Case finding advantage of HIV rapid tests in community settings: men who have sex with men in 12 programme areas in China,2011

Abstract

We sought to describe the advantage of rapid tests over ELISA tests in community-based screening for HIV among men who have sex with men (MSM) in urban areas of China. Data of 31,406 screening tests conducted over six months in 2011 among MSM across 12 areas were analyzed to compare the differences between those receiving rapid testing and ELISA. Rapid tests accounted for 45.8% of these screening tests. The rate of being screened positive was 7.2% among rapid tests and 5.3% for ELISA tests (χ²= 49.161, p < 0.001). This advantage of rapid test in HIV case finding persisted even when socio-demographic, behavioural, screening recruitment channel and city were controlled for in logistic regression (exp[beta] = 1.42, p < 0.001, 95% CI = 1.27,1.59). MSM who received rapid tests, compared with those tested by ELISA, were less likely to use condoms during last anal sex (50.8% vs. 72.3%, χ²= 1706.146, p < 0.001), more likely to have multiple sex partners (55.7% vs. 49.5%, χ²= 238.188, p < 0.001) and less likely to have previously undergone HIV testing (38.8% vs. 54.7%, χ²= 798.476, p < 0.001). These results demonstrate the robustness of the advantage of rapid tests over traditional ELISA tests in screening for MSM with HIV infection in cooperation with community-based organizations in urban settings in China.

Keywords

HIV AIDS diagnosis rapid test point-of-care test men who have sex with men MSM homosexual men screening China

Background

In the last 10 years, China’s HIV/AIDS epidemic among men who have sex with men (MSM) has worsened.^1,2 Despite expanded behavioural interventions and HIV testing and counselling (HTC) among MSM, the HIV epidemic among MSM is still on the rise.^1,3–5 The proportion of newly reported HIV/AIDS cases attributed to MSM increased from 3% in 2006 to 14% in 2011.¹ The national HIV sentinel surveillance data show the prevalence among MSM increased from 1% in 2006 to 5% in 2011. A 2008 national survey among MSM conducted in 61 cities showed that prevalence in some areas was over 15%.⁶

Promoting HIV testing and case finding is one of the key prevention strategies for MSM in China.^7–9 It aims to reduce the number of people who are unaware of their HIV infection status, and link those infected to care and treatment. However, promoting HTC among MSM faces challenges in China. HTC is primarily organized by Centers for Disease Control and Prevention (CDCs) and hospitals across the country at different administrative levels. Most HTC is provided by voluntary counselling and testing (VCT) clinics based in local CDCs and general hospitals.⁸ MSM are expected to turn up on their own at these clinics to get tested. However, this passive testing approach is not quick enough to scale up HIV case finding among MSM. With already heavy regular workload, local CDC staff do not have enough time to conduct outreach to MSM. Inconvenient operations hours and location of local CDC VCT clinics and fear of stigma and discrimination deter MSM from using government testing services.^8,10,11 In addition, HTC provided by local CDCs and hospitals in China predominately uses the conventional testing of enzyme-linked immunosorbent assay (ELISA). Because ELISA testing usually takes days or even weeks to produce results and requires clients to visit the clinics multiple times after initial screenings to receive results, a high proportion of cases initially screened positive are often lost to follow-up for confirmatory and CD4 tests.¹²

In response to the emerging HIV epidemic among MSM in China, the Chinese Ministry of Health and the Bill & Melinda Gates Foundation established the six-year China-Gates Foundation HIV Programme in 2007 focusing on HIV prevention among MSM in 14 major Chinese cities and Hainan Province. The main interventions of the programme were to: promote HTC among MSM by partnering with community-based organizations (CBOs), improve the linkage between HIV screening and HIV confirmatory test and improve retention in HIV care and treatment for people diagnosed with HIV.¹³ To tackle the challenges faced in scaling up HTC among MSM, two primary approaches were followed in the China-Gates HIV Programme supporting CBOs to mobilize the MSM community to get tested for HIV, and promoting the use of rapid tests to screen MSM for HIV.^14–17

The key advantage over CDCs and hospitals in HIV testing intervention lies with MSM CBOs’ unique community knowledge enabling them to reach often hidden MSM. CBOs typically engage fellow MSM from the community to provide stigma-discrimination-free services at gay-friendly settings and convenient hours.¹³ Rapid tests allow service providers to give immediate post-test counselling and initiate a confirmatory testing process for those screened positive. This helps to reduce the rate of individuals who fail to receive confirmatory tests since there is no need for testers to return a few days later for test results as required by ELISA.¹⁸ Rapid tests can also be performed with minimal equipment. This advantage makes rapid tests well suited for settings with limited laboratory technicians and equipment, particularly community settings where MSM CBOs operate such as CBO offices, gay bars and bathhouses.^17,19 Thus, equipping them with rapid tests gives MSM CBOs the added advantage to increase HIV testing uptake among MSM and reduce losses to follow-up services. However, only approved medical institutions can perform HIV testing in China as required by the National Regulation for detection of HIV/AIDS.²⁰ Although some CBOs have shown capability of providing rapid testing services and are being supported by some international programmes to provide HIV rapid testing services for MSM occasionally, local CDCs have concerns on legality, confidentiality and biological safety if allowing CBOs to perform testing independently. Therefore, in the China-Gates programme, CBOs were only responsible for conducting intervention and recruitment of MSM for testing, and HIV testing was performed by CDC staff, which limited CBO’s advantage in HIV testing promotion.

