November 2003

Expanding Workplace HIV Prevention Programs for a Highly Mobile Population in
Ho Chi Minh City

Key Findings

PEs contacted more workers and were better at distributing condoms than HCs. Among workers surveyed, a higher proportion of workers at PE sites were reached by the intervention than were workers at the HC sites. At six months follow-up, 73 percent of workers at PE sites compared to 57 percent at HC sites had direct experience with program activities, such as through group or one-on-one sessions. Exposure to the program reported at one year about the prior six months declined to 61 percent of the workers at PE sites and 45 percent for the HC sites. The reduction in exposure to PE between rounds two and three likely reflects the dilution of the intervention as both PEs and workers moved on to new construction sites.

Training session. Photo credit: Dr. Vu Ngoc Bao

The PEs appeared to be more successful in distributing condoms to fellow workers than the HCs. At six months, 88 percent of workers at PE sites who had contact with PEs reported that they had received condoms from them, compared to 76 percent of workers who had contact with HCs (p < .01). At 12 months, 78 percent of workers at PE sites compared to 65 percent of workers at HC sites reported the same (p < .01). Reports from HCs reveal that they distributed fewer condoms than they were provided, and qualitative data from workers suggest they were more comfortable getting condoms from PEs than HCs.

Although turnover among both groups was a problem, PEs had better retention rates than HCs, and PEs continued program activities at new work sites. An ongoing concern has been the high level of dropouts from the HC program and the need to keep recruiting and training new HCs. However, there was also a concern that many PEs would be lost when they moved to new work sites. The results indicate that there was significant turnover with both groups, although the turnover was less among PEs than HCs. Approximately two-thirds of the PEs who started the program remained after six months, compared to 55 percent of the HCs. At 12 months, 52 percent of the PEs remained, compared to 43 percent of the HCs.

One factor that may have contributed to dropout among female HCs was teasing by male workers. A few female health communicators described embarrassing exchanges with male construction workers when discussing sensitive issues. According to one HC, "Some male construction workers asked these HCs to display how to use a condom. These people felt very embarrassed and consequently quit this work." This raises the need to be sensitive to the appropriate match between educators and their audience, in particular with issues related to gender.

An important objective of the study was to explore whether PEs can continue education activities with their fellow workers as they move to different work sites, and therefore reach a highly mobile work force. Overall, peer education activities were conducted at 31 construction sites through the 12 months of the intervention, expanding from the original six sites. The great majority of PEs who moved from the original six construction sites went to work at other sites in Ho Chi Minh City managed by the companies participating in the study, and the project was able to follow up all of these peer educators. Focus group discussions with PEs revealed that they generally wished to continue activities even after moving to new sites, and often did, but it was difficult do this without the support of the labor union coordinators.

PEs’ comfort levels in discussing sexual topics and knowledge about HIV/AIDS greatly improved over time and equaled or surpassed those of HCs at follow-up. An index of comfort in discussing sexual behavior was constructed out of six items. At baseline, HCs were significantly more comfortable discussing topics related to sexual behavior than were PEs (p < .001). After six months, PEs who remained in the program were more comfortable discussing sensitive topics with workers than were HCs. This gap widened further after one year for those who remained in the program (Figure 1).

Eight general questions about HIV/AIDS (e.g., "Can a person who looks healthy be infected with HIV?") were combined into an index of HIV/AIDS knowledge. At baseline, levels of knowledge were higher among HCs than PEs (p < .05). This difference disappeared by the second round of data collection. Equivalent levels of knowledge were also found after one year.

Knowledge increased for both groups of workers, but diffusion of information to non-exposed workers was greater at the PE sites than the HC sites. At baseline, the level of knowledge about HIV/AIDS and STIs among workers was high, with a mean of more than six of eight items correct for workers at both intervention sites. At six months, the mean knowledge score increased to 7.3 for workers at both the PE and the HC sites (p < .01).

Knowledge also increased significantly for workers in the PE sites who did not have direct exposure to activities with PEs (p < .01), but this increase was not found among workers in the HC sites who were not directly exposed to health communicators (Figure 2). This is likely a result of increased communication about HIV/AIDS between exposed and unexposed workers at the PE sites.

