- Top of page
Objective To evaluate the efficacy of voluntary counselling and testing (VCT) for HIV/AIDS in changing risky sexual behaviour in central Mozambique.
Method Longitudinal cohort study of men and women aged at least 18 years from October 2002 to June 2003. We interviewed 622 participants in VCT groups and 598 in non-VCT groups. The interviews occurred before counselling and 4 and 6 months afterwards.
Results Reported use of condoms while having sex with a friends/prostitute increased over each time period in the VCT group and between baseline and first visit in the non-VCT group. Both men and women in the VCT group increased their condom use over time, but the women in the non-VCT group did not. Reported always/sometimes use of condoms for both literate and illiterate subjects was higher and rose over time in the VCT group.
Conclusion People who undergo voluntary counselling and testing fro HIV/AIDS change their behaviour, presumably as a result of their counselling.
Objectif Evaluer l'efficacité de la consultation et du test volontaire pour le VIH/SIDA dans le changement du comportement sexuel à risque dans le centre du Mozambique.
Méthode Une étude longitudinale de cohorte d'hommes et de femmes agés d'au moins 18 ans, d'octobre à juin 2003. Nous avons interviewé 622 participants au sein de groupes de consultation et test volontaires (CTV) et 598 participants sans CTV. Les interviews ont eu lieu avant la consultation puis à quatre mois et à six mois après la consultation.
Résultats Le nombre de reports d'utilisation de préservatifs lors de rapport sexuel avec un(e) ami(e) ou avec une prostituée augmentait au cours de chacune des périodes dans le groupe avec CTV et entre le début et la première visite dans le groupe sans CTV. Hommes comme femmes dans le groupe avec CTV ont augmenté leur utilisation de préservatifs au cours du temps. Mais cela n’était pas le cas chez les femmes dans le groupe sans CTV. L‘utilisation «toujours ou parfois» de préservatifs était plus élevée et a augmenté au cours du temps dans le groupe avec CTV autant chez les pesonnes lettrées que chez les illettrées.
Conclusion Les personnes qui prennent part aux consultations et tests volontaires pour le VIH/SIDA changent leur comportement sexuel, grâce probablement à la consultation.
Objetivo Evaluar la eficacia del Aconsejamiento Voluntario y Prueba (AVP) de VIH/SIDA en el cambio del comportamiento sexual de riesgo, en la región central de Mozambique.
Métodos Estudio de cohorte longitudinal de hombres y mujeres de al menos 18 años de edad, entre octubre del 2002 y Junio del 2003. Entrevistamos 622 participantes en el grupo de Aconsejamiento Voluntario y Prueba (AVP) y 598 en el grupo de No-AVP. Las entrevistas ocurrieron antes del aconsejamiento y 4 y 6 meses después del mismo.
Resultados El uso de preservativos en contactos sexuales con amistades/prostitutas aumentó en cada periodo de tiempo en el grupo AVP, y entre el valor basal y la primera visita en el grupo No-AVP. Tanto hombres como mujeres en el grupo AVP aumentaron su uso de preservativos en el tiempo, pero no en las mujeres del grupo No-AVP. El reporte de uso de preservativos siempre/algunas veces fue más alto y aumentó en el tiempo en el grupo AVP tanto para sujetos alfabetos como para analfabetos
Conclusión Las personas que participaron en el aconsejamiento voluntario y examen de VIH/SIDA cambian su comportamiento, presumiblemente como resultado del aconsejamiento.