Use of HIV rapid tests has been increasing gradually in China.^16,21–23 In 2009, the National Center for AIDS/STD Control and Prevention, China CDC, published the Guidelines for Detection of HIV Antibody Using Rapid Test to promote use of rapid tests.²⁴ Even with the guidelines in place, HIV screening in China is still predominately conducted through traditional ELISA methods. While evaluations have demonstrated the accuracy of rapid tests,^15,16,23 few of these studies focused on documenting the case finding advantage of rapid tests. Therefore, in order to raise awareness and promote the use of HIV rapid tests in community settings, we analyzed six months (July–December 2011) of programme data of 12 programme areas which participated in the China-Gates HIV Programme of community-based testing among MSM. We compared data between those receiving rapid tests with those receiving traditional ELISA tests. Two key objectives of this analysis were (a) to ascertain the robustness of the case finding advantage of rapid tests over traditional ELISA tests and (b) to profile rapid test clients vs. traditional ELISA test clients to illuminate on this advantage.

Methods

Study population

This is a retrospective analysis of programmatic data from the 11 Chinese cities (Beijing, Changsha, Chongqing, Guangzhou, Hangzhou, Nanjing, Qingdao, Shanghai, Tianjin, Wuhan, and Xi’an) and Hainan Province covered by the China-Gates HIV Programme. From July 2011, for six months we collected detailed socio-demographic, behavioural and HIV testing data for each MSM who was mobilized by CBOs for HTC through the support of the China-Gates HIV Programme. As all screening tests were anonymous rendering it impossible to link these tests to unique individuals. Thus, in order to minimize the problem of homogeneity in which multiple data records (i.e. HIV tests) in the dataset belong to the same individual, we limited our analysis to six months of data between July and December 2011.

Interventions among MSM by CBOs

The China-Gates HIV Programme actively promoted cooperation between local CDCs and CBOs. There were on average 4–6 MSM CBOs in each city. CBOs were responsible for conducting health education and behavioural intervention among MSM and mobilizing them to receive HIV tests. CBOs’ interventions focused on testing mobilization through web-based interventions such as gay websites and gay chat rooms using banners, popups, posting testing IEC materials or chatting with visitors. Another is through gay QQ groups (an instant messaging software service) or through Blue D (an instant communicating software used on smart phones for gay chat and dating). CBOs also directly recruited MSM for testing at gay gathering areas and entertainment establishments such as gay bars and bathhouses. Testing mobilization often involved peer education through interpersonal relationship networks. And through word of mouth, many MSM sought testing on their own at CBO offices (with technical support by local CDCs).

Referral mechanisms

Referral mechanisms were established between CBOs and local CDCs to ensure that MSM screened positive would receive HIV confirmatory tests (WB) at local CDCs. The key tool was a referral card that has three portions: (1) referral for the MSM client screened positive to take to local CDC for confirmatory tests; (2) record of the MSM client the CBO would pass to local CDC and (3) record of the MSM client kept by the CBO. This provides the basis for the China-Gates HIV Programme to track screening tests conducted by CBOs.

HIV screening services

Through several years of the China-Gates HIV Programme implementation, two primary models of HIV testing emerged in programme areas as shown in Figure 1.

Figure 1.

Illustration of HIV screening models in 12 programme areas.

Model 1: CBO + Rapid test

As a result of HIV testing intervention among MSM, HIV testing services were mainly provided in three ways. (1) MSM directly went to or were referred to CBO-operated testing points for rapid tests such as a CBO office located near popular gay venues in the city. (2) CBOs conducted outreach intervention at MSM gathering areas and provided on-site rapid tests such as in gay bars and bathhouses. In these two scenarios above, health workers from local CDCs usually worked together with CBOs at the testing point to perform HIV rapid tests themselves, or provided technical support to CBOs to conduct the rapid tests, and CBOs staff are responsible for pre- and post-test counselling and data collection; (3) MSM directly went to or were referred by CBOs to local CDCs for HIV rapid tests. As all tested in model I used rapid test, model I was referred as “CBO + Rapid Test”.

