Overestimation of the South African HIV incidence using the BED IgG assay?

Abstract

We have been conducting HIV/AIDS behavioural surveillance research at a large public health clinic that provides sexually transmitted infection (STI) services in Cape Town and have collected data that can help shed light on this urgent problem. In anonymous behavioural surveys collected from 1 729 men and 470 women receiving STI services we have found that 41% of men and 37% of women have experienced condom failure, defined as a broken, torn, or slipped-off condom. In a subsample of 202 patients who reported condom failure, 12% had used oil-based condom lubricants that are known to degrade latex, such as hand creams, vaseline, or oils. In another separate subsample of 214 patients who had experienced condom failure, 7% reported having practised dry sex, although we do not know if the dry-sex practices were directly associated with condom failure. These rates of 30-40% of persons experiencing condom failure are similar to those reported in the US studies cited by Dr Khumalo. 2,3 Our behavioural surveillance data confirm that condom failure is prevalent in at least some high-risk populations in South Africa and may be of particular concern in the populations at highest risk. The causes of condom failure remain undocumented as we found only a minority of cases potentially attributable to improper use of lubricants or dry-sex practices. As stated by Dr Khumalo, there are interventions that reduce condom failure and there are now brief counselling interventions that increase condom uptake and proper use in STI patients tested in South Africa. 4,5 We must also remember that condoms succeed in preventing pregnancy, STI and HIV infection far more often than they fail. We therefore applaud Dr Khumalo's call for more research as well as evidence-based guidelines that include skill-building techniques for improving correct and consistent use of condoms.

Full text

July 2007, Vol. 97, No. 7 SAMJ
B
RIEWE
476
Condom failure in South Africa
To the Editor: It was with great interest that we read the recent
editorial by Dr Khumalo
1
in which she expressed concern
regarding potential condom failure in Africa. The issue of
condom failure is certainly important and we were most
alarmed by the lack of prevalence data on condom failure in
South Africa. In her literature search Dr Khumalo did not find
any research on the prevalence of condom failure in Africa
aside from that in pregnant women.
We have been conducting HIV/AIDS behavioural
surveillance research at a large public health clinic that provides
sexually transmitted infection (STI) services in Cape Town
and have collected data that can help shed light on this urgent
problem. In anonymous behavioural surveys collected from
1 729 men and 470 women receiving STI services we have
found that 41% of men and 37% of women have experienced
condom failure, defined as a broken, torn, or slipped-off
condom. In a subsample of 202 patients who reported condom
failure, 12% had used oil-based condom lubricants that are
known to degrade latex, such as hand creams, vaseline, or
oils. In another separate subsample of 214 patients who had
experienced condom failure, 7% reported having practised
dry sex, although we do not know if the dry-sex practices
were directly associated with condom failure. These rates of
30 - 40% of persons experiencing condom failure are similar to
those reported in the US studies cited by Dr Khumalo.
2,3
Our
behavioural surveillance data confirm that condom failure is
prevalent in at least some high-risk populations in South Africa
and may be of particular concern in the populations at highest
risk. The causes of condom failure remain undocumented as
we found only a minority of cases potentially attributable to
improper use of lubricants or dry-sex practices.
As stated by Dr Khumalo, there are interventions that
reduce condom failure and there are now brief counselling
interventions that increase condom uptake and proper use in
STI patients tested in South Africa.
4,5
We must also remember
that condoms succeed in preventing pregnancy, STI and HIV
infection far more often than they fail. We therefore applaud
Dr Khumalo’s call for more research as well as evidence-based
guidelines that include skill-building techniques for improving
correct and consistent use of condoms.
Leickness C Simbayi
Human Sciences Research Council
Cape Town
Seth C Kalichman
University of Connecticut
USA
seth.k@uconn.edu
1. Khumalo NP. How common is condom failure? S Afr Med J 2007; 97:143.
2. Crosby R, DiClemente R, Wingood GM,
et al. Correlates of condom failure among adolescent
males: An exploratory study. Prev Med 2005; 41:873-876.
3. Bortot AT, Risser WL, Cromwell PF. Condom use in incarcerated adolescent males:
Knowledge and practice. Sex Transm Dis 2006; 33(1):5.
4. Simbayi LC, Kalichman SC, Skinner D,
et al. Theory-based HIV risk reduction counseling
for sexually transmitted infection clinic patients in Cape Town, South Africa. Sex Transm Dis
2004; 31: 727-733.
