728

TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE

One year surveillance

AND HYGIENE

of HIV-1 infection

(1990)

84,

72tL730

in Johannesburg,

South Africa

Desmond J. Martin’, Barry D. Schoub’, Gopalan N. Padayachee’, Alan N. Smith’, Susan F. LyonsI, Gillian M. McGillivray’, Shan Naidoo2 and Hilliard S. Hurwitz2 ‘Department of Virology, University of the Witwatersrand; ‘Johannesburg City Health Department and Department of Community Health, University of the Witwatersrand Abstract A sero-epidemiological surveillance study to monitor the prevalence of HIV-l infection in Johannesburg, South Africa, was commenced in February 1988. The population selected for study were attenders at clinics for sexually transmitted diseases (STD) and at family planning (FP) clinics. In the 12 months of the study 6631 sera were tested. Of the STD attenders, 15 of 1224 black females (1.2%) and 21 of 2482 black males (0.8%) were positive. Of the 449 white males tested 49 were homosexual, amongst whom 10 (20.4%) were positive; in the heterosexual white male group 4 of 400 (1.0%) were positive. Of the FP clinic attenders, 4 of 1459 black females (0.3%) were positive. 68 of the 6631 sera tested were indeterminate for infection. No attenders were positive for HIV-2 infection. These data confirmed the entry of HIV infection into the black population in South Africa. Table 1. Attendance at family 1988-31 January 1989

planning

clinics

White

and sexually

The programme was established jointly by the National Institute for Virology and the City Health Department of the Johannesburg City Council. Subjects and Methods Blood specimens from attenders at sexually transmitted diseases (STD) and family planning (FP) clinics were submitted to the laboratory anonymously, although data collected included sex, age, race and sexual preference. The study was sanctioned by the ethics committees of the University of the Witwatersrand and the Johannesburg City Council. The total number of attenders at the clinics used in the survey is shown in Table 1. The ethnic distribution is representative of the population in greater Johannesburg. A sample was drawn from those persons attending. The sampling technique used was a systematic sample with fixed sample intervals from a random start. The sample size was determined by the transmitted

Coloured

attendance attendances

Sexually transmitted New Cases Re-visits Total

in Johannesburg,

1 February

Black

Asiatic

Total

F

M

F

M

F

M

F

M

F

1::

2395 19086

494 561

2061 23556

1::

841 6300

1013 1590

7621 69243

1607 2422

12918 118185

184 164 348

242 323 565

2

:4

73

28

7167 8929 16096

1910 1797 3707

8121 10566 18687

2374 2384 4758

disease 736 1434 2170

clinics 206 252 458

Introduction The dominant acquired immunodeficiency syndrome (AIDS) epidemic in western countries and in the white population of South Africa has been in the male homosexual population, while in Central Africa the epidemic has occurred mainly among the heterosexual population (SCHOUB et al., 1988). Until recently in South Africa the infection was absent from black heterosexual people attending clinics for sexually transmitted diseases (SCHOUB et at., 1987). This study was undertaken to momtor the prevalence of human immunodeficiency virus (HIV) infection in samples from various populations. Active surveillance commenced in February 1988 and is continuing; it will establish epidemiological trends over time. The data for the first 12 months are the basis of this report. Requests Research Africa.

clinics

M

Family planning clinics First Total

diseases

for offprints to Dr D. J. Martin, MRC Unit, Private Bag X4, Sandringham

AIDS Virus 2131, South

capacity of the laboratory to handle the specimens. The survey reported here was conducted between 1 February 1988 and 31 January 1989. From February to June 1988 a first generation enzyme-linked immunosorbent assay (ELISA) was used for detection of antibodies to HIV-l (Elavia I@, Diagnostics Pasteur, Paris, France). In June an ELISA for detection of antibodies to both HIV-l and HIV-2 (Rapid-Elavia Mixt@, Diagnostics Pasteur) was introduced and used throughout the remainder of the study for the initial screening. Repeatedly positive specimens were further evaluated using individual ELISAs for HIV-l and HIV-2 (Elavia I@ or Elavia II@ respectively, Diagnostics Pasteur); positive specimens were confirmed using either an in-house indirect immunofluorescence assay (IFA) (SHER et al., 1985), or, if the sera were IFA negative or non-specific, commercial Western blots (WB) for HIV-l (lav-Blot I@, Diagnostics Pasteur) or HIV-2 (lav-Blot II@, Diagnostics Pasteur or DuPont HIV-2 IgG Blot@,

729

Table 2. Prevalence esburg, 1 February

of HIV seropositivity among attenders 1988-31 January 1989

Female? No. examined No. positive White Black

1224 117

Coloured

15 8

160 (1.1% (-> [05-1.71)

?

21 1; (04% [!;l% [o+1.2])b [l-5-4.7])

2482 449

1 (2.0% [O*l-9.31)

1395 5

disease clinics in Johann-

Malesa No. examined No. positive

?