Model II: CBO + ELISA

MSM who agreed to have HIV testing through intervention by CBOs went directly to or were referred to local CDCs to receive HIV screening tests, using traditional ELISA. Model II were referred as “CBO + ELISA”.

Due to the limitation of data collection, the specific breakdowns of these testing scenarios were not captured. According to our field experience, as a result of HIV testing intervention by CBOs, MSM who went directly to CDCs or CBO operated testing points were the main testing ways, because the project cities have been conducting HIV testing promotion programmes since 2008.

HIV testing assays

As a measure to promote use of rapid tests, finger prick test kits (the Alere Determine HIV-1/2® Rapid Test, Abbott Laboratories, Chiba, Japan) were provided to project areas for 50% of their annually planned test quota. If not enough, local CDCs provided additional rapid test kits based on local situations. Types of ELISA kits used in project areas varied based on local situations.

Data collection

For each MSM who agreed to be screened for HIV, a questionnaire was also administered. The questionnaire included two components. CBOs were responsible for conducting the interviews and filled in the first component of the questionnaire and local CDCs were responsible for the second component. The first component includes socio-demographic characteristics, sexual behaviour, condom use, HTC history and the channel through which MSM were recruited to be screened for HIV. Channels captured by the questionnaire include bars, public bathhouses, Internet, parks or public toilets, and “others”. While the majority of MSM who reported the “others” recruitment channel were those who went to a CBO’s testing point on their own as a result of their previous contacts with the CBO, the remaining minority in this “others” group include other MSM who learned about CBO screening services from their friends or attended privately organized parties which CBO outreach workers might have also attended. The questionnaire was not designed to differentiate these MSM within the “others” category. The second component includes results of HIV screening, confirmatory testing, CD4 testing and notification of testing results. All information were entered into a web-based programme database on a monthly basis. The two sets of information were linked using a unique code (anonymous ID assigned to the client without the need to have the client’s ID card number) which was assigned to each questionnaire. Only when a case is confirmed positive with WB that the client ID card number would be collected by local CDCs, with the help of CBOs who screened the client positive, to facilitate the client’s access of subsequent government HIV services (CD4 tests and ART).

Data analysis

Although the China-Gates HIV Programme was implemented in 14 cities and Hainan Province, only 12 programme areas were included in this analysis. Three programme cities (Harbin, Shenyang and Kunming, n = 2694) were excluded because rapid tests were the only screening option used for the MSM programmes.

To document the case finding impact of rapid over ELISA tests, we first compared the difference in rates of MSM screened positive between these two tests (univariate) of screening tests done in the remaining 12 cities over a 6-month period from July to December 2011. Then we tested the robustness of this difference using multivariate logistic regression to control for other possible factors (screening test recruitment channel and city, socio-demographics and behaviour) that could make the rapid test’s advantage over ELISA spurious. Studies have shown that MSM visiting gay bath houses have higher risk behaviours and higher HIV prevalence than other MSM in China;^25–27 thus recruiting through bathhouses might bias the differences between two groups. Studies have also shown that risk for HIV infection among MSM might be related to age, marriage status and educational level.^4,5 Furthermore, other city-level contextual factors (such as HIV prevalence among MSM, coordination between local CDCs and CBOs, and level of support for rapid tests as an HIV screening tool) might also make the observed advantage of rapid tests spurious. Thus, in the multivariate logistic regression, the dependent variable “screening test result” was dummy coded with “1” being screened positive and “0” otherwise. Other independent variables that were included were socio-demographic (age, education, marital status and current address), behavioural (number of male sexual partners in past three months, sex with women in past three months, used condom during last anal sex, ever tested for HIV), and screening test recruitment channel and city information. Two behaviour variables (anal sex with men in past three months, always used condoms during anal sex with men in past three months) were not included in the logistic regression as they were found to be correlated with the other four behaviour variables already included in the logistic regression. Lastly, after establishing the robustness of the advantage of rapid over ELISA tests, we profiled these two sets of testers to further illuminate on the reasons behind this advantage. Differences in socio-demographic characteristics, risk behaviours, post-screen notification and referral services between MSM in model I (hereafter referred as “Rapid” group) and those in model II (hereafter referred as “ELISA” group) were compared using Chi-square tests for categorical variables (screened positive, socio-demographic, behavioural, screening test recruitment channel information).

The Institutional Review Board of the National Center for AIDS/STD Control and Prevention, China CDC, approved the project. This project intended to increase access of HTC for MSM. All HIV consent and testing procedures followed relevant Chinese national guidelines. In order to simplify HTC procedures and improve HIV testing rates, verbal consent was provided prior to receiving HTC.

