The selection of cities, populations of focus and sample size were based on implementing partners’ service delivery platform,[1] available resources and relative size of each city. Prospectively, we aimed to recruit 3 500 participants from 11 sites across the seven cities. Cities were divided into two categories; those above, and those below, 1.5 million total population based on the latest census (2011)[40]. For SWs and PWUD/ID, 400 of each population group were to be recruited from larger cities and 250 from smaller cities. There were more SW sites due to the larger SW HIV and health service programme relative to other KP programmes at the time. PWUD/ID were recruited from cities that operated harm reduction services. PWUD: PWID were recruited in the ratio 1:4 due to the increased risk of blood borne infection transmission through injecting practices. Two hundred and fifty MSM were recruited at each site in the three cities that implementing partners operated MSM health services. Fewer MSM were recruited in each city, due to the existence of some data for this population and due to the limited resources available. The sample size per population was 1 550 SWs (400 each in Cape Town and Durban; 250 each in Pietermaritzburg, Port Elizabeth and Mthatha), 750 MSM (250 each from Cape Town, Pretoria and Johannesburg), and 1 200 PWUD/ID (400 each in Cape Town, Durban and Pretoria).
Consenting participants self-defined as being 18 years or older and having met the following criteria, as relevant, in the past 12 months i) SW: having exchanged sex for money[2]; ii) MSM: having been born biologically male and having had penetrative or receptive oral or anal sex with another male; iii) PWUD: having used heroin, cocaine or methamphetamine and iv) PWID: having injected a substance for non-therapeutic purposes, irrespective of the type of drug injected or the mode of injection.
Participants were conveniently sampled. Study activities were conducted as an extension of existing HIV prevention, sexual health and harm reduction services provided by non-profit organisations targeting KPs in the respective cities. Sites were designated to recruit from the specific population they already served. Each of the service providers operate dedicated teams that provide targeted services to specific KPs. In Cape Town, for example, routine health services for SWs and PWUD/ID were provided by a single service provider out of a single physical location, but were provided in parallel by different mobile teams and were designated as two distinct study sites recruiting SW or PWUD/ID respectively. Although there is alignment of services, each site tended to have separate clients depending on how the clients themselves best identified. Participants recruited through a SW site and who met inclusion criteria would have been counted as a SW irrespective of substance use behaviour and vice versa for PWUD/ID. Similarly, at MSM sites, people accessing HIV testing and/ or treatment services were informed about the study and invited to participate and would have been recruited as MSM. Refreshments were offered during the last month of the study’s recruitment period to encourage enrolment at PWUD/ID sites to meet the recruitment targets within the time frame allocated for fieldwork.
Data collection took place between August 2016 and October 2017. In all cases, a study team member administered a standard, confidential health-screening questionnaire to participants who provided written informed consent.
The questionnaire was the standard form that PWUD/ID and SW implementing partners used as part of their service delivery. The questionnaire captured data on demographics (age, sex, race and housing), substance use (substances used in the last month and injecting in the last month, including injecting frequency, needle and syringe reuse and sharing), use of opioid substitution therapy and sexual risk behaviour (condom use, penile-vaginal intercourse, anal intercourse, transactional sex and substance use in the context of sex) and prior health screens and previous HIV testing.
The questionnaire was completed in English. Questionnaires were administered in private spaces. Where closed rooms were not possible, questionnaires were completed out of hearing of other people. Service providers and study staff were all trained in ethical research practices and signed research confidentiality agreements. The paper-based questionnaire was filed and then locked away. The questionnaire captured participants’ names, surnames, dates of birth and included a participant identification code. Participants were not obligated to provide identifying information. Access to participant names and surnames was restricted to clinicians providing health services and staff conducting the questionnaire. All other study documentation, including blood samples, used unique participant identification codes.
A total of 20 ml whole blood was drawn from consenting participants and used for HIV, HBV and HCV point-of-care (POC) testing. POC HIV testing was done in line with national protocols [41]. POC HBsAg testing was performed using the DetermineTM HBsAg (Alere Inc, MA, USA). POC HCV testing was done using OraQuickÒ HCV rapid antibody test (Orasure Technologies Inc, PA, USA). Samples that were HCV POC reactive were sent to the National Institute for Communicable Diseases (NICD) in Johannesburg for additional testing. Samples received by the laboratory underwent anti-HCV testing (ARCHITECT Anti-HCV assay) on the ARCHITECT i1000SR system (Abbott Laboratories, Diagnostics Division, IL, USA). The COBAS® AmpliPrep/COBAS ® TaqMan® HCV Quantitative Test v2.0 on the COBAS® Ampliprep Taqman® Analyzer (Roche Molecular Systems, CA, USA) was used for quantitative assessment of HCV antibody positive samples. PCR amplification and HCV genotyping was performed using the Versant LiPA amplification and Versant HCV genotyping 2.0 assays (LiPA, Innogenetics, Ghent, Belgium), respectively.
