(introduction...)

Many VCT services are monitored by examining data on demand and uptake of service, number of people being tested following counselling and rates for collecting HIV results.

7.1 Uptake, return rates and acceptability

Uptake of services has often been regarded as an important measure of VCT services “success”. Uptake of VCT varies widely within and between VCT service types, communities and countries.

MTCT settings

Research settings

Uptake and return rate of VCT associated with MTCT interventions has been studied in detail. Thirteen studies from west (Abidjan, Bobo-Dioulasso), east (Addis Ababa, Dar Es Salaam, Mombassa, Nairobi,) and southern Africa (Blantyre, Durban, Harare, Lusaka, Soweto) and one from Thailand were included in a cross-sectional mailing survey about the acceptability of VCT and various MTCT interventions in antenatal clinics (Cartouxet al., 1998)²⁰⁷. All respondents represented research projects, rather than operational sites. The median overall acceptability was 65%, ranging from 33-95%. Where several studies were conducted in the same country, pregnant women had similar attitudes to HIV testing. The main reason for not wanting an HIV test was “wanting to discuss with partner”. Five studies had overall acceptability rates of >70%, with high return rates. In the five studies where all information was available 1-4% of all women offered VCT and 11-35% of all seropositive women, were finally included in the MTCT intervention projects. These results were obtained from pilot research sites. The authors note that “fighting against HIV discrimination” will be important to improve VCT acceptability and uptake.

Table 8: Uptake of VCT at MTCT sites

	Acceptance rate (%)	Return rate (%)	Return rate of HIV+	Overall acceptance
		(%) rate	(%)
Bobo Dioulasso	92	82	81	76
Abidjan	78	58	38	46
Abidjan	77	63	59	49
Nairobi	99	70	54	69
Mombasa	95	68	62	68
Dar Es Salaam	88	86	72	75
Blantyre	53	83	75	45
Blantyre	90	68	69	61
Lusaka	81	100*	100	81
Harare	100	33	36	33
Soweto	97	83	50	80
Durban	98	98	100	95
Bangkok	100	95	90	95

* 100% return rate likely to be due to simple/rapid testing employed at the site.

A study from Zimbabwe offering VCT to women at delivery or post par-tum had very poor uptake (Tavengwa et al., 2000²⁰⁸). This study offered vitamin A to mothers and/or babies in the 96 hours post delivery and offered infant feeding coun-selling to women who tested seropositive. Of 1 311 women who delivered 4 requested HIV tests at delivery and 118 (9%) between 2 and 8 weeks post partum. Even with free VCT and education and counselling on infant feeding most women choose not to learn their HIV status. This may reflect that the immediate/early post-partum period is not a practical time to offer VCT.

Operational settings

It is likely that the uptake and return rates will be different in operational settings. There is limited information available, but preliminary results from a large MTCT programme in Botswana show relatively low uptake of VCT during the first eight months of operation (Mazhaniet al., 2000²⁰⁹). This is due to fear of a seropositive result, lack of facilities where partners can receive counselling and testing, worry about partners’ reactions and the lack of effective treatment available for infected women themselves. It should not be seen as a “failing” of the project; it demonstrates that women are able to make choices on testing without coercion from counsellors. Furthermore, as the programme develops and the benefits of VCT associated with MTCT interventions (for both mothers and children) are more widely known and understood it would be expected that uptake would increase.

In Thailand VCT associated with MTCT interventions provided by the antenatal clinic is routinely offered in some provinces. In a study exploring 24 465 women attending 27 hospital antenatal clinics 99% of women accepted VCT (Koetsawang et al., 2000²¹⁰).

Uptake of VCT in other operational settings varies considerably in the UNICEF/UNAIDS MTCT pilot sites. Differences in testing schedules, maturity of the epidemic, seroprevalence in the community and attitudes to and availability of VCT in the community are thought to be important. It is also proposed that counsellors’ attitudes towards testing at the sites may be a key factor in uptake.

Table 9: Uptake of VCT at MTCT operational sites

Country	Antenatal attendees	# (%) counselled (of those attending antenatal clinic)	# (%) tested (of those who received pre-test counselling)	% HIV +ve (of those testing)	# (%) returned for test results
1. Uganda*		1111	776 (70%)	105 (13.5%)	776 (100%)
2. Botswana** (Gaborone & Francistown)	8781	5078 (58%)	2325 (46%)	933 (40%)
3. Rwanda* Kicukiro HC, Kigali	1122	1122 (100%)	781 (70%)	181 (23%)	624 (80%)
4. Cd’Ivoire** (Yopongon, Abidjan)	4309	3756 (87%)	3452 (92%)	445 (13%)	2382 (69%)
5. Cd’Ivoire	9652**		6976 (72%)	925 (13%)	422 (45.6%) (results for +ve only)
(Project Retro-CI Abidjan)	3942*		2710 (68.5)	294 (10.9%)	294 (100%) (results for +ve only)
6. Zimbabwe** (Highfield, Harare & other Harare site)	3571	621 (17%)	470 (76%)	138 (29%)
7. South Africa** Khayelitsha Mother-to-Child Pilot Project	6927		5459 (79%)	768 (14%)
8. Thailand** (27 sites)		24465	(99%)	1509 (6%)	1353 (89.6%) (results for +ve only)
9. Brazil**		(14-60%)	(40-70%)		(95%)

* Simple/rapid testing with test results given within 1 hour
** ELISA- test results given at next ANC visit
Source: UNICEF interim project report May 2000 and other published/unpublished MTCT reports.