Sample

Between July and December 2011, 32,155 screening tests were performed among MSM, who also filled out a questionnaire, in 12 programme cities. Incomplete questionnaires with less than 80% of items filled in were excluded from the analysis (n = 749). In total, 31,406 or 97.7% were included in the analysis. Sample descriptive statistics are shown in Table 1.

Table 1.

Sample descriptive statistics in 12 programme areas, July–December 2011 (n = 31406).

Variables	Screening tests	%
Socio-demographic
Age (years)
≤20	1530	4.9
21–30	14,546	46.3
31–40	6679	21.3
41–50	2593	8.3
≥51	1227	3.9
Unknown	4831	15.4
Education
High school attendance or less	5532	17.6
Completed high school or vocational school	9861	31.4
University attendance or higher	15,481	49.3
Unknown	532	1.7
Marital status
Single	21,162	67.4
Married	5958	19.0
Divorced or widowed	1157	3.7
Living with male partner	2240	7.1
Unknown	889	2.8
Current address
In project city	26,859	85.5
Outside project city, within project province	2462	7.8
Outside project province	1817	5.8
Unknown	268	0.9
Behaviour
Number of male sexual partners in past three months
1	12,262	39.0
2–4	13,739	43.7
≥5	2680	8.5
Unknown	2725	8.7
Sex with women in past three months
Yes	4399	14.0
No	20,794	66.2
Unknown	6213	19.8
Used condom during last anal sex
Yes	19,614	62.5
No	9592	30.5
Unknown	2200	7.0
Ever tested for HIV
Yes	14,892	47.4
No	16,514	52.6
Anal sex with men in past three months
Yes	30,611	97.5
No	763	2.4
Unknown	32	0.1
Always used condoms during anal sex with men in past three months
Yes	15,020	47.8
No	14,170	45.1
Unknown	2216	7.1
Screening test recruitment channel
Bar	4440	14.1
Bathhouse	1991	6.3
Park or public toilet	1648	5.2
Internet	6013	19.1
Other	16,937	53.9
Unknown	377	1.2
Screening test city
Beijing	7165	22.8
Changsha	1933	6.2
Chongqing	3199	10.2
Guangzhou	625	2.0
Hainan	192	0.6
Hangzhou	1255	4.0
Nanjing	2895	9.2
Qingdao	1957	6.2
Shanghai	2129	6.8
Tianjin	4357	13.9
Wuhan	2412	7.7
Xi'an	3287	10.5
Screening test type
ELISA	17,015	54.2
Rapid	14,391	45.8
Screening test result
Positive	1940	6.2
Negative	29,466	93.8

Results

Univariate and multivariable analyses

The rate of being screened positive was 7.2% among rapid tests and the corresponding rate for ELISA was 5.3% (χ²= 49.161, p < 0.001), representing a 36% advantage of rapid tests over ELISA in finding positive cases among the MSM in urban settings in China (Table 2).

Table 2.

Screening test results by screening test type in 12 programme areas, July–December 2011.

Screening test results	ELISA		Rapid		χ ²	p-Value
	(n = 17015)		(n = 14391)
	n	%	n	%
Positive	902	5.3	1038	7.2	49.161	<0.001
Negative	16,113	94.7	13,353	92.8

Consistent with the results of univariate analysis, the multivariate logistic regression results showed the case finding advantage of rapid over ELISA tests persisted even when other factors (socio-demographic, behavioural, screening test recruitment channel and city) were controlled for (odds ratio = 1.42, p < 0.001, 95% CI = 1.27,1.59), indicating a 42% (27–59%) advantage of rapid tests over ELISA in finding positive cases (Table 3).

Table 3.

Logistic regression of screening test results (positive = 1, negative = 0) on screening test type, controlling for socio-demographic and behavioural factors, screening test recruitment channel and screening test city (n = 31406).