Study team members provided participants with appropriate counselling, condoms and lubricant, and PWID received sterile injecting equipment. Participants who screened positive for hepatitis B or who were diagnosed with confirmed HIV or HCV infection were referred to a pre-identified health facility for work-up and/or assessment for treatment. HBV vaccination was offered to all clients who were HBsAg POC negative[3].
Anonymised data was imported into STATA 14.0 (Statacorp LLC, Texas, USA) for data analysis. Data was analysed using descriptive statistics (proportions, medians and inter quartile range (IQR)), stratified by KP. Age was also categorised based on age groups used by the Joint United Nations Programme on HIV/AIDS, with a binary cut off value of 25 years [42]. Racial groupings were based on those used by Statistics South Africa [43]. POC seroprevalence was calculated for HIV and HBsAg for tests completed as per manufacturers guidelines and testing protocols. HBsAg seroprevalence was also calculated for people born before and after the introduction of childhood HBV vaccination in 1995. Anti-HCV prevalence was based on laboratory results. The HCV viraemic rate was calculated by dividing the number of participants with detectable viral load over the number of eligible participants with laboratory based ELISA confirmation.
The study was approved by the Human Research Ethics Committee of the University of Cape Town (ref: 004/2016), the Research Ethics Committee of the University of the Witwatersrand (ref: M160510) as well as the Eastern Cape (ref: EC_2016RP19_818), Western Cape (ref: WC_2016RP19_818) and KwaZulu-Natal (ref: KZN_2016RP59_986) Provincial Department of Health Ethics Committees. Participants did not receive any remuneration.
Findings
In total, 3509 participants were recruited and enrolled into the study. Seventy participants were excluded from the per protocol analysis. Reasons for exclusion included: ineligibility (did not meet criteria for KP) (n=23), study informed consent form could not be located (n=11), source documents missing (n=4), error in testing procedure for HBV and/or HCV or test not done (n=32). In a per protocol analysis, 3 439 KP were included (1 528 SWs, 746 MSM, 1 165 PWUD/ID).
Socio-demographics
Table 1 outlines participant socio-demographics. Participants’ median age was 29 years (IQR 25 – 35) and most were black (60%, 2 078/3 439). Gender distribution was similar when all KPs were pooled (males were 52%, 1 798/3 439), but varied between KPs. SWs were almost exclusively female, in all locations other than Cape Town, where male SWs were recruited (12%, 47/284). PWUD/ID were predominantly male (85%, 999/1 175). Almost a quarter of participants (23%, 807/3 439) were between the ages of 18 and 24 years. Age did not vary particularly between KP sub-groups, bar two exceptions: in Mthatha close to half (42%, 104/248) of the SWs were under 25 years of age and in Cape Town PWUD/ID tended to be older, with only 8% (28/367) under 25 years. PWUD/ID disproportionately reported homelessness; ranging from 55% (201/367) in Cape Town to 70% (278/398) in Durban. In Cape Town, SWs also reported higher levels of homelessness (17% (66/384)) than were reported in other SW groups.