Provisional results from the Khayelitsha Mother-to-Child Pilot Project (funded by the provincial Government of the Western Cape in South Africa) show a high rate of acceptance of HIV testing (Sigxaxhe, 2000²¹¹).

In Zimbabwe 186 women attending an antenatal clinic in Chitungwisa were offered VCT as part of their antenatal care (Martin-Hetz et al., 2000²¹²). Although most women endorsed the multiple benefits of VCT, uptake was low, with only 23% of women consenting to VCT.

MTCT interventions have recently been introduced in antenatal clinic in Lusaka, as part of routine antenatal care. Prior to this a survey among antenatal attenders showed that when VCT was offered to women at less than 28 weeks, 28 to 36 weeks and more than 36 weeks, 71%, 62% and 64%, respectively, of women said that they would agreed to be tested (Chibwesha et al., 2000²¹³). It is not known whether this proposed high uptake will be translated into action when the services are running.

VCT centres

Uptake

Uptake of VCT in communities is dependent on societal factors as well as factors associated with delivery of the service. There may also be great differences in theoretical and actual uptake rates. For example, in Lusaka when students were asked if they wished to be tested for HIV there was a very high rate of interest. When the service was provided initially, uptake was very low. However, with time, there has been increasing demand for VCT in Lusaka (Baggaley et al., 1997²¹⁴). Another study from Zambia examined the readiness to utilize VCT services offered to 4 812 participants from rural and urban sites. Although 37% initially expressed willingness to use VCT services only 3.6% actually came for VCT (Rosensvard et al., 1998²¹⁵).

Return rates

The multi-site study from Nairobi and Dar Es Salaam randomized people to either receive VCT or health information. Of those assigned VCT 95% were actually tested. 75% of those in the United Republic of Tanzania and 85% of those in Kenya returned to collect their results (Balmer et al., 1998²¹⁶). In the United Republic of Tanzania, those who were seropositive, or enrolled as couples were less likely to return for their test results. When given access to unlimited counselling sessions most people chose to attend only two sessions, though those who tested seropositive had more coun-selling sessions.

In the Central African Republic a remarkably high 98.5% of a sample of 2 800 clients came back for their HIV test result after 1 week (Sehonou et al., 1999²¹⁷).

Couple counselling

When couples can be tested together it has been shown that this can be a very successful way of preventing HIV transmission in discordant couples. Uptake and acceptance of this option has been shown to vary considerably. In Lusaka, where it was developed as part of a VCT services, it was not popular and couples were very reluctant to be tested together (Baggaley et al., 1997).

Table 10: Couple counselling uptake

#(%) couples invited	#(%) couples confirmed	#(%) couples attended for group counselling	#(%) couples who attended for couple counselling	~#(%) couples tested
1927	1310 (68%)	424 (22%)	301 (15.6%)	136 (7%)

Kara couple counselling project

In Rwanda, where couple counselling was offered as part of a research project looking at discordant couples, it was a popular option. No couple declined testing, and none declined to authorize the investigators to divulge test results to a partner (King et al., 1993²¹⁸).

The striking difference in uptake of couple counselling in these two projects may represent a difference in the approaches of the counsellors and difference in the target populations. However, the uptake of VCT services by a community is dependent on a wide range of factors, often beyond the control of the VCT service itself.

7.2 HIV testing methods: Simple-rapid (S/R) testing/same-day testing

If people are able to obtain their HIV test result within a few hours they are much more likely to receive their test result than if they have to wait one to two weeks. In the UNICEF/UNAIDS MTCT sites return rates were significantly higher in sites using S/R testing than ELISA(see table above).

Other MTCT projects using S/Rs or changing to S/R testing have shown high uptake rates:

· In Zambia, when S/R testing was offered in a pilot project at antenatal clinics, the overall acceptance was very high (81%) (Bhat et al., 1998²¹⁹). In the VCT services in Lusaka return rates were found to be higher when S/R testing replaced ELISA (Kayawe, 2000²²⁰).

· In Rwanda, HIV rapid testing was offered to all women attending the antenatal clinic in the central hospital in Kigali. Of the 1 223 women screened for HIV 68.7% returned to collect their HIV test results. Women were more likely to return if they were seronegative (71.3%) than if they were seropositive (63.9%) (Ladner et al., 1996²²¹).

· In Malawi, uptake of VCT had been low but increased four fold when simple/rapid testing was available (Msowoya et al., 2000²²²). Furthermore, 100% of clients received their test results compared with 68% when ELISAtesting was used.