1; ;A;‘% [06-1.81)

49

Etai

at sexually transmitted

3;

72

1:

0

30;:

350 6) (1.2%

[O-S-1.6])

3:

‘Numbers in squarebrackets indicate 95% confidenceintervals; ?=indeterminate (seetext). bHomosexual:lo/49 (20.4%); heterosexual4/400 (1.0%). Singapore).All assayswere performed accordingto the manufacturers’directions. For specimensto be confirmed positive for HIV-l we used the criteria recommendedby the CONSORTIUM FOR RETROVIRUS SEROLOGYSTANDARDIZATION(1988). Reactivity with p24 or p31 and gp41or gp120/160was regardedas being a positive WB for HIV-l and, by extrapolation, reactivity with p26 and gp41 or gp140aspositive for HIV-2. Other banding patterns were reported as indeterminate. Results

A total of 6631 sera was tested, 4410 from individuals attending clinics for STD and 2221from individuals attending an FP clinic. STD

clinic attendm

The resultsareshownin Table 2. The ratio of black male seropositivesto black femaleseropositiveswas 1:1.4. FP clinic attenders

All of the 4 seropositive black females were Table 3. Prevalence of HIV seropositivity among attenders at family planning clinics in Johannesburg, 1 February 198&31 January 1989 No.

Black White p;rd

examinedNo. positivea 1459 ; if;'% [O-0*6]) 593

116 53

? 12 :

8 )I{

Total 2221 4 0.2 (O-0.4) 1: “Numbersin squarebracketsindicate 95% confidence intervals;?=indeterminate (seetext). Table 4. Prevalence at 30 June 1989a

Black White Coloured Asiatic Total

of HIV

seropositivity

Females 101/180335(0.06%) 11543450

(O-0002%)

l/60247 (0.002%).

among

asymptomatic.No seropositives wereobtainedamong white, coloured or Asiatic females(Table 3). Sixty-eight serawere categorizedas being indeterminate for infection (Tables 2 and 3). No attenders were positive for HIV-2 infection. Discussion

In South Africa, both pattern 1 (transmissionof HIV predominantlyamongstmalehomosexuals)and pattern 2 (transmissionpredominantly by heterosexual contacts)are significantcomponentsof the overall epidemicsof HIV-l and AIDS. Passivesurveillance seroprevalence data gainedfrom routine screeningof blood donorsprovide low risk population seroprevalences(Table 4). Among white donors the O-004% prevalenceof HIV antibodies(confirmedby WB test) waslow andfalls betweenthe 0.001%reportedfor the United Kingdom (COMMUNICABLEDISEASESURVEILLANCECENTRE,1988)and 0.012%for the USA (CENTERS FORDISEASECONTROL,1987), and the male/female(M/F) sexratio of 26:1 is consistentwith pattern 1 AIDS. These figures undoubtedly underrepresentthe generalpopulation prevalencebecause of voluntary self-exclusionby individualsin high risk groups. The prevalenceof infection in black blood donors (0*040/o) was much higher than in the white donors,althoughin this populationthere is probably no significantvoluntary self-exclusionand the prevalencefigureswould be morecloselyrepresentativeof the overall population. The M:F ratio of 1:1.6 amongstblack blood donorsis consistentwith pattern 2 HIV transmission(Table 4). The presentstudy examinedthe prevalencein both high and low risk populations.In a previousstudy, in 1987,on 2 cohorts of promiscuousblack women in Johannesburg(a prostitute cohort and a cohort of STD clinic attenders),only 1 seropositivepersonwas found-a Malawian woman (SCHOUB et al., 1987). blood

donors

in the

Males 81/233493 (0.03%) 5511072363 6/96075 l/83942

(0.005%) (O.OOS%j

Republic

of South

Total 0.04% 0.003% 0.004%

(0.001%) 0.001% 103/784032(0.01%) 14311485873 (0.01%) 0.01% aFiguresby courtesy of Dr C. Prior, Medical Director, Natal Blood Transfusion Service.

Africa

Ratio (M:F) ~ , 1:1.6 26:l 4:l

-

730

It is clearly evident from the figures obtained during this surveillance that there has been an entry of HIV ihfection into the black heterosexual popula&on, especially among those individuals attending STD cl&tics. The se&prevalence rate in the black-female group (1.2%) is alarmingly high considering the relatively short space of time since the last survey in this group. The finding of 1% HIV infection prevalence among the white heterosexual males is to be treated with considerable reserve as it is difficult to be certain that these subjects were truly heterosexual. It should be noted that there were no positives in the white female group. That STD act as cofactors in the acauisition of HIV infection is well known and a high incidence of HIV infection occurs in patients with STD (PIOT et al., 1987; GREENBLATT et al., 1988). Female prostitutes play a major role in the spread of AIDS in Africa; they consistently exhibit the highest HIV seroprevalence rates and are also an important reservoir of other STD (VAN DE PERRE et al.. 1985: MANN et al.. 1988: ~