Variables	Coefficient	Odds Ratio [exp(coefficient)]	95% Confidence Interval	p-Value
Screening test type
ELISA		1.00
Rapid	0.351	1.42	1.27, 1.59	<0.001
Socio-demographic
Age (years)
≤30		1.00
≥31	0.311	1.37	1.20, 1.55	<0.001
Unknown	0.053	1.05	0.87, 1.28	0.592
Education
High school attendance or less		1.00
Completed high school or vocational school	−0.106	0.90	0.79, 1.03	0.122
University attendance or higher	−0.475	0.62	0.54, 0.71	<0.001
Unknown	−1.455	0.23	0.13, 0.43	<0.001
Marital status			,
Living with male partner		1.00
Single	−0.037	0.96	0.80, 1.17	0.702
Married	0.031	1.03	0.82, 1.29	0.785
Divorced or widowed	0.092	1.10	0.82, 1.46	0.529
Unknown	−0.151	0.86	0.56, 1.32	0.488
Current address
Outside project province		1.00
In project city	−0.432	0.65	0.54, 0.79	<0.001
Outside project city, within project province	−0.052	0.95	0.75, 1.20	0.669
Unknown	0.012	1.01	0.64, 1.61	0.959
Behavioural
Number of male sexual partners in past three months
1		1.00
≥2	0.308	1.36	1.23, 1.51	<0.001
Unknown	−0.086	0.92	0.68, 1.24	0.570
Sex with women in past three months
Yes		1.00
No	0.218	1.24	1.06, 1.46	0.007
Unknown	0.104	1.11	0.92, 1.34	0.283
Used condom during last anal sex
Yes		1.00
No	0.385	1.47	1.32, 1.63	<0.001
Unknown	0.439	1.55	1.18, 2.04	0.002
Ever tested for HIV
Yes		1.00
No	0.204	1.23	1.11, 1.36	<0.001
Screening test recruitment channel
Other		1.00
Bar	−0.057	0.94	0.80, 1.12	0.500
Bathhouse	0.692	2.00	1.66, 2.40	<0.001
Park or public toilet	0.321	1.38	1.11, 1.71	0.004
Internet	0.623	1.86	1.62, 2.14	<0.001
Unknown	0.029	1.03	0.66, 1.61	0.901
Screening test city
Beijing		1.00		<0.001
Changsha	0.176	1.19	0.95, 1.49	0.123
Chongqing	0.714	2.04	1.74, 2.40	<0.001
Guangzhou	0.336	1.40	1.00, 1.95	0.048
Hainan	−0.807	0.45	0.21, 0.96	0.040
Hangzhou	0.107	1.11	0.88, 1.41	0.372
Nanjing	−0.475	0.62	0.50, 0.78	<0.001
Qingdao	−1.223	0.29	0.22, 0.39	<0.001
Shanghai	0.276	1.32	1.08, 1.61	0.007
Tianjin	−0.899	0.41	0.32, 0.52	<0.001
Wuhan	−0.882	0.41	0.33, 0.53	<0.001
Xi'an	−0.070	0.93	0.76, 1.14	0.500
Constant	−3.064	0.05	–	<0.001

Profile of rapid test vs. ELISA test users

Individuals in the Rapid group, in comparison with the ELISA group, were more likely to be university educated (56.6% vs. 43.1%, x² = 1384.516, p < 0.001), single (71.9% vs. 63.6%, x² = 941.510, p < 0.001) and recruited from bathhouses (9.5% vs. 3.7%, x² = 1016.110, p < 0.001) (Table 4), but less likely to consistently use condoms during anal sex with men in the last three months (38.3% vs. 55.8%, x² = 1004.424, p < 0.001), use condom during last anal sex (50.8% vs. 72.3%, x² = 1706.146, p < 0.001) and to have previously undergone HTC (38.8% vs. 54.7%, x² = 798.476, p < 0.001).

Table 4.

Socio-demographic, behavioural and screening test recruitment channel profiles by screening test type in 12 programme areas, July–December 2011.

Variable	ELISA		Rapid
	(n = 17,015)		(n = 14,391)
	n	%	n	%	χ ²	p-Value
Socio-demographic
Age (years)					1748.829	<0.001
≤20	618	3.6	912	6.3
21–30	8129	47.8	6417	44.6
31–40	4236	24.9	2443	17.0
41–50	1714	10.1	879	6.1
≥51	857	5.0	370	2.6
Unknown	1461	8.6	3370	23.4
Mean [SD]	31.6 [17.2]		29.5 [9.0]
Education					1384.516	<0.001
High school attendance or less	4027	23.7	1505	10.5
Completed high school or vocational school	5562	32.7	4299	29.9
University attendance or higher	7339	43.1	8142	56.6
Unknown	87	0.5	445	3.1
Marital status					941.510	<0.001
Single	10,819	63.6	10343	71.9
Married	3947	23.2	2011	14.0
Divorced or widowed	810	4.8	347	2.4
Living with male partner	1079	6.3	1161	8.1
Unknown	360	2.1	529	3.7
Current address					37.190	<0.001
In project city	1671	5.7	146	7.5
Outside project city, within project province	25,291	85.8	1568	80.8
Outside project province	2260	7.7	202	10.4
Unknown	244	0.8	24	1.2
Behavioural
Number of male sexual partners in past three months					238.188	<0.001
1	6829	40.1	5433	37.8
2–4	7197	42.3	6542	45.5
≥5	1219	7.2	1461	10.2
Unknown	1770	10.4	955	6.6
Used condom during last anal sex					1706.146	<0.001
Yes	12,310	72.3	7304	50.8
No	3566	21.0	6026	41.9
Unknown	1139	6.7	1061	7.4
Sex with women in past three months					885.116	<0.001
Yes	2956	17.4	1443	10.0
No	10,027	58.9	10767	74.8
Unknown	4032	23.7	2181	15.2
Ever tested for HIV					798.476	<0.001
Yes	9314	54.7	5578	38.8
No	7701	45.3	8813	61.2
Unknown	0	0.0	0	0.0
Anal sex with men in past three months					0.565	0.754
Yes	28,721	97.5	1890	97.4
No	716	2.4	47	2.4
Unknown	29	0.1	3	0.2
Always used condoms during anal sex with men in past three months					1004.424	<0.001
Yes	9502	55.8	5518	38.3
No	6347	37.3	7823	54.4
Unknown	1166	6.9	1050	7.3
Screening test recruitment channel					1016.110	<0.001
Bar	3018	17.7	1422	9.9
Bathhouse	625	3.7	1366	9.5
Park or public toilet	1126	6.6	522	3.6
Internet	2926	17.2	3087	21.5
Other	9186	54.0	7751	53.9
Unknown	134	0.8	243	1.7