|
SWs
(n= 1 528)
|
MSM
(n= 746 )
|
PWUD/ID
(n = 1 165)
|
TOTAL
|
|
CT
|
PE
|
Mthatha
|
DBN
|
PMB
|
CT
|
JHB
|
PTA
|
CT
|
DBN
|
PTA
|
|
N
|
384
|
248
|
248
|
399
|
249
|
250
|
250
|
246
|
367
|
398
|
400
|
3 439
|
Age
|
|
|
|
|
|
|
|
|
|
|
|
|
Median (IQR)
|
31
(26 – 37)
|
28
(24 – 35)
|
26
(23 -30)
|
30
(25 – 35)
|
27
(24 – 31)
|
31
(26 – 39)
|
29
(24 – 34)
|
29
(24 – 37)
|
31
(28 – 35)
|
27
(25 – 31)
|
30
(26-34)
|
29
(25 – 35)
|
<25 years
|
80
(21%)
|
78
(31%)
|
104
(42%)
|
90
(23%)
|
78
(31%)
|
49
(20%)
|
68
(27%)
|
65
(26%)
|
28
(8%)
|
96
(24%)
|
71
(18%)
|
807
(23%)
|
³ 25 year
|
304
(79%)
|
170
(69%)
|
144
(58%)
|
309
(77%)
|
171 (69%)
|
201
(80%)
|
182
(73%)
|
181 (74%)
|
339
(92%)
|
302
(76%)
|
329
(82%)
|
2 632
(77%)
|
Sex
|
|
|
|
|
|
|
|
|
|
|
|
|
Female
|
337
(88%)
|
247 (>99%)
|
248 (100%)
|
397 (>99%)
|
246 (99%)
|
0
|
0
|
0
|
69
(19%)
|
42
(11%)
|
55
(14%)
|
1 641
(48%)
|
Male
|
47
(12%)
|
1
(<1%)
|
0
|
2
(<1%)
|
3
(1%)
|
250
(100%)
|
250 (100%)
|
246 (100%)
|
298
(81%)
|
356
(89%)
|
345
(86%)
|
1 798
(52%)
|
Race*
|
|
|
|
|
|
|
|
|
|
|
|
|
Black
|
106
(28%)
|
182
(73%)
|
247
(99%)
|
369
(92%)
|
249 (100%)
|
102
(41%)
|
197
(79%)
|
118 (48%)
|
15
(4%)
|
218
(55%)
|
275
(69%)
|
2078
(60%)
|
Coloured
|
247
(64%)
|
49
(20%)
|
0
|
12
(3%)
|
0
|
47
(19%)
|
9
(4%)
|
9
(4%)
|
274
(75%)
|
45
(11%)
|
13
(3%)
|
705
(21%)
|
White
|
18
(5%)
|
17
(7%)
|
0
|
5
(1%)
|
0
|
92
(37%)
|
37
(15%)
|
109 (44%)
|
71
(19%)
|
96
(24%)
|
109
(27%)
|
554
(16%)
|
Indian/ Asian
|
1
(<1%)
|
9
|
0
|
10
(3%)
|
0
|
4
(2%)
|
2
(1%)
|
4
(2%)
|
1
(<1%)
|
39
(10%)
|
1
(<1%)
|
62
(2%)
|
Other
|
12
(3%)
|
0
|
1
(<1%)
|
3
(1%)
|
0
|
5
(2%)
|
5
(2%)
|
6
(2%)
|
6
(2%)
|
0
|
2
(1%)
|
40
(1%)
|
Homeless
|
66
(17%)
|
1
(<1%)
|
0
|
0
|
0
|
14
(6%)
|
1
(<1%)
|
3
(1%)
|
201
(55%)
|
278
(70%)
|
262
(66%)
|
826
(24%)
|
General population**
|
3 740 026
|
1 152 115
|
451 710
|
3 442 361
|
618 536
|
3 740 026
|
4,434,827
|
2 921 488
|
3 740 026
|
3 442 361
|
2 921 488
|
|
* Based on race categories used by Statistics South Africa. Coloured is a South African term which refers to people of mixed ancestry. The national distribution of race is: Black 80,9%, Coloured 8,8%, White 7,8%, Indian/ Asian 2,5% [43]
** Population of municipality based on most recent census (2011)
CT: Cape Town; DBN: Durban; JHB: Johannesburg; PE: Port Elizabeth; PMB: Pietermaritzburg; PTA: Pretoria
|
Table 1
Participants’ socio-demographic characteristics (n= 3 439)
Substance use
Overall, 82% (2 818/3 439) of participants reported use of at least one of the substances enquired about in the previous month. Alcohol was the most commonly reported substance used (46%, 1 568 / 3 439), and use was higher among SWs and MSM than among PWUD/ID. Heroin (including heroin in combination with other substances, for example cannabis – known locally as nyaope or whoonga) was the most frequently reported illegal/unregulated substance used in the last month (33%, 1 138/3 439) followed by methamphetamine (14%, 486/3 439), cannabis (13%, 433/3 439), cocaine (6%, 204/3 439) and methcathinone (2%, 68/3 439) (Table 2). No SWs in Mthatha reported any illegal/non-regulated substance use in the last month. Almost all PWUD/ID had used heroin in the last month, ranging from 82% (300/367) in Cape Town to 99% (394/398) in Durban. Methamphetamine use in the last month was highest in Cape Town across all three sub-groups: 80% of PWUD/ID (293/367), 28% of SWs (107/384) and 11% of MSM (27/250). PWUD/ID living in Pretoria had the highest reported use of cannabis (30%, 121/400) and cocaine (30%, 118/400). This was substantially higher than that reported for any other KP sub-group or region for all three substances. Cannabis use in other KPs was variable, ranging from none reported among SWs in Mthatha to roughly a fifth of MSM in Johannesburg.