5. Kalichman SC, Simbayi LC, Vermaak R,
et al. HIV/AIDS risk reduction counseling for
alcohol using sexually transmitted infections clinic patients in Cape Town South Africa. J
Acquir Immune Defic Syndr (Epub ahead of print).
Overestimation of the South African
HIV incidence using the BED IgG
assay?
To the Editor: We thank Rehle et al. for their important study
of HIV incidence in South Africa,
1
which we read with great
interest. We agree with the authors that the incidence of HIV
in South Africa is probably extremely high, particularly among
young women, and believe that the study will help us focus
HIV prevention efforts on appropriate subgroups. We have
serious concerns, however, about the applicability of the BED
IgG assay to the South African HIV epidemic. In light of recent
evidence, we are concerned that Rehle et al. have overstated the
true absolute incidence of HIV in South Africa.
As the name implies, the BED assay was developed using
sequences from HIV subtypes B, D and E.
2
To compensate for
imperfect sensitivity and specificity, Rehle et al. use a correction
factor based on McDougal et al.’s study of subtype B virus.
3
Given that the majority of HIV infections considered by Rehle
et al. were (apparently) of subtype C,
1
the applicability of the
McDougal correction, and indeed of the BED assay itself, to
these samples is problematic. More questions arise in light
of a recent report by Karita et al.
4
that the BED assay does not
perform well in subtype C virus infections; investigators found
a specificity of 71% (95% confidence interval (CI) 54 - 84%),
4
substantially different from one estimate of specificity used in
the McDougal correction
3
(94% for infections more than 360
days in the past). In addition, Karita et al. found that using
the BED assay with the McDougal correction resulted in
overestimation of incidence in prospective Ugandan samples
(subtype not available, but probably A and D
5
), reporting a
corrected BED incidence of 6.4% and a true incidence of 1.3 -
1.7%.
4
We are therefore concerned that the incidence figures
reported by Rehle et al. may be overestimates. If indeed
these figures are incorrect, this will make future comparisons
with more accurate measures of incidence difficult and could
lead to spurious conclusions with regard to the course of the
epidemic. Given these concerns and the current UNAIDS
recommendation against using the BED assay for incidence
estimation,
6
it would be helpful if the authors clarified their
findings with a quantitative sensitivity analysis of their
estimates. Until the BED assay has been further validated, we
Pg 474-480.indd 476 6/20/07 8:07:55 AM

July 2007, Vol. 97, No. 7 SAMJ
B
RIEWE
478
believe that BED-derived estimates of HIV incidence must be
interpreted with caution.
Daniel Westreich
Audrey Pettifor
Department of Epidemiology
University of North Carolina at Chapel Hill
USA
westreic@email.unc.edu
Etienne Karita
Projet San Francisco
Kigali
Rwanda
Matthew Price
International AIDS Vaccine Initiative
New York
USA
Agnes Fiamma
UCLA Program in Global Health
University of California
Los Angeles
USA
Susan Fiscus
Department of Microbiology and Immunology
University of North Carolina at Chapel Hill
USA
Myron Cohen
Division of Infectious Diseases
School of Medicine
University of North Carolina at Chapel Hill
USA
1. Rehle T, Shisana O, Pillay V, Zuma K, Puren A, Parker W. National HIV incidence measures
– new insights into the South African epidemic. S Afr Med J 2007; 97: 194-199.
2. Parekh B, Kennedy S, Dobbs T,
et al. Quantitative detection of increasing HIV type 1
antibodies after seroconversion: A simple assay for detecting recent HIV infection and
estimating incidence. AIDS Res Hum Retroviruses 2002; 18: 295-307.
3. McDougal JS, Parekh, BS, Peterson ML,
et al. Comparison of HIV-1 incidence observed
during longitudinal follow-up with incidence estimated by cross-sectional analysis using the
BED capture enzyme immunoassay. AIDS Res Hum Retroviruses 2006; 22: 945-952.
4. Karita E, Price M, Hunter E,
et al. Investigating the utility of the HIV-1 BED capture enzyme
immunoassay using cross-sectional and longitudinal seroconverter specimens from Africa.
AIDS 2007; 21: 403-408.
5. Yirrell DL, Kaleebu P, Morgan D, Hutchinson S, and Whitworth JA. HIV-1 subtype dynamics
over 10 years in a rural Ugandan cohort. Int J STD AIDS 2004; 15(2):103-106.