· In Kenya, same-day S/R testing has also been found to be acceptable, but worries about confidentiality are still a concern (Arthur et al., 2000²²³).

· In Guatemala, introduction of S/R testing significantly increased return rates (Samayoa et al., 2000²²⁴).

· In Cd’Ivoire, although S/R testing increased the proportion of women receiving an HIV test result during the antenatal period (compared with ELISA testing), there was, overall, very poor adherence to ARV interventions to prevent MTCT (Sibailly et al., 2000²²⁵). This highlights that while uptake rates of VCT can be improved using S/R testing unless women receive adequate counselling to understand the benefits of testing it may not be translated into beneficial outcomes for the mother and child.

Box 15: Barriers to the uptake of VCT Uptake of VCT services varies greatly between settings and between countries. There are several factors that may affect uptake rates or deter people from accessing services: 1. Method of reporting/confidentiality VCT services have different methods of reporting HIV results, and this may influence uptake of the service, especially where people are worried about confidentiality or belong to groups that are already unsupported or marginalized. Many VCT services offer a confidential service where the individual and his or her counsellor share the test result. Test results may be kept by the clinic or there may be a policy of reporting seropositive to a referral centre (named reporting). The rationale for named reporting is that it is important for surveillance of HIV within communities/countries. Others VCT sites offer an anonymous service, where someone wanting a test can attend without giving his or her name. A study from the United States looked at the effect of named reporting on the uptake of VCT services in publicly funded VCT programmes, where approximately 2.5 million people are tested for HIV each year (Nakashima et al., 1998226). It was feared that the introduction of named reporting to aid surveillance would cause some individuals to avoid testing. However, there was no significant effect on the use of testing facilities following the introduction of named reporting, though in some states there was a statistically non-significant reduction in testing among African Americans and IDUs. Other studies suggest, however, that the introduction of anonymous testing increases testing in higher-risk populations, such as MSM and IDUs (Fehrs et al., 1988227, Hoxworth et al., 1994228, Hirano et al., 1994229). Other studies from the United States have also reported that ending anonymous services results in a decline in testing of vulnerable populations (Hertz-Picciotto et al., 1996230, Irwin et al., 1996231). 2. Stigma/societal factors/current events Stigma and discrimination may be important factors in the uptake of VCT in different communities. It has been postulated that political commitment to HIV prevention and care have led to less discrimination and, hence, higher demand for VCT in Uganda when compared to neighbouring countries. It has also been argued that, conversely, it is the large number of people who have been tested that is a major factor in promoting normalization and reducing stigma and discrimination associated with HIV It has been shown that a role model or valued member of the community declaring that he or she has been tested is important in reducing stigma and increasing the uptake of HIV testing. For example, when Magic Johnson announced that he had been tested and was seropositive there was a significant rise in people requesting VCT in the United States (CDC 1993232). 3. Availability of treatment and ARVs In countries where ARVs and other effective medical interventions (such as interventions to prevent MTCT) are available for people with HIV, there are considerable advantages to people with HIV being diagnosed early. This has changed attitudes of health workers as well as people who are at risk from HIV infection in industrialized countries to VCT, resulting in a greater uptake. In developing countries, the lack of ARVs and medical and social support services available for people with HIV is reported as a reason for the poor uptake of VCT (Baggaley et al., 1995)233. 4. Simple-rapid (S/R) testing/same-day testing Many studies have shown that if people are able to obtain their HIV test result within a few hours using simple/rapid technologies they are much more likely to receive their test result than if they have to wait one to two weeks. Thus, uptake rates of VCT can be improved using S/R testing. However, with S/R testing there is a possibility that people may not have adequate thinking time to make an informed and voluntary decision about testing. 5. Community mobilization and IEC Unless VCT services are promoted as part of comprehensive HIV prevention, care and support uptake is likely to be poor. Provision of adequate IEC and community mobilization is thought to be an important element in ensuring uptake of MTCT services associated with MTCT interventions (UNICEF, 2000). 6. Poor quality of services In some countries VCT services are under-utilized because the services they offer are inadequate and do not meet client needs. A study from India found that some clients who had tested at hospital sites were found to be unaware of their HIV status and had not received any follow up (Abraham et al., 1998234). The authors state that services should match the needs of the target group, including the medical and emotional needs, to maximize client uptake, participation and involvement. Poor quality of VCT services was also noted to be a considerable problem in a study from South Africa (Viljoen et al., 1998235). Only 85% of people interviewed had given consent before being tested and only 48% had had “adequate” pre-test coun-selling. The authors state that the lack of behavioural changes following the VCT services could be attributable to the poor standard of counselling observed. In Kinshasa (DRC) counsellors noted that in clinical settings pre-test coun-selling was not often done. Consequently, results were often not relayed to the client. Medical staff themselves often had an unhelpful attitude to VCT and were reluctant to give positive results to clients (Denolf et al., 2000236).