|
SWs
(n= 1 528)
|
MSM
(n= 746 )
|
PWUD/ID
(n = 1 165)
|
TOTAL
|
|
CT
|
PE
|
Mthatha
|
DBN
|
PMB
|
CT
|
JHB
|
PTA
|
CT
|
DBN
|
PTA
|
|
N
|
384
|
248
|
248
|
399
|
249
|
250
|
250
|
246
|
367
|
398
|
400
|
3 439
|
Substances used in last month
|
|
|
|
|
|
|
|
|
|
|
|
|
Any substance*
|
296
(77%)
|
216
(87%)
|
194
(78%)
|
305
(76%)
|
160
(64%)
|
176
(70%)
|
199
(80%)
|
119
(48%)
|
361
(98%)
|
396
(99%)
|
396
(99%)
|
2 818
(82%)
|
Alcohol
|
217
(56%)
|
205
(83%)
|
194
(78%)
|
295
(74%)
|
148
(59%)
|
152
(61%)
|
192
(77%)
|
102
(42%)
|
29
(8%)
|
19
(5%)
|
15
(4%)
|
1 568
(46%)
|
Heroin**
|
32
(8%)
|
1
(<1%)
|
0
|
0
|
12
(5%)
|
17
(7%)
|
3
(1%)
|
1
(<1%)
|
300
(82%)
|
394
(99%)
|
378
(95%)
|
1 138
(33%)
|
Methamphetamine
|
107
(28%)
|
17
(7%)
|
0
|
2
(1%)
|
1
(<1%)
|
27
(11%)
|
11
(4%)
|
6
(2%)
|
293
(80%)
|
8
(2%)
|
14
(4%)
|
486
(14%)
|
Cannabis
|
55
(14%)
|
14
(6%)
|
0
|
32
(8%)
|
9
(4%)
|
39
(16%)
|
43
(17%)
|
31
(13%)
|
45
(12%)
|
44
(11%)
|
121
(30%)
|
433
(13%)
|
Cocaine
|
12
(3%)
|
0
|
0
|
31
(8%)
|
4
(2%)
|
11
(4%)
|
4
(2%)
|
2
(1%)
|
9
(2%)
|
13
(3%)
|
118
(30%)
|
204
(6%)
|
Methcathinone
|
4
(1%)
|
2
(1%)
|
0
|
20
(5%)
|
0
|
8
(3%)
|
11
(4%)
|
10
(4%)
|
1
(<1%)
|
1
(<1%)
|
11
(3%)
|
68
(2%)
|
Injected a drug in the last 12 months
|
0
|
0
|
0
|
0
|
0
|
24
(10%)
|
9
(4%)
|
8
(3%)
|
292
(80%)
|
315
(79%)
|
313
(79%)
|
961
(28%)
|
Injected a drug in last month
|
0
|
0
|
0
|
0
|
0
|
22
(9%)
|
8
(3%)
|
8
(3%)
|
290
(79%)
|
313
(79%)
|
311
(78%)
|
952
(28%)
|
Currently on OST for ≥30 days
|
0
|
0
|
0
|
0
|
0
|
7
(3%)
|
1
(<1%)
|
1
(<1%)
|
12
(3%)
|
16
(4%)
|
21
(5%)
|
58
(2%)
|
* Refers to any of the substances listed in this table
** Heroin includes heroin/ cannabis combinations, known locally as nyaope and whoonga
CT: Cape Town; DBN: Durban; JHB: Johannesburg; PE: Port Elizabeth; PMB: Pietermaritzburg; PTA: Pretoria
|
Table 2
Substance use practices among participants (n=3 439)
In the complete cohort, 28% reported injecting a drug in the last month (952/3 439). Among PWUD/ID, close to 80% reported injecting a drug in the last month. No SWs reported having injected any drug in the last month; in this group the reported heroin use was consumed through smoking or inhalation. Injecting in the last month among MSM ranged from 3% in Johannesburg and Pretoria (8/250 and 8/236, respectively) to 9% in Cape Town (22/250). In Cape Town, proportionately more MSM last injected methamphetamine than PWUD/ID (18% (4/22) versus 6% (15/243)) and injected at a lower frequency (64% (14/22) injecting less than 4 times a day, versus 74% (179/243) injecting 4 or more times per day, respectively). A similar pattern was seen among MSM reporting injecting in the last month in Pretoria compared to male PWUD/ID counterparts, with 75% (6/8) MSM reporting to have last injected methamphetamine while 99% PWUD/ID (268/270) last injected heroin. Most MSM who injected rin Pretoria eported less frequent injecting than among male PWUD/ID with 63% (5/8) injecting less than 4 times a day, versus 73% of PWUD/ID (196/ 270) injecting 4 or more times per day.