6. UNAIDS.
Statement on the Use of the BED-assay for the Estimation of HIV-1 Incidence for
Surveillance or Epidemic Monitoring. Report of a meeting of the UNAIDS Reference Group
for Estimates, Modelling and Projections, Athens, Greece, 13-15 December 2005. Geneva:
UNAIDS, 2005.
Drs Rehle, Shisana, Parker and Puren reply: Westreich and
colleagues express concerns about the applicability of the BED
capture enzyme immunoassay to the South African epidemic
with HIV subtype C as the predominant HIV clade.
The BED assay uses a multi-subtype peptide designed
to cover all major HIV subtypes, not just subtypes B, E and
D as its name may imply. The three main variants of the
immunodominant region of gp41 were used to synthesise the
BED peptide (B Parekh, Centers for Disease Control (CDC)
– personal communication). These consensus sequences are
well preserved and the inclusion of those sequences from the
three subtypes B, E and D was found to be sufficient to cover all
major (group M) subtypes of HIV prevalent in different areas
of the world.
1
The BED peptide is equivalently reactive among
these HIV subtypes as assessed by saturation binding and end-
point titres.
In May 2006, an incidence validation meeting was held
at the CDC where new study results were presented from
China, Cote d’Ivoire, South Africa, Thailand, Uganda, the
USA and Zimbabwe to address the concerns expressed by the
UNAIDS Reference Group in December 2005.
2,3
Working groups
developed guidelines with detailed adjustment procedures
for the estimation of HIV-1 incidence in cross-sectional,
population-based serosurveys.
4
Two separate studies showed
similar misclassification rates among subtype B and subtype C
infections and proposed their own adjustment formulae
5
(and
Hargrove J, et al., ‘Improved HIV-1 incidence estimates using
the BED Capture Enzyme Immunoassay’ – in review).
Values for the imputed variables for both adjustment factors
were validated in 2 532 specimens from 1 192 people with
known date of seroconversion in HIV-1 subtypes B and C.
The key imputed value in these adjustments is the false recent
rate among long-term (> 1 year) infected people. It is 5.57%
in both adjustments (1-γ in McDougal’s adjustment is equal
to ε in Hargrove’s adjustment). Therefore, the McDougal and
Hargrove adjustments have only been validated for HIV-1
subtypes B and C where the proportion of long-term infection
misclassifying as recent infections were quantified. The
performance of these adjustments in populations with HIV-1
subtypes A, D and E is not yet known and is being validated.
The study of Karita et al.
6
quoted by Westreich and colleagues
questions the validity of the adjustments applied in our
analysis. However, in view of the large samples from which the
McDougal and Hargrove adjustments were derived, a major
limitation of the analysis by Karita et al. was the small sample
size used in the BED performance assessment in subtype C
specimens – only 117 samples from 26 Zambian volunteers.
Furthermore, based on previous analysis of HIV subtype C
seroconverter samples (Ethiopia, Zimbabwe) done at the CDC
we have applied a window period of 180 days in our incidence
calculation. This is in contrast to the window period of 153 days
used by Karita et al.
In order to examine the plausibility of our HIV incidence
estimates we compared the adjusted BED estimates with
estimates derived from mathematical modelling, using the
ASSA2003 AIDS and Demographic model.
7
BED HIV incidence
in the population aged 2 years and older was 1.4%, compared
with 1.3% estimated by the ASSA model. A BED HIV incidence
rate of 2.4% was found among individuals aged 15 - 49 years.
The modelled HIV incidence was 2.2% for this age group.
We therefore conclude that the adjusted BED HIV incidence
estimates appear to provide plausible national HIV incidence
estimates for South Africa.
Pg 474-480.indd 478 6/20/07 8:07:55 AM

July 2007, Vol. 97, No. 7 SAMJ
B
RIEWE
480
Notwithstanding these encouraging results we remain
actively involved in further validation studies not limited to
the BED-CEIA but will also explore the suitability of testing
algorithms involving, for example, antibody avidity testing.
There is emerging consensus that validated laboratory
based tests are the method of choice to estimate national
HIV incidence and assess the impact of national prevention
programmes.
1. Parekh B, Kennedy S, Dobbs T, et al. Quantitative detection of increasing HIV type 1
antibodies after seroconversion: A simple assay for detecting recent HIV infection and
estimating incidence. AIDS Res Hum Retroviruses 2002; 18(4): 295-307.
