Title: Can a second rapid HIV test discriminate false positives as effectively as a Western Blot? The NJ Experience
1Can a second rapid HIV testdiscriminate false
positives as effectively as a Western Blot? The
NJ Experience
Cadoff EM, Cadoff RA, Salaru G, Paul SM, Martin EG
- Evan M. Cadoff, MD
- Robert Wood Johnson Medical School
- New Brunswick, NJ
2NJHIV Rapid Testing Program
- Under direction of RWJMS
- OraQuick started November 1, 2003
- Oversees about 2/3 of NJ CTS HIV testing
- Confirmatory testing at PHEL in Trenton
- NJHIV oversees followup of all discordant
- NJHIV does not evaluate linkage to care
3Preliminary Positive Followup
- 7.1 refused blood draw for confirmation
- 25.8 of those drawn did not return for results
- 70.1 of confirmed positives got their results
and post-test counseling - Rapid confirmation could improve effectiveness of
prevention and referral/entry to care and
treatment services
4Data Needs for Proposed Testing Strategies
- Are there false negative screening tests?
- Can true positive screening tests be confirmed by
a second rapid test, rather than WB/IFA/NAT? - Can false positive screening tests (ie
discordants) be detected by a second rapid test,
rather than waiting for WB/IFA/NAT? - Impact on linkage to care?
- How well can inconclusive second-round test
results (eg, WB vs a second rapid test) be
resolved?
5Two NJHIV Datasets
- Confirm true positives
- Evaluate confirmed false positives
- Resolve discordants and indeterminate
confirmatory test results
6Nov 2003 April 2005 data
- Rapid testing by blood only using OraQuick
- Western Blot specimens sent to PHEL
- All avialable WB specimens evaluated by
- Repeat OraQuick on blood
- Trinity Uni-Gold
- MedMira Reveal
- BioRad Multispot
- Followup of clients with discordant results
7Nov 2003 April 2005 data
- 15,923 OraQuick tests statewide
- 363 preliminary postive samples to state lab for
Western Blot - 355 Western Blot positive confirmed by other
rapids - 8 Western Blot negative discordants
- all repeat OraQuick positive
- one reactive Multispot by one reader
- others all negative
- 6 preliminary positive is false positive
- 25 with no clinical followup
Strategy 1, Data Need 3
8Discordant followup
- Four visits
- Rapid test. If prelim pos, draw blood and send
to Trenton for Western Blot. - Return in a week, and get discordant confirmatory
test result. - Return at a month to have more blood drawn
(repeat antibody and NAAT). - Come back a week later for definitive result.
9Discordant followup
10Revised discordant followup
- Three Visits
- If rapid test is reactive, draw Western Blot.
- Return in a week. Get discordant confirmatory
test result and draw NAAT. - Third visit for NAAT result.
- or
- Two Visits
- If rapid test is reactive, draw both Western Blot
and NAAT. - NAAT is run if WB is discordant, and both results
are available at second visit.
11Revised discordant followup
- Three Visits
- If rapid test is reactive, draw Western Blot.
- Return in a week. Get discordant confirmatory
test result and draw NAAT. - Third visit for NAAT result.
- or
- Two Visits
- If rapid test is reactive, draw both Western Blot
and NAAT. - NAAT is run if WB is discordant, and both results
are available at second visit. - or
- One visit with rapid confirmation would be better!
12Jan 2006 Oct 2007 data
- Retained specimens from followup testing for
discordant Western Blot results. - Cannot confirm true positives.
- Can evaluate discordants and indeterminates (if
they had followup).
13Jan 2006 Oct 2007 data
- Samples are not from the same time as screening
OraQuick. - Used CLIA-waived tests
- Repeat OraQuick on blood
- Trinity Uni-Gold
- Clearview StatPak
14Jan 2006 Oct 2007 data
- 108 false positive 72 without followup may be
false positive Strategy 1, Data need 3 Strategy
4, Data need 5a - No false positive strategy results Strategy
4,Data need 1 - No false negative A2 or A3 Strategy 4, Data need
4, 5b
15Jan 2006 Oct 2007 data
- No false positive strategy results Strategy 3,
Data need 1 - 1 false positive A1-blood result Strategy 3,
Data need 4a,5 - 91 false positive A1-oral results Strategy 3,
Data need 5 - No false negative A2 results Strategy 3, Data
need 4b
16Jan 2006 Oct 2007 data
No false positive strategy results Strategy 2,
Data need 1 17 Inconclusive results with false
positive A1 Strategy 2, Data need 4
17Data Needs for Proposed Testing Strategies
- All followups on negative Western Blot were NAAT
negative (OraQuick false positive). - All were "second rapid" negative.
- Oral OraQuick followed by blood OraQuick
- 3 tested at CTS site on False Positives
- 2 blood negative 1 blood positive
- 56 tested on followup blood specimen
- all blood negative
- OraQuick blood discordant
- 11 tested on followup blood specimen
- 7 negative
- 4 repeat positive
18Observations
- Indeterminate Western Blot
- 12 total
- 4 no followup 3 QNS
- 3 NAAT negative were "second rapid" negative
- 2 NAAT positive were "second rapid" positive
19Rapid Confirmation Is Needed
- Rapid confirmation of preliminary positive
results is necessary - 30-40 without followup would have benefited from
rapid confirmatory testing.
20Rapid Confirmation Works
- All 355 true positives confirmed using a second
rapid test (from a pool of about 16,000 clients
tested) - All 65 false positives with followup available
were negative using a second rapid test (from a
pool of about 120,000 clients tested) - Indeterminate Western Blots were resolved by a
second rapid
21Conclusions
- Rapid confirmation of preliminary positive
results is necessary. - Rapid confirmation of preliminary positive
results works at least as well as Western Blot
testing.
22(No Transcript)