Five percent (54/1 138) of all participants who reported having used heroin in the last month also reported taking some form of OST for the last 30 days or more. Among PWUD/ID this ranged from 4% (12/300 and 16/394) in CT and Durban, respectively, to 5% (19/378) in Pretoria, and 41% (7/17) of MSM in Cape Town.
Sexual risk practices
Sexual activity in the past month was reported in 75% (2 589/3 439) of all participants and highest among SWs (98%) (Table 3). MSM reporting sexual activity in the last month varied from 69% (169/246) in Pretoria to 79% (197/250) in Cape Town. Sexual activity was lowest among PWUD/ID; ranging from 29% (114/400) in Pretoria to 55% (217/398) in Durban. The median number of sex partners in the past week for those reporting sexual activity was one among MSM and PWUD/ID, and between two in Mthatha and eight in Pietermaritzburg among SWs.
Among SWs, almost all (96% - 100%) had engaged in penile-vaginal sex in the last week. Condom use at last penile-vaginal sex was lowest in Mthatha (34%, 81/248) and over 90% in Cape Town, Durban and Pietermaritzburg (93%, 339/384; 100%, 397/397 and 94%, 230/249, respectively). PWUD/ID reports of penile-vaginal sex in the last week was lowest in Pretoria (25%, 100/400) and highest in Durban (52%, 208/398). Among these, reported condom use at last penile-vaginal sex was 61% (61/100) and 59% (122/208) in Pretoria and Durban, respectively.
|
SWs
(n= 1 528)
|
MSM
(n= 746 )
|
PWUD/ID
(n = 1 165)
|
Total
|
|
CT
|
PE
|
Mthatha
|
DBN
|
PMB
|
CT
|
JHB
|
PTA
|
CT
|
DBN
|
PTA
|
|
N
|
384
|
248
|
248
|
399
|
249
|
250
|
250
|
246
|
367
|
398
|
400
|
3 439
|
Sexually active in the last month
|
376
(98%)
|
244
(98%)
|
245
(99%)
|
396
(99%)
|
248
(>99%)
|
197
(79%)
|
189
(76%)
|
169
(69%)
|
194
(53%)
|
217
(55%)
|
114
(29%)
|
2 589
(75%)
|
Number of people had sex with in last week, if sexually active (median, IQR)
|
5
(3 – 10)
|
3
(2 – 6)
|
2
(1 -5)
|
8
(5 – 10)
|
4
(2 – 6)
|
1
(1-1)
|
1
(0-1)
|
1
(0-1)
|
1
(1-1)
|
1
(1-1)
|
1
(1-1)
|
2
(1 -5)
|
Penile-vaginal sex in last week
|
361
(96%)
|
240
(97%)
|
246
(>99%)
|
397
(100%)
|
245
(98%)
|
8
(3%)
|
9
(4%)
|
10
(4%)
|
178
(49%)
|
208
(52%)
|
100
(25%)
|
2 002
(59%)
|
Condom used at last penile-vaginal sex among those who had penile-vaginal sex in last week
|
339
(93%)
|
147
(61%)
|
81
(34%)
|
397
(100%)
|
230
(94%)
|
6
(75%)
|
4
(44%)
|
3
(30%)
|
95
(53%)
|
122
(59%)
|
61
(61%)
|
1 485
(74%)
|
Receptive anal intercourse in last week
|
49
(13%)
|
7
(3%)
|
9
(4%)
|
82
(21%)
|
9
(4%)
|
91
(36%)
|
76
(30%)
|
81
(33%)
|
4
(1%)
|
0
|
3
(1%)
|
411
(12%)
|
Condom used at last receptive anal sex among those who had receptive anal intercourse in last week
|
41
(84%)
|
7
(100%)
|
0
|
81
(99%)
|
7
(78%)
|
53
(58%)
|
45
(59%)
|
51
(63%)
|
2
(50%)
|
0
|
3
(100%)
|
290
(71%)
|
Lubricant used at last receptive anal sex among those who had receptive anal intercourse in last week
|
30
(61%)
|
1
(14%)
|
0
|
76
(95%)
|
5
(56%)
|
71
(78%)
|
67
(88%)
|
72
(90%)
|
2
(50%)
|
0
|
1
(33%)
|
325
(80%)
|
Drugs/ goods received in exchange for sex in last month