Regarding the post-screen results notification and referral services, MSM screened positive in the Rapid group, in comparison to the ELISA group, were more likely to be aware of their screening results (97.4% vs. 93.3%, x² = 290.677) (Table 5). And among those screened positive, individuals in the Rapid group were also more likely than their ELISA counterparts to receive their positive screening test results on the same day (68.1% vs. 16.4%, x² = 570.349) and receive the WB confirmatory tests (81.1% vs. 64.9%, x² = 71.577). The overall rate of awareness of positive confirmatory test results among all MSM screened for the Rapid group was 4.8% (692/14,391) and for the ELISA group was 3.0% (508/17,015), giving Rapid test a 61% advantage over ELISA in finding and notifying MSM who were confirmed HIV positive.

Table 5.

Post-screen notification and referral profiles by screening test type in 12 programme areas, July–December 2011.

	ELISA		Rapid		χ ²	p-Value
Variable	n	%	n	%	χ ²	p-Value
Screening test results received by MSM screened					290.677	<0.001
Yes	15,880	93.3	14024	97.4
No	1135	6.7	367	2.6
Total (screening tests)	17,015	100	14391	100
Time between screening and screening results received by MSM among cases screened positive (days)					570.349	<0.001
0 (same day)	148	16.4	706	68.1
1–3	294	32.6	186	17.9
4–7	182	20.3	49	4.7
≥8	166	18.4	27	2.6
Unknown	111	12.3	69	6.6
Total (screened positive)	902	100	1038	100
Confirmatory tests conducted among MSM screened positives					71.577	<0.001
Yes	585	64.9	842	81.1
No	287	31.8	189	18.2
Unknown	30	3.3	7	0.7
Total (screened positive)	902	100	1038	100
Results of confirmatory tests					50.120	<0.001
Positive	525	89.7	715	84.9
Negative	5	0.9	79	9.4
Ambiguous	55	9.4	48	5.7
Total (confirmatory tests conducted)	585	100	842	100
Notifications received by MSM confirmed positive					0.000	0.983
Yes	508	96.8	692	96.8
No	17	3.2	23	3.2
Total (confirmed positive)	525	100	715	100

Discussions

The results of this study provide evidence to support that, when cooperating with CBOs to conduct HTC intervention among MSM in community settings, rapid testing is more efficient than conventional ELISA testing in finding MSM infected with HIV. Such an advantage is robust even when the potential influence of other related factors such as socio-demographic, behavioural and screen test recruitment channel and city information are accounted for. This advantage of rapid tests over ELISA in community settings stems from its ability to empower CBOs to strengthen their relationships with hard-to-reach MSM, particularly those who are at higher risk for acquiring HIV, and reduce losses to follow-up by government services such as confirmatory and CD4 tests.

As shown by the results, MSM tested by rapid tests are more likely to have risky sexual behaviours and be HIV positive. This may be due to the easy use of the rapid test for HIV screening in non-clinical settings. As reflected by the results, there was a higher proportion of MSM recruited from bathhouses in the Rapid vs the ELISA group. This provides an opportunity for CBOs to provide HTC with the support of CDCs, which may attract more MSM with higher risk behaviours who are reluctant to go to local CDCs for HIV testing. Previous studies have shown that MSM visiting gay bathhouses have higher risk behaviours and higher HIV prevalence than the others in China.^25–27 However, after controlling for screening test recruiting channel in the logistic regression, MSM with higher risk sexual behaviours still are more likely to use rapid test for HIV screening than traditional ELISA. This may be related to the streamlined process of rapid tests (no blood draw is needed and results are provided immediately after the test) and friendly services by CBOs, which may attract higher-risk MSM for HIV testing.

