Both injecting drug use and HIV among injecting drug users (IDUs) have become major international public health problems. HIV infection has been reported among IDUs in over 100 countries [1]. Travel by IDUs, particularly along drug distribution routes, appears to be a major mechanism for the spread of both injecting drug use and HIV among IDUs. HIV spread north [2] and south [3] from New York City along the East Coast in the U.S. Stimson [4], and Beyrer and colleagues [5] have reconstructed the spread of HIV among IDUs in South East Asia. Beyrer et al. used molecular epidemiology (mapping the different subtypes of HIV) in their reconstruction. A recent study by Kato and colleagues (2001) found patterns of HIV genetic subtyping consistent with cross-border transmission either from Vietnam to China or from China to Vietnam.

While these regional and country level analyses have great value in understanding the worldwide spread of HIV among IDUs, they have important limitations with respect to HIV prevention efforts. Reducing HIV spread by attempting to disrupt regional and country drug distribution routes may have the unintended consequence of displacing distribution to new routes, leading to additional spread of injecting drug use and HIV among IDUs.

Successful prevention efforts will be greatly facilitated by more detailed understanding of the spread of injecting drug use and transmission of HIV among IDUs within smaller geographic areas. Understanding of HIV transmission across international borders is particularly important, as few HIV prevention programs are coordinated across such borders. We present here data on injecting drug use and HIV among IDUs in the adjacent border provinces of Lang Son, Vietnam and Guangxi, China. HIV among IDUs was noted in this area in 1996 [6] and since then there has been substantial transmission among IDUs in both provinces. The present situation shows a clear geographic pattern, with the potential for additional spread across the border between the provinces and within each of the provinces.

The data reported here were collected as part of baseline surveys of IDUs conducted before implementation of a cross-border HIV prevention intervention in Lang Son Province, Vietnam and-Ning Ming County, Guangxi Province, Vietnam [7]. Figure 1 shows a map of the area, with the project sites – in Lang Son Province and in Ning Ming County.

There is considerable official and entirely legal movement of commercial goods across the border in both directions. There are also semi-official and informal crossing points and pathways through the hills that permit local residents to cross the border with little or no regulation. Crossing the border is a regular aspect of life for many people who live near the border, for example, to attend market days in the larger villages. Drug dealers cross the border to sell drugs and drug users also cross to obtain drugs of higher purity and at better prices. The frequency of border crossing can vary, influenced by current price and purity of heroin, the ebb and flow of law enforcement activity and, factors such as the outbreak of severe acute respiratory syndrome (SARS) in China.

There is also substantial trade and migratory employment in the region. Lang Son City, Aidian, and Puzhai (near Pingxiang) are bustling centers of legitimate cross-border trade, as well as drug trafficking and sex work. Many people cross the border daily and seasonally to find work and many are employed as porters in the cross-border trade. The region is home to many ethnic minority groups (e.g., Zhuang, Tay, Nung), some of whom live on both sides of the border (for example, the Zhuang in China and the Tay in Vietnam are the same ethnic group but are known by different names in the two countries). There is frequent intermarriage across the border, although this might be illegal and might lead to loss of nationality. Kinship ties, like migratory employment and trade, result in additional cross-border movement.

Data collection methods for the IDU surveys reported on here were essentially parallel in Ning Ming County and Lang Son Province, with some variation in the community-based subject recruitment strategies used. The availability of large known drug use gathering places and of officially registered IDUs in Lang Son permitted greater use of probability-based methods in Vietnam, while there was more reliance on peer recruitment in China.

Figure 2 shows the relationship between mean number of years injecting drugs and the site's distance from the border. There is a general gradient of decreasing mean years of drug injecting running from the site farthest from the border on the Vietnamese side (Lang Son City) to the site farthest from the border on the Chinese side (Ning Ming City). This gradient was statistically significant, p = -0.002. The estimated slope was -0.02, so the average number of years of injection use decreased by about a week (0.02 of a year) for each kilometer in the Chinese direction. This was based on a weighted linear regression model fitted to the mean number of years used in a site and distance to the border of the site. The weights were based on the sample size at each site.

Figure 3 shows the HIV prevalence by distance to the border for the sites in Vietnam and China for the baseline data. There is a strong gradient of decreasing prevalence from the largest Vietnamese site and the one farthest from the border (Lang Son City) to the largest Chinese site and the one farthest from the border (Ning Ming City). Only one site- (Tan Thanh in Vietnam) does not fall along this gradient. Overall baseline seroprevalence among IDUs was 46% in Lang Son and in 17% Ning Ming. This gradient was statistically significant, p < 0.001. The estimated odds ratio was 0.97, so the odds of an individual being HIV positive decreased by 2–3% for each kilometer in the Chinese direction. This relationship between the proportion of subjects in a site testing positive and distance to the border was modeled using logistic regression.

Table 2 shows factors related to crossing the border (in the 6 months prior to the interview) among the subjects in Vietnam and China. Overall, 10.6% of subjects in China and 24.5% of subjects in Vietnam reported having crossed the border in the six months prior to the interview. Being a member of an ethnic minority group and living closer to the border were strongly related to crossing the border in both countries. In Vietnam, being HIV seropositive was negatively related to crossing the border, while in China being HIV seropositive was positively related to crossing the border (although this relationship did not reach statistical significance, p = .12). Being a member of an ethnic minority group was related to being HIV seropostive in China (19% HIV seropositive among ethnic minority subjects versus 10% seropositive among ethnic majority subjects, (p < 0.05). Ethnicity was not related to HIV status in among the Vietnamese subjects.