At baseline, about two-thirds of workers at PE sites and HC sites knew where they could obtain a condom. At six months, a greater proportion of workers directly exposed to PE activities (94 percent) as well as non-exposed workers at the PE sites (75 percent) knew where to obtain a condom, and these differences were statistically significant (p < .01). At HC sites, the figure increased to only 75 percent of exposed workers who knew where to get a condom six months after the intervention began (p < .01), but for non-exposed workers there was little change.

There were statistically significant improvements in key outcomes among exposed workers in both arms, although improvements on several variables were greater at PE sites. Apart from knowledge, the HC and PE interventions aimed to increase workers’ perception of risk, confidence to make risk-reduction decisions, and confidence to use condoms correctly. Findings show that there were significant, positive changes among workers exposed to both the PE and HC interventions (p < .01).

For an additional outcome—stigmatizing attitudes toward people living with HIV and AIDS—an index was constructed from five items. The survey results show that the interventions did have an impact on reducing negative attitudes among exposed workers in both groups (p < .01); however, significantly greater reductions were observed among workers at the PE sites.

Further analysis reveals that support for risk reduction norms increased significantly only among exposed workers in the PE group. To determine support for risk reduction norms, a 15-item index was constructed. The mean score increased significantly for exposed workers in the PE sites (p < .01), but there was no comparable improvement among exposed workers at the HC sites.

There is some evidence of increased condom use, although workers report low levels of sexual risk behavior. In the surveys, reported levels of sexual risk behavior among workers were low. Averaged over the three rounds, only about half of workers reported sexual activity in the previous six months and, for the great majority, sex occurred with a spouse. Reported sex with non-marital partners varied across survey rounds but was still relatively low: In the second survey, 5 percent reported sex with a sex worker and 15 percent reported having sex with other partners in the previous six months.

Male construction workers in peer education session. Photo credit: Dr. Vu Ngoc Bao

HCs and PEs felt that more workers engage in risk behavior than was reported to the survey interviewers. This was corroborated by data from the in-depth interviews with workers. For example, one worker noted:

"I have a medium risk [of HIV] because I sometimes have sex with my girlfriend, and once in a while I visit sex workers. One time having sex with a sex worker, I did not use condoms."
Male worker, 33, divorced, PE site

Almost all of the workers who reported having sex with a sex worker at both survey rounds used a condom at last sex (12 of 14 workers at baseline, and 24 of 30 workers at six months follow-up).

Survey data show an increase in condom use with non-marital partners other than sex workers. At baseline, condom use at last sex was reported by about a third (n = 17/51) of workers, and this proportion increased to 49 percent (n = 42/86) at six months follow-up. This increase was greater among workers exposed to the PE program (71 percent; n = 17/24) than the HC program (45 percent; n = 15/33).

Findings from the qualitative data suggest a shift toward condom use among workers, particularly those reached by PEs. According to a married male worker from a PE site: "Mr. C. [PE] distributed condoms to us. As a man he understood our sexual desire. Therefore I am not embarrassed to ask him." A male PE noted:

"Construction workers living far away from home used to drink a lot on the weekend, and then go to ‘karaoke bar’ to have ‘the second shift’ or ‘the third shift’ [visit sex workers], but never used a condom. After listening to us [PEs], they carry condoms whenever they go there."

Cost per worker reached was lower for the PE program than the HC program. The total costs over one year were greater for the PE program (US$14,638) than for the HC program (US$11,374), due to greater training costs and more supervision. However, the productivity analysis, which takes into account that a greater percentage of workers at the PE sites reported being contacted (61 vs. 45 percent), shows that the cost per worker reached was lower for the PE program than the HC program. The cost per worker reached for the PE program was US$30.37 and for the HC program was US$34.16, a 16 percent lower cost per worker reached.

Female construction workers in peer education session. Photo credit: Dr. Vu Ngoc Bao

Support of management was key, and management became involved for varied reasons. An important part of the PE program was involving company management, since PE programs require permission from authorities to train their workers and to allow them time to engage in PE activities. Interviews with program coordinators indicate that the most successful way to motivate management was to work with both top management and site management. Findings from construction company managers indicate that their support of workplace HIV/AIDS programs was motivated by multiple factors. These range from concerns about the wellbeing of their workers to the desire to reduce stigma in the workplace. Having a reputable organization like the labor union address concerns about costs and time spent on the intervention appears to be an important part of a successful strategy to get management on board.

Table of Contents | Next >

	Print this page
@	E-mail this page

This page updated
19 Oct 2007