2. Centers for Disease Control (CDC), Surveillance and Survey and Laboratory Working
Groups. Expert meeting on the validation of the BED HIV-1 incidence assay for HIV-1
incidence surveillance. CDC, Atlanta, USA, 9-10 May 2006.
3. UNAIDS.
Statement on the Use of the BED-assay for the Estimation of HIV-1 Incidence for
Surveillance or Epidemic Monitoring. Report of a meeting of the UNAIDS Reference Group
for Estimates, Modelling and Projections, Athens, Greece, 13-15 December 2005. Geneva:
UNAIDS, 2005.
4. Centers for Disease Control (CDC), Surveillance and Survey and Laboratory Working
Groups. Guidelines for the Use of the BED Capture Enzyme Immunoassay for Incidence Estimation
and Surveillance. Atlanta, USA: CDC, 2006.
5. McDougal JS, Parekh, BS, Peterson ML,
et al. Comparison of HIV-1 incidence observed
during longitudinal follow-up with incidence estimated by cross-sectional analysis using the
BED capture enzyme immunoassay. AIDS Res Hum Retroviruses 2006; (10): 945-952.
6. Karita E, Price M, Hunter E,
et al. Investigating the utility of the HIV-1 BED capture enzyme
immunoassay using cross-sectional and longitudinal seroconverter specimens from Africa.
AIDS 2007; 21: 403-408.
7. Rehle T, Dorrington R, Shisana O,
et al. National HIV incidence estimates: direct measures
compared with mathematical modelling. Paper presented at the 3rd South African AIDS
Conference, Durban, 5-8 June 2007.
African section of e-journal Rural and
Remote Health
To the Editor: We read with interest the SAMJ article ‘Scope
and geographical distribution of African medical journals active
in 2005’ by Siegfried et al.,
1
and would like to bring to your
readers’ attention the recent launch of an African section of the
e-journal Rural and Remote Health (RRH). This regional section
has a particularly African flavour, owing to its own editorial
board and peer-review panel, but is under the umbrella of the
international journal.
We hope that the African section will add to the initiatives
described by Siegfried et al. and address some of the issues
raised in their article. RRH is an international, peer-reviewed,
open-access journal. It is Medline-listed. It aims to offer wider
world exposure for quality African research in the area of rural
and remote health care education, policy and practice. We
believe the issues of rural and remote health are relevant to
most of Africa.
Because RRH is an electronic journal it affords authors timely
publication on an article-by-article basis. In addition, the
electronic format means that RRH is not geographically bound,
and therefore offers rural and remote authors and users an all-
of-Africa approach to publication.
In a recent RRH editorial to coincide with the launch of the
African section, we recognised the impact of inadequate access
to information on the problems of health and health care in
Africa.
2
We also discussed the issue of inequity in access to the
Internet, which has been highlighted for urgent attention by
the Commission for Africa,
3
and recent initiatives to improve
the current situation of variable access.
4,5
We offer the African
section of RRH as a small contribution towards this.
The Journal can be accessed at www.rrh.org.au. Users should
select ‘African section’ from the main menu on the home page.
Jennifer Richmond
Production Editor, RRH
Australian Rural Health Education Network
Canberra, ACT
Australia
Ian Couper
Editor, African section, RRH
Professor of Rural Health
Department of Family Medicine
University of the Witwatersrand
Johannesburg
couperid@medicine.wits.ac.za
Paul Worley
Editor-in-chief, RRH
Professor and Director
Rural Clinical School
Flinders University, South Australia
1. Siegfried N, Busgeeth K, Certain E. Scope and geographical distribution of African medical
journals active in 2005. S Afr Med J 2006; 96: 533-538.
2. Couper ID, Worley PS. Health and information in Africa: the role of the journal
Rural and
Remote Health. Rural and Remote Health 6 (online), 2006: 644. http://rrh.deakin.edu.au (last
accessed 14 September 2006).
3. Dare L, Buch E. The future of health care in Africa.
BMJ 2005; 331: 1-2.
4. Katikireddi SV. HINARI: bridging the global information divide.
BMJ 2004; 328: 1190-1193.
5. Beveridge M, Howard A, Burton K, Holder W. The Ptolemy project: a scalable model for
delivering health information in Africa. BMJ 2003; 327: 790-793.
Pg 474-480.indd 480 6/20/07 8:07:56 AM