In this report, we present data on injecting drug use and HIV infection among IDUs in Lang Son province, Vietnam and Ning Ming County, Guangxi Province, China from a baseline survey conducted before implementation of a peer-based cross-border HIV prevention intervention.

Several limitations should be noted. First, we were working with cross-sectional data, where longitudinal data from the initial spread of injecting drug use in the area would have certainly been preferable. Second, there are measurement and sample size limitations. It would have been helpful to have a measure of ease of travel to the border rather than simple physical distance to the border. Also it would have been helpful to have sufficiently large sample sizes so that the relationship between being a member of an ethnic minority group and being HIV seropositive could be examined within individual geographic sites. Despite these limitations, there are very clear patterns in the data.

There are similar gradients for mean length of injecting history and baseline HIV prevalence running in descending order from the Vietnamese site farthest from the border to the Chinese site farthest from the border. These patterns are consistent with the theory that both the practice of drug injection and the prevalence of HIV infection among IDUs spread from Northern Vietnam to Southern China along a major heroin trans-shipment route [5]. The patterns in our data suggest that, in some circumstances, it may be possible to reconstruct histories of the diffusion of injecting drug use and HIV among IDUs using cross-sectional data.

There is clearly a potential for further cross-border transmission of HIV in both directions. Our discussions with the peer educators in the cross-border HIV prevention project suggest three primary reasons for these IDUs crossing the border: 1. Obtaining higher quality/lower priced drugs, 2. Avoiding police pressure on drug injectors, which can be unpredictably variable and involves large-scale periodic crackdowns, and 3. Personal factors, such as migratory employment, commerce, and family ties. It would appear to be very difficult to reduce these reasons for IDUs crossing the border.

Because of the possibility of arrest, IDUs who cross the border are unlikely to carry needles and syringes with them, even if they are crossing the border at places without any supervision or inspection.

Prevention of risky injections among border crossing IDUs will require very good supplies of sterile injection equipment on both sides. If IDUs who crosses the border cannot readily access sterile injection equipment on both sides of the border, then their fellow IDUs will need to have sufficient supplies of sterile injection equipment for use by themselves and the IDUs who cross the border. This will require large-scale safer injection programs on both sides of the border.

The majority of subjects in this study belong to ethnic minority groups, primarily Zhuang in China and Tay and Nung in Vietnam. Ethnic minority IDUs were also overrepresented among the border crossers. The issues of ethnic minority status, injecting drug use, and HIV infection deserve much more research and policy development. Ethnic minority IDUs are more likely to be infected with HIV in many places, from African-American and Latino/a IDUs in New York City [10] to Roma in Eastern Europe [11] to First Nations in Vancouver [12] to Vietnamese in Australia [13] to Manipuris in India [14]. As noted above, there was a strong relationship between ethnic minority status and HIV serpositivity in Ning Ming (OR = 5.08 (95% CI 1.41,18.26, p = .013) [15]. Social stigmatization of ethnic minority communities may make them more vulnerable to illicit drug use, including injecting drug use. Employment discrimination against ethnic minority communities may increase the extent to which drug distribution occurs in these communities, and to which drugs are transported by minority community members. Persons belonging to ethnic minority groups also may have important factors facilitating international travel, such as social support systems and persons that speak the same language on the other side of international borders.

The data presented here illustrate many of the factors in the international diffusion of HIV among IDUs at modest geographic scale. (There is a total distance of 71 kilometers between the two most distant sites in the study). These include gradients of length of injecting drug use and HIV seroprevalence across the international border, border crossing by IDUs and its association with HIV infection for those crossing from China into Vietnam, overrepresentation of ethnic minority persons among the border crossers. Both injecting drug use and HIV among IDUs are already well established among IDUs on the Vietnamese side of the border and injecting drug use is well established on the Chinese side of the border. HIV is present among IDUs on the Chinese side of the border, but at lower seroprevalence levels than in Lang Son.

There are multiple reasons that people cross the border in this area, and it would not appear to be possible to stop IDUs from crossing the border or from injecting drugs across the border. Thus, HIV prevention goals must include increasing the safety of injections among border crossers (as well as reducing risk behavior among the IDUs who do not cross the border). This will require coordinated HIV prevention that increases the likelihood that IDUs will inject safely on both sides of the border. Such a program has been implemented in Lang Son and Guangxi provinces. It includes peer outreach, increased access to sterile injection equipment through syringe distribution and exchange and a pharmacy voucher program. IDUs may exchange used injection equipment for new needles/syringes or for vouchers that can be redeemed at local pharmacies for needles/syringes, sterile water, and condoms. IDUs may also directly receive new needles/syringes or pharmacy vouchers even if they do not return used equipment. The program also includes large-scale collection and safe disposal of used needles/syringes, general community education about drugs and HIV, and social support for people living with HIV or AIDS [7].