|
40
(10%)
|
137
(55%)
|
204
(82%)
|
3
(1%)
|
26
(10%)
|
10
(4%)
|
7
(3%)
|
1
(<1%)
|
17
(5%)
|
2
(1%)
|
10
(3%)
|
457
(13%)
|
Money received in exchange for sex in last month
|
363
(95%)
|
212
(85%)
|
244
(98%)
|
388
(97%)
|
240
(96%)
|
13
(5%)
|
9
(4%)
|
3
(1%)
|
18
(5%)
|
7
(2%)
|
16
(4%)
|
1 513
(44%)
|
Alcohol or substance use at last sex
|
253
(66%)
|
124
(50%)
|
133
(54%)
|
290
(73%)
|
11
(4%)
|
65
(26%)
|
87
(35%)
|
54
(22%)
|
129
(35%)
|
5
(1%)
|
162
(41%)
|
1 313
(38%)
|
CT: Cape Town; DBN: Durban; JHB: Johannesburg; PE: Port Elizabeth; PMB: Pietermaritzburg; PTA: Pretoria
|
Table 3
Sexual risk practices (n=3 439)
Receptive anal intercourse was reported by 12% (411/3 439) of participants overall. Very few (<1%, 7/1 165) PWUD/ID reported receptive anal intercourse, among whom 4 were male. Roughly one third (33%, 248/746) of MSM reported receptive anal intercourse in the last week. Of these, 60% (149/248) reported condom use at last receptive anal sex with 85% (210/248) reporting the use of lubricant. A total of 156 SWs reported receptive anal sex in the last week, highest in Durban (21%, 82/399). Seventeen male SWs reported anal sex in the last week, all from Cape Town. Condom use at last receptive anal sex in SWs was generally high, between 78% (7/9) in Pietermaritzburg to 99% (81/82) and 100% (7/7) in Durban and Port Elizabeth, respectively. None of the nine women engaging in receptive anal sex in Mthatha reported condom use. Lubricant use at last receptive anal sex act varied from none in Mthatha to 95% (76/82) in Durban.
The 44% (1 513/3 439) of participants reporting receiving money in exchange for sex in the last month were mainly SWs; between 85% (212/248) in Port Elizabeth and 98% (244/248) in Mthatha.
Overall, 38% (1 313/3 439) of participants reported substance use at last sex. This was generally highest among SWs (53%, 811 / 1 528), who mainly reported using alcohol, with those in Durban reporting the highest prevalence (73%, 290/399). Among PWUD/ID, substance use at last sex ranged from 1% (5/400) in Pretoria to 35% (129/398) in Durban. Among MSM, substance use at last sex was more consistent varying between 22% (54/246) in Pretoria to 35% (87/250) in Johannesburg.
Access to HIV services
Overall, for 11% (373/3 439) of study participants it was the first time that they had ever had a health screen and been tested for HIV. PWUD/ID had the highest levels of first health screens and HIV tests, with the majority of their previous contacts being linked to accessing sterile injecting equipment. A quarter of all participants (25%, 867/3 439) reported to be HIV-infected at the time of participation, ranging from 2% among PWUD/ID in Cape Town and Pretoria (9/367 and 9/400, respectively) to 62% (154/249) among SWs in Pietermaritzburg. The proportion of people living with HIV reporting current antiretroviral therapy (ART) was highest among MSM recruited from MSM health clinics (93%, 255/274) and lowest among PWUD/ID recruited largely as part of community outreach activities (41%, 21/51) (See Table 4).