- Top of page
Voluntary counselling and testing (VCT) for HIV has become an integral part of HIV prevention and care programs in many countries (Pronyk et al. 2002; Chimzizi et al. 2004; Liechty 2005). Voluntary counselling and testing is usually set up as a separate service, where an individual receives counselling about being tested for HIV, enabling him or her to make an informed choice. The person undergoes the HIV test if they choose, and is counselled afterwards according to the result. Voluntary counselling and testing is also the entry point for comprehensive, long-term medical care and for psychosocial, legal and material support after a positive HIV test. Health care workers advocate VCT in the hope that seropositive individuals will change their behaviour to prevent transmission of HIV and that seronegative individuals will extend their protective behaviour. But it is difficult to access behaviour change in these populations, and further studies are needed (Kawichai et al. 2004; Magnussen et al. 2004). In several countries in Africa, VCT has proven to be effective in changing risk behaviour among HIV seropositive persons (Muller et al. 1995; Allen et al. 2003).
In Mozambique, with a population of 19 million, the overall HIV seroprevalence among adults is estimated between 10% and 14%, reaching 20% along the central corridor of Beira and in Tete Province. In 2002, about 1.2 million Mozambicans were HIV positive; 2 90 000 of them had AIDS (Ministry of Health Mozambique 2000). The vast majority have not been tested due to limited access to VCT, and even where it is available people are afraid to take it up because of the stigma attached to a positive test result.
Thirty VCT centres existed in Mozambique as of 2001, most of them located in the provincial capitals of Maputo, Beira, Chimoio, Nampula and other key cities. Since 2001, another five VCT centres have been established at health facilities in Beira and Chimoio, which are run by the Ministry of Health and supported by NGOs. These centres have been very well attended in the 2 years since their inauguration, each providing HIV testing and counselling services to 250–600 clients per month.
Little information is available about sustained effectiveness of VCT for both HIV positive and HIV negative people. Our aim was to determine the effectiveness of VCT for HIV-infected and uninfected people in a poor country, using the example of Mozambique.
- Top of page
The study was designed as a longitudinal cohort study among men and women aged at least 18 years, running for 9 months, from October 2002 to June 2003. The study was conducted in two cities in central Mozambique: Beira (population 5 45 000; HIV prevalence 20%) and Chimoio (population 2 50 000; HIV prevalence 22%). Both cities have VCT centres and the Ministry of Health and NGOs conducted a mass-media campaign to popularise VCT there.
Potential participants had to be at least 18 years old and plan to remain in the same area during the period of the study. Each consenting subject was asked to participate in three interviews: at baseline, after 4 months and after 6 months. Every client attending VCT was asked by interviewers if he/she would agree to participate in the study. The interviewers were blinded to the HIV results. Controls attended the regular outpatient clinics, such as general medical and antenatal care, in the same cities for any other medical reasons in the study period. Verbal informed consent was given by all participants because of low literacy rates. The survey was designed to provide face-to-face contact with interviewers before the counselling and testing sessions began. We designed a questionnaire to collect demographic and social information and also details about sexual activities.
The baseline interviews took place before the first counselling sessions. All interviews were conducted in a private section of the waiting room. A cash payment of approximately $5.00 was offered to participants at each follow-up interview. Confidentiality was assured by coding; the interviewers were blinded to HIV test results.
Data were analysed using Epi-info 2002 (CDC 2002) and Microsoft ACCESS (Microsoft Corporation). Bivariate analysis was done using chi-square.
The study was approved by the Human Subjects Division at University of Washington in Seattle, USA, and by the National Institute of Health, Maputo, Mozambique.
- Top of page
Between October 2002 and June 2003, 622 study participants attended individual VCT in Beira and Chimoio. A 598 controls attended the outpatient clinics for general medical and antenatal care in the same cities. Table 1 shows the basic socio-demographic attributes of the study population. At baseline, there was no significant difference between the VCT and non-VCT group with respect to gender, marital status, education and parenthood. The VCT group was significantly younger than the control group, more likely to be non-Christian, and less likely to be employed.