Rapid tests also help access MSM who have not been tested previously. One reason might be that rapid tests were used in outreach intervention in non-clinical settings, such as gay bars and bathhouses. This increases access to those unwilling or uncomfortable having HTC in clinical settings.^17,28 A study in the USA showed similar results that rapid testing helps to attract more MSM to receive HTC in bathhouses.²⁸ In addition, participation of CBOs may help to approach hard-to-reach MSM.¹⁹ MSM who have not been previously tested may be more willing to test if there are friendly counselling, testing and follow-up services provided by CBOs.

A loss to follow-up services such as confirmatory testing after being screened positive decreases HIV testing efficiency. Nearly 20% of MSM who screen positive in China do not receive a WB confirmatory test.¹³ Unlike traditional ELISA methods, rapid testing allows informing results and counselling within 30 minutes immediately after the test, which greatly decreases the possibility of the individuals tested not returning for test results. Our results show that a much higher proportion of screened positives received test results and confirmatory testing in the Rapid group, which is similar to several studies in the USA.^28,29

However, results showed that 25% of MSM who received rapid tests did not receive same day test results. The high proportion is unexpected since rapid testing service providers are required to provide post-test counselling after the test for both positives and negatives. According to our communication with local service providers, when local CDCs are busy, some staff will collect intravenous blood and conduct rapid tests at a later time. This reflects a limitation of the adoption of rapid testing by local CDCs. On the other hand, this reflects the urgency of accelerating the training of CBOs to perform rapid HIV testing instead of the requirement that only medically-approved institution staff are allowed to perform rapid HIV testing.²⁰ CBOs, with more flexibility and community knowledge of MSM who are often not comfortable with traditional public health clinics, are better positioned than CDCs in case finding, and can be empowered when they are equipped with rapid testing technology.^15,19,30 Further evaluation and policy advocacy on CBO’s participation in HIV rapid testing are needed.

Expanding use of HIV rapid testing is helpful to achieve China’s goals of expanding HTC and ART treatment. Rapid tests can help simplify the HIV testing procedure from screening to CD4 test. Currently in China, the standard procedure for screening, diagnosis and treatment of HIV is complex, requiring several patient-provider encounters at different medical and public health institutions. HIV screening tests are usually performed in county CDCs or hospitals using ELISA. Results take several days to a week – in our study only 16% of MSM received same day results from ELISA. If positive, the individual is contacted and required to come back to the clinic for another blood draw. Based on the current protocol, a return for second blood draw is compulsory in order to make sure the second blood sample belonged to the screened positive. The second sample is then sent to a centralized laboratory at the prefecture CDC, which conducts a repeat ELISA and, if positive, a confirmatory WB test. At the prefecture CDC, blood samples for confirmatory testing are batched for processing and results typically take 10–15 days. By using rapid tests, blood samples for WB confirmatory test and CD4 test can be drawn on site immediately after the post-screen counselling. This can greatly shorten the time from screening to confirmatory test and reduce the number of patient-provider encounters.

As a retrospective analysis of programmatic data, this study has limitations. Although all HIV testing among MSM were through the efforts of CBO’s intervention, the disparity of CBO’s development and capacity across project cities may affect MSM’s willingness to access HTC at CDCs or at CBO-operated testing points. In addition, since data analyzed covered a period of six months, some MSM may have participated in HIV testing more than once. However, based on an analysis in 2011, the average HIV testing frequency was less than twice a year (unpublished report). Therefore, considering over 52% of MSM receiving HTC for the first time, we assume that only a small proportion of MSM may have multiple HIV tests over the six-month period. In addition, as all data were collected in metropolitan cities, results may be not applicable to rural areas of China. Furthermore, due to limitation of data collection, differences between MSM who received HIV testing at CDCs by rapid or ELISA testing cannot be analyzed. This will be further explored to provide extra evidence for the advantage of rapid testing in improving HIV case finding.

In light of the rising HIV epidemic among MSM and China’s key HIV prevention strategy among MSM being testing and treatment expansion, promoting CBOs’ conducting rapid tests to screen for HIV in community settings instead of limiting CBOs to recruiting MSM to get tested at local CDCs through the traditional ELISA tests can help promote HIV case finding and reduce loss of follow-up for screened positives.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This programme was funded by the China-Gates Foundation HIV Cooperation Program (Opp49277). We acknowledge the significant contribution of project officers and community-based organizations in 12 project areas for data collection and reporting. We would like to thank Nora Kleinman of the US CDC Global AIDS Program – China Office and Peter Mok, an independent consultant of the Gates Foundation, for their review of the manuscript.

References

China Ministry of Health, UNAIDS and WHO. 2011 Estimate of China's ADIS epidemic. Chin J AIDS STD 2012; 18: 1–5.

. China facing new situation and challenges for AIDS response. Chin J Public Health 2011; 27: 1505–1507.