|
SWs
(n= 1 528)
|
MSM
(n= 746 )
|
PWUD/ID
(n = 1 165)
|
Total
|
|
CT
|
PE
|
Mthatha
|
DBN
|
PMB
|
CT
|
JHB
|
PTA
|
CT
|
DBN
|
PTA
|
|
N
|
384
|
248
|
248
|
399
|
249
|
250
|
250
|
246
|
367
|
398
|
400
|
3 439
|
First health-screen that includes HIV testing
|
19
(5%)
|
2
(1%)
|
2
(1%)
|
0
|
1
(<1%)
|
15
(6%)
|
2
(1%)
|
3
(1%)
|
58
(16%)
|
86
(22%)
|
184
(46%)
|
372
(11%)
|
Self-reported HIV +ve
|
18
(5%)
|
67
(27%)
|
109
(44%)
|
194
(49%)
|
154
(62%)
|
90
(36%)
|
110
(44%)
|
74
(30%)
|
9
(2%)
|
33
(8%)
|
9
(2%)
|
867
(25%)
|
Self-reported in care on ART, if self-reported HIV-infected
|
8
(44%)
|
54
(81%)
|
93
(85%)
|
168
(87%)
|
134
(87%)
|
87
(97%)
|
97
(88%)
|
71
(96%)
|
1
(11%)
|
12
(34%)
|
8
(89%)
|
733
(84%)
|
CT: Cape Town; DBN: Durban; JHB: Johannesburg; PE: Port Elizabeth; PMB: Pietermaritzburg; PTA: Pretoria
|
Table 4
Health care access (n= 3 439)
Testing results
Details of the testing results are provided in Table 5. HIV prevalence was 37% (1 258/3 439) overall; 47% (711/1 528) among SWs, 43% (320/746) among MSM, 19% (227/1 165) among PWUD/ID. HIV prevalence was highest among SWs in Pietermaritzburg at 80% (199/249) and lowest among PWUD/ID in Cape Town (6%, 22/367). There was considerable variation in prevalence of HIV between cities among SWs and PWUD/ID but it remained fairly similar among MSM.
HBsAg positivity was 4% (141/3 439), ranging from 2% (6/246) among MSM in Pretoria to 5% among PWUD/ID in Cape Town (20/367) and Pretoria (20/400) and SWs in Pietermaritzburg (13/249). HBsAg positivity was 1% (3/307) among people born after 1995 (the year that childhood immunisation was implemented) and 4% (138/3 132) among those born before 1995.
|
SWs
(n= 1 528)
|
MSM
(n= 746 )
|
PWUD/ID
(n = 1 165)
|
TOTAL
|
|
CT
|
PE
|
Mthatha
|
DBN
|
PMB
|
CT
|
JHB
|
PTA
|
CT
|
DBN
|
PTA
|
|
N
|
384
|
248
|
248
|
399
|
249
|
250
|
250
|
246
|
367
|
398
|
400
|
3 439
|
HIV +ve
|
69
(18%)
|
90
(38%)
|
133
(56%)
|
220
(55%)
|
199
(80%)
|
101
(40%)
|
122
(49%)
|
97
(41%)
|
22
(6%)
|
65
(16%)
|
140
(35%)
|
1 258
(37%)
|
HBsAg +ve
|
13
(3%)
|
8
(3%)
|
10
(4%)
|
17
(4%)
|
13
(5%)
|
9
(4%)
|
10
(4%)
|
6
(2%)
|
20
(5%)
|
15
(4%)
|
20
(5%)
|
141
(4%)
|
Anti-HCV +ve ELISA
|
1
(<1%)
|
0
|
0
|
0
|
0
|
16
(6%)
|
3
(1%)
|
1
(<1%)
|
136
(37%)
|
113
(28%)
|
288
(72%)
|
558
(16%)
|
HCV viral load detectable
|
0
|
0
|
0
|
0
|
0
|
15
(6%)
|
2
(1%)
|
0
|
102
(28%)
|
90
(23%)
|
227
(57%)
|
436
(13%)
|
HCV viraemic rate (among anti_HCV +ve)
|
0
|
0
|
0
|
0
|
0
|
94%
|
67%
|
0
|
75%
|
80%
|
79%
|
78%
|
Anti-HCV-HIV co-infection
|
0
|
0
|
0
|
0
|
0
|
6
(2%)
|
3
(1%)
|
1
(<1%)
|
10
(3%)
|
25
(6%)
|
113
(28%)
|
158
(5%)
|
Anti-HCV prevalence among HIV +ve participants
|
0
|
0
|
0
|
0
|
0
|
6
(6%)
|
3
(2%)
|
1
(1%)
|
10
(45%)
|
25
(38%)
|
113
(81%)
|
158
(13%)
|
Anti-HCV-HBV co-infection
|
0
|
0
|
0
|
0
|
0
|
1
(<1%)
|
0
|
0
|
5
(1%)
|
5
(1%)
|
17
(4%)
|
28
(<1%)
|
HIV-HBsAg co-infection
|
5
(1%)
|
3
(1%)
|
9
(4%)
|
13
(3%)
|
12
(5%)
|
5
(2%)
|
7
(3%)
|
3
(1%)
|
3
(1%)
|
6
(2%)
|
9
(2%)
|
75
(2%)
|
HIV-anti-HCV-HBV co-infection
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
2
(1%)
|
8
( 2%)
|
10
(<1%)
|
CT: Cape Town; DBN: Durban; JHB: Johannesburg; PE: Port Elizabeth; PMB: Pietermaritzburg; PTA: Pretoria
|
Table 5
HIV, HBsAg, anti-HCV, HCV PCR prevalence and HCV viraemic rate among participants (n=3 439)
Anti-HCV positivity was 16% (558/3 439) overall - mostly driven by the PWUD/ID sub-group. Among PWUD/ID, anti-HCV positivity ranged from 28% (113/318) in Durban to 72% (288/400) in Pretoria, with HCV viraemic rate of 80% (90/113) and 79% (227/288), respectively. Six percent (16/250) of MSM in Cape Town had detectable anti-HCV antibodies, among whom all but one had detectable HCV RNA. Only one SW was anti-HCV antibody positive but had no detectable viraemia. The overall viraemic rate among those who were anti-HCV positive was 78% (436/558).