Table 1. Socio-demographic characteristics of VCT and non-VCT groups
|Characteristic||VCT, n (%)||non-VCT, n (%)||P value|
|Participants||622 (51)||598 (49)||0.49|
|Gender/Male||322 (52)||297 (48)||0.35|
| <21||262 (43)||186 (31)||<0.01|
| 21–40||317 (51)||380 (64)|| |
| >40||37 (6)||32 (5)|| |
| Catholic||216 (36)||267 (45)||<0.01|
| Protestant||82 (14)||170 (28)|| |
| Other||270 (50)||156 (14)|| |
| Single||299 (49)||289 (48)||0.79|
| Married||178 (29)||184 (31)||0.71|
| Non-marital union||109 (18)||115 (19)||0.88|
| Widow (er) ship/divorced||28 (4)||9 (2)||0.56|
| No education||39 (6)||33 (6)||0.12|
| Complete primary school||278 (45)||249 (42)|| |
| Complete secondary||228 (37)||265 (44)|| |
| Post-secondary||60 (12)||50 (8)|| |
| Currently employed||101 (16)||173 (29)||<0.01|
| Have children||318 (50)||250 (46)||0.76|
Drop out rates between the three visits for the VCT group were 16% (102/622) from baseline to the first follow up visit, and 15%(70/520) from the first to the second follow-up visit. The drop out rate in the non-VCT groups was 10% (66/598) between baseline and the first follow-up visit, and 5% (28/532) between the first and second follow-up visit. By the end of the study, the VCT group numbered 450 people and the non-VCT group numbered 504 people.
We differentiated between two types of sexual partners: friends and prostitutes. A friend is a person who agrees to have sex without immediate pay, but will receive some gift later. A prostitute demands payment at the time of the sexual encounter. The reported use of condoms while having sex with either a friend or a prostitute increased over each time period in the VCT group and between baseline and first return for the non-VCT group. By the second follow-up visit the VCT group reported significantly higher condom use during these casual encounters than the control group. Reported condom use ‘always’ or ‘sometimes’ was not significantly different between the VCT and non-VCT groups at baseline, but was significantly higher in the VCT group at both follow-up visits, and the difference increased from the first to the second follow-up visit. Condom use at the most recent sex act was also the same in both groups at baseline but became significantly more frequent in the VCT group at both follow-up visits (Table 2).
Table 2. Reported sexual practices among voluntary counseling and testing (VCT) and non-VCT groups
|Characteristic||VCT n (%)||non-VCT n (%)||P value|
|Sex with friend/prostitute in last month|
| Baseline||181 (57)||156 (42)||0.01|
| First return||157 (55)||160 (42)||0.02|
| Second return||135 (64)||153 (50)||0.02|
|Condom in last sex among those having sex with friends/prostitutes in last month|
| Baseline||105 (58)||92 (59)||0.93|
| First return||124 (79)||124 (78)||0.87|
| Second return||117 (88)||111 (73)||0.01|
|Always or sometime use condom|
| Baseline||328 (54)||295 (50)||0.55|
| First return||320 (60)||276 (52)||0.05|
| Second return||263 (76)||254 (63)||0.01|
|Condom last sex|
| Baseline||191 (30)||169 (29)||<0.86|
| First return||348 (65)||253 (47)||0.01|
| Second return||248 (71)||188 (47)||<0.01|
In the VCT group both men and women used condoms more often over time. In the non-VCT group women did not, and men used condoms only slightly more often (Figure 1). Reported always/sometimes use of condoms for both literate and illiterate subjects was higher in the VCT group then the non-VCT group and increased over time in the VCT group (Figure 2). Condom use during the most recent sexual act increased over time in both HIV+ and HIV− groups, but the increase was greater in the HIV+ group (Figure 3). The difference from baseline is statistically significant in both groups.
Figure 1. Voluntary counselling and testing and condom use (always/sometimes) for men and women over the study period.
Download figure to PowerPoint
- Top of page
In our study people who chose to participate in VCT, compared to people from a general clinic population, reported to use condoms significantly more often after VCT, although reported condom use increased consistently to each follow-up visit in both groups. Only at the second return visit did VCT attendees report more frequent use of condoms during sex with a friend or prostitute than the control group. In the VCT group condom use increased over time in both HIV-negative and HIV-positive subjects, but the increase at the second follow-up visit was greater in HIV-positive people.
