Parsing social network survey data from hidden populations using stochastic context-free grammars.

BACKGROUND: Human populations are structured by social networks, in which individuals tend to form relationships based on shared attributes. Certain attributes that are ambiguous, stigmatized or illegal can create a OhiddenO population, so-called because its members are difficult to identify. Many hidden populations are also at an elevated risk of exposure to infectious diseases. Consequently, public health agencies are presently adopting modern survey techniques that traverse social networks in hidden populations by soliciting individuals to recruit their peers, e.g., respondent-driven sampling (RDS). The concomitant accumulation of network-based epidemiological data, however, is rapidly outpacing the development of computational methods for analysis. Moreover, current analytical models rely on unrealistic assumptions, e.g., that the traversal of social networks can be modeled by a Markov chain rather than a branching process. METHODOLOGY/PRINCIPAL FINDINGS: Here, we develop a new methodology based on stochastic context-free grammars (SCFGs), which are well-suited to modeling tree-like structure of the RDS recruitment process. We apply this methodology to an RDS case study of injection drug users (IDUs) in Tijuana, México, a hidden population at high risk of blood-borne and sexually-transmitted infections (i.e., HIV, hepatitis C virus, syphilis). Survey data were encoded as text strings that were parsed using our custom implementation of the inside-outside algorithm in a publicly-available software package (HyPhy), which uses either expectation maximization or direct optimization methods and permits constraints on model parameters for hypothesis testing. We identified significant latent variability in the recruitment process that violates assumptions of Markov chain-based methods for RDS analysis: firstly, IDUs tended to emulate the recruitment behavior of their own recruiter; and secondly, the recruitment of like peers (homophily) was dependent on the number of recruits. CONCLUSIONS: SCFGs provide a rich probabilistic language that can articulate complex latent structure in survey data derived from the traversal of social networks. Such structure that has no representation in Markov chain-based models can interfere with the estimation of the composition of hidden populations if left unaccounted for, raising critical implications for the prevention and control of infectious disease epidemics.

A tale of two cities: Social and environmental influences shaping risk factors and protective behaviors in two Mexico-US border cities.

The economic, social, cultural, and political milieus that influence HIV risk behaviors along the US-Mexico border are understudied. In an effort to appropriately inform interventions targeting structural influences, we compared injecting drug using populations living in two cities--Ciudad Juárez, Chihuahua and Tijuana, Baja California--situated on the Mexico-US border. These populations presented with similar demographic profiles, but differed significantly in terms of social and environmental influences that can influence both risk and protective factors (e.g., family drug use, migration, drug use patterns). We observed distinct behavioral and structural influences in these two border cities that will require tailored intervention strategies to reduce