Overall, 5% (158/3 439) were anti-HCV-HIV co-infected, which was highest among PWUD/ID in Pretoria (28%, 113/400), with no co-infection detected among SWs. 13% (158/1 258) of the HIV-infected study participants were co-infected with HCV; highest among PWUD/ID in Pretoria (81%, 113/140).
The overall prevalence of HCV-HBV co-infection was less than 1%, and 2% (75/3 439) were HIV-HBV co-infected. 10 participants were HCV-HBV-HIV triple-infected, all of whom where PWUD/ID.
Among male SWs reporting anal intercourse in the last week (all of whom were from Cape Town), 41% (7/17) were HIV infected, 6% (1/17) HBV infected, one person had HIV-HBV co-infection and no HCV infections were detected. Among male PWUD/ID reporting anal intercourse in the last week, 50% (2/4) were HCV infected, 50% (2/4) HIV infected, one person was HIV-HCV co-infected and none were HBV infected.
Of the 38 participants recruited from MSM sites reporting injecting drugs in the last month, HCV prevalence was 37% (14/38) [45% (10/22) in Cape Town, 38% (3/8) in Johannesburg and 13% (1/8) in Pretoria], HIV prevalence 37% (14/38) [23% (5/22) in Cape Town, 63% (5/8) in Johannesburg and 50% (4/8) in Pretoria)] and HBV prevalence 3% (1 infection in Cape Town).
HCV viral load on 435 samples showed a median of 5.45 log10 IU/ml (IQR 4.81 - 5.95). PWUD/ID had a higher median viral load compared to MSM, which was not statistically significant (5.38 versus 5.18 log10IU/ml, p=0.3). Genotyping was possible on 92% (400/435) of the samples with detectable viral load. Among the samples that could be genotyped, the most prevalent HCV genotypes were 1a (74%, 296/400) and 3a (14%, 56/400). Mixed genotype infections were seen in 11 PWUD/ID with most from Cape Town (6/11). Genotype 4 was found only in MSM. Of note, no genotype 5 was detected (see Table 6).
|
SW
|
MSM
|
PWUD/ID
|
Total
|
Genotype1
|
|
314
|
1
|
0
|
1
|
11
|
12
|
1a
|
0
|
12
|
284
|
296
|
1b
|
0
|
0
|
6
|
6
|
Genotype 3
|
|
71
|
3
|
0
|
1
|
11
|
12
|
3a
|
0
|
0
|
56
|
56
|
3c
|
0
|
0
|
3
|
3
|
Genotype 4
|
|
4
|
4
|
0
|
2
|
|
2
|
4a/4c/4d
|
0
|
2
|
|
2
|
Mixed genotype
|
|
11
|
1a and 3
|
0
|
0
|
1
|
1
|
1a and 3a
|
0
|
0
|
8
|
8
|
1a and 3c
|
0
|
0
|
1
|
1
|
1b and 3a
|
0
|
0
|
1
|
1
|
Total
|
0
|
18
|
382
|
400
|
Table 6
HCV genotype among participants with detectable HCV viral load (n=400)
2 At the time of study implementation TB HIV Care operated HIV prevention services for SWs (in Cape Town, Port Elizabeth, Durban, Pietermaritzburg) and PWUD/ID (in Cape Town and Durban) through mobile outreach and from community based health centres. OUT Wellbeing provided HIV prevention and treatment services to PWUD/ID and MSM in Pretoria through mobile outreach and from a community based health centre. Anova Health Institute provided HIV prevention and treatment services to MSM in Cape Town and Johannesburg from community based health centres.
3The sites which recruited sex workers routinely provided services to people who accessed services tailored towards people who identify as sex workers, but no specific distinction was made between transactional sex and sex work.
4This decision was made based on the acceptable safety profile and tolerance of HBV vaccination and insufficient funds for additional HBV testing.