Several studies on VCT and behaviour change show risk reduction in the study populations: prostitutes in the Gambia reduced their risky behaviour after VCT (Pickering et al. 1993), as did couples in Rwanda (Van der Straten et al. 1995). In San Francisco counselling could prevent HIV transmission (Padian et al. 1993). Voluntary counselling and testing prompted sustained (albeit imperfect) condom use in 963 cohabiting heterosexual couples with discordant HIV status in Lusaka (Allen et al. 2003). In a 4 years observational study in Bangui (Gresenguet et al. 2002) VCT proved helpful in reducing unprotected sex. Similarly, a retrospective study in Cambodia found that VCT may change the risky behaviour of attendees (Kruy et al. 2001). Weinhardt et al. (1999) in a meta-analysis of 27 published studies which assessed behaviour before and after counselling and testing found that HIV-positive participants used condoms more often than untested participants.
Our study suggests the same in the central Mozambican cities of Beira and Chimoio: condom use increased significantly after VCT attendance, although this increase was modest in the HIV positive population. We observed negative or weak associations, or none, between VCT and reported sex with casual partners or use of condoms during those encounters. Behaviour change was greatest in the illiterate subcategory of the VCT group: a jump of always/sometimes use of condoms from 10% at baseline to 64% at the final visit in the illiterate VCT group compared to 21% at baseline and 17% at the final visit in the illiterate non-VCT group. This finding is an important indication of the value of interpersonal communication for illiterates in changing behaviours in sensitive areas such as risky sexual activities. The reported sexual behaviour of people in central Mozambique changed over the time of the study. People in Beira and Chimoio reported more frequent condom use at follow-up visits than at baseline, independently of VCT use. Whether this reflects a true change or reporting differences is not clear. Within the VCT group HIV+ participants seem to have made greater and more positive behaviour changes than HIV− participants.
The reported prevalence of sex with friends and prostitutes seems to have increased in both arms over the 6 months of the study. Because of the similarity of these increases it seems likely that this trend reflects more honest reporting rather than a real change over time. However, more than half of all participants reported having sex in the previous month with a non-primary partner. This very high rate suggests that special efforts need to be made to point out the risk of HIV infection to those who engage in unprotected casual sex.
The study had several limitations. The two groups were not entirely comparable. The VCT group had more pregnant women, was younger, had more non-Christians, and was more likely to be unemployed than the non-VCT group. Being younger, the VCT group may engage in higher risk behaviour than controls. In addition, the VCT group was self-selected, perhaps as a consequence of having different perspectives or risk taking behaviours. The differences between the VCT group and others are also reported by other studies (Kalichman & Simbayi 2003; Adewole & Lawoyin 2004).
There was likely measurement error in this study as well. The questions were sensitive and subjective, and it is well known that people are less than honest about sexual behaviour or condom use. However, misclassification would have been same in both groups at entry into the study. Under reporting of risky sexual behaviour and over reporting of safer behaviours may have been greater over time after VCT, as participants may have felt more pressure to give the ‘right’ answer. It was interesting that the VCT group reported more sexual relations with risky sexual partners, but also claimed more frequent condom use with them. Both groups reported increased risk behaviour in terms of having sex with friends or prostitutes over the three time periods of the study, perhaps indicating a greater level of ease and willingness to speak honestly.
Sample size was not a limiting problem overall, but when broken down by gender, site, or HIV status the sample size was small for detecting significant differences. Follow-up rates were less than hoped, but not significantly different between the study and comparison groups. Those not followed up may have had more risky sexual behaviour than those followed up, but it is not possible to know that from this study.
This study suggests the need for the Mozambiquan government to make VCT accessible to the general population wherever possible, especially in areas where HIV prevalence is high or where the population is particularly vulnerable. In the interests of the illiterate and of women, who are seen as especially vulnerable, VCT implementation in larger population groups is highly desirable.