Health Center
Hunterdon County Health Department
Hyacinth Foundation
Martin Luther King Outreach
Morristown Memorial Hospital
Newark Community Health Center
NJCRI
Figure 1
Ocean County Health Department
Paterson Health Department
Plainfield Community Health Center
Proceed
Robert Wood Johnson Medical School
Trinitas Hospital
20 DISCORDANT
Month to Month
Pale colors indicate pending sites
NA
T
• An evolving infection – HIV screen is positive prior to traditional EIA or
Western Blot
• Cross-reacting non-specific antibodies
• Over-reading by testing personnel
mobile
HA
VP
oly
What Causes Discordants?
UMDNJ-RWJMS/ NJ DHSS
AIDS PREVENTION GRANTEES
Primary
Satellite fixed
H IV
TWO TYPES OF DISCORDANTS
TYPE I
Positive Oraquick ®, NEGATIVE Western Blot
No bands present
Client is considered HIV negative and not likely to be in an HIV window.
TYPE II
Positive Oraquick ®, INDETERMINATE Western Blot
Some bands not meeting the criteria to be declared positive are present
Possibility the client is in the process of seroconverting.
14
tiv e
DEFINITION: A reactive OraQuick® rapid HIV test followed by a negative or
indeterminate Western blot (WB) or immunofluorescent assay (IFA) result.
Frequency of Discordants – NJ (2003-5)
Po
si
Background – Discordant HIV Result
Rapid Testing in New Jersey – NJ HIV
NJ 2004
RAPID
HIV TESTING
RAPID
HIV TESTING
Background: A statewide case series of patients with discordant rapid HIV results at
publicly funded counseling and testing sites is described.
Methods: Initial fingerstick testing by either OraQuick or OraQuick Advance was
confirmed with Vironostika HIV-1 (BioMerieux) enzyme immunoassay (EIA) and
HIV-1 Western blot (Biorad). Discordant results (Oraquick positive/Western blot
negative) were followed by repeat Oraquick screening at 4-6 weeks, confirmation
of the original
EIA and Western blot results, collection of additional serum for hepatitis A
(HAV), hepatitis B (HBV), hepatitis C (HCV), HIV by standard enzyme
immunoassay, Epstein-Barr virus (EBV), and Rheumatoid factor (RF); collection of
additional plasma for ultrasensitive, quantitative RNA determination of HIV.
Demographic data were collected using the standard CDC form.
Results: Rapid testing began at a single New Jersey site on 11/1/03. Through
10/5/05, 125 sites were conducting rapid testing with 32,463 tests completed.
Twenty (0.06%) patients were discordant with a preliminary positive Oraquick rapid
test and a negative Western blot. Two patients refused follow-up testing. Three
other patients were lost to follow-up. Nine of 10 patients tested were Oraquick
positive upon re-examination 4-6 weeks later, but continued to test HIV negative by
traditional EIA and Western blot. Other testing included: 12 of 12 tested were
negative by ultrasensitive RNA analysis. Eight of 12 (67%) were HAV polyvalent
antibody positive, 1 of 12 (8%) had acute HBV, 2 of 9 (22%) had HCV infection, 1
of 12 (8%) had RF, 12 of 12 (100%) had evidence of a distant EBV infection. No
instances of an evolving HIV infection were identified.
Conclusion: Rapid testing is a reliable, reproducible screening test in publicly
funded counseling and testing sites. The 0.06% observed false positive rate in sites
with a rigorous QA program is acceptable.
– Robert Wood Johnson Medical School and 2New Jersey Department of Health and Senior Services
•
Figure 3
•
True Pos
Jul-06 Aug-06 Sep-06
No Followup
YTD
Total
Figure 6
Between 2003-5, 20 discordant rapid HIV tests were identified (Fig. 2).
A discordant occurs when a + Rapid HIV test fails to confirm by
confirmatory Western blot.
While evidence of distant EBV and HAV infection exist in more than half
of individuals with discordant HIV results, the frequency is similar to that
in the US population (Fig. 3)
With the growth of NJ statewide program, additional discordant results
have been identified. Through September 2006, 58 additional
discordants have been identified. Of these, 78% were associated with
oral specimens, while 20% were associated with fingerstick specimens.
A slightly lower degree of specificity is reported by the manufacturer
when using oral specimen types and is apparent in NJ.
The possibility of seasonal variation is suggested by monthly data, but
remains within the manufacturer’s specificity claims.
With the exception of a single NJ case, discordant results in 2006 have
not been associated with an evolving infection as demonstrated by > 1
month follow-up utilizing qualitative DNA and quantitative RNA HIV
testing, as well as HIV1/2 EIA testing.