Wong

Huang

Wang

. STIs and HIV among men having sex with men in China: a ticking time bomb? AIDS Educ Prev 2009; 21: 430–446.

Zhang

. Internet use and risk behaviours: an online survey of visitors to three gay websites in China. Sex Transm Infect 2007; 83: 571–576.

Zhang

. Changes in HIV prevalence and sexual behavior among men who have sex with men in a northern Chinese city: 2002–2006. J Infect 2007; 55: 456–463.

Liu

. HIV and syphilis prevalence among men who have sex with men: a cross-sectional survey of 61 cities in China. Clin Infect Dis 2013; 57: 298–309.

Granich

Gilks

Dye

. Universal voluntary HIV testing with immediate antiretroviral therapy as a strategy for elimination of HIV transmission: a mathematical model. Lancet 2009; 373: 48–57.

Sun

Sullivan

. HIV testing in China. Science 2006; 312: 1475–1476.

Raymond

. Untested and undiagnosed: barriers to HIV testing among men who have sex with men, Beijing, China. Sex Transm Infect 2012; 88: 187–193.

10.

Hesketh

Duo

. Attitudes to HIV and HIV testing in high prevalence areas of China: informing the introduction of voluntary counselling and testing programmes. Sex Transmit Infect 2005; 81: 108–112.

11.

Choi

Lui

Guo

. Lack of HIV testing and awareness of HIV infection among men who have sex with men, Beijing, China. AIDS Edu Prevent 2006; 18: 33–43.

12.

Zou

Xin

. HIV testing among men who have sex with men in China: a systematic review and meta-analysis. AIDS Behav 2012; 16: 1717–1728.

13.

Zhang

Han

. The impact of community based organizations in HIV testing mobilization among men who have sex with men. Chin J Prev Med 2013; 47: 431–434.

14.

Molesworth

Ndhlovu

Banda

. High accuracy of home-based community rapid HIV testing in rural Malawi. J Acquir Immune Defic Syndr 2010; 55: 625–630.

15.

Plate

. Evaluation and implementation of rapid HIV tests: the experience in 11 African countries. AIDS Res Hum Retrovirus 2007; 23: 1491–1498.

16.

Zou

Chen

Zhang

. Effect on HIV detection for four different HIV rapid kits. Lab Med Clinic 2010; 2010: 93–93.

17.

World Health Organization. Rapid HIV tests: Guidelines for use in HIV testing and counseling services in resource-constrained settings, Geneva: World Health Organization, 2004.

18.

Guenter

Greer

Barbara

. Rapid point-of-care HIV testing in community-based anonymous testing program: a valuable alternative to conventional testing. AIDS Patient Care STDS 2008; 22: 195–204.

19.

Clark

Bowles

Song

. Implementation of rapid HIV testing programs in community and outreach settings: perspectives from staff at eight community-based organizations in seven U.S. cities. Public Health Rep 2008; 3: 86–93.

20.

China National Centers for AIDS/STD Control and Prevention. National Guideline for detection of HIV/AIDS. 2009.

21.

Han

Feng

Jiang

. Application of rapid test, combined HIV antigen/antibody EIA and pooled PCR in the detection of HIV-1 among MSM. Int J Lab Med 2011; 32: 1185–1186.

22.

Tian

Jiang

Sun

. Evaluation of rapid HIV testing strategy in HIV voluntary counselling and testing sites. Chin J AIDS STD 2008; 14: 220–222.

23.

Dange

. The application of the on-site rapid detection of HIV antibody among MSM in public bathroom. Zhejing Prevent Med 2011; 23: 6–9.

24.

Chinese Center for Disease Control and Prevention. Guidelines for detection of HIV antibody using rapid test. Chinese Center for Disease Control and Prevention: Beijing, 2009.

25.

Cheng

Yang

. HIV prevalence and associated factors among gay bathhouse attendees in China. Chinese J AIDS & STD 2003; 19: 127–130.

26.

. An Investigation on the Infection Rate of Syphilis and HIV among MSM Population from Public Bath. Zhejiang J Prevent Med 2011; 23: 14–16.

27.

Zeng

Shandan

. Investigation on correlated attitude and behavior on AIDS among man having sex with man. Chinese J Public Health 2007; 23: 1422–1423.

28.

Huebner

Binson

Dilworth

. Rapid vs. standard HIV testing in bathhouses: what is gained and lost? AIDS Behav 2010; 14: 688–696.

29.

Keenan

. Rapid HIV testing in urban outreach: a strategy for improving posttest counseling rates. AIDS Educ Prevent 2001; 13: 541–550.

30.

Shrestha

Sansom

Schulden

. Costs and effectiveness of finding new HIV diagnoses by using rapid testing in transgender communities. AIDS Educ Prevent 2011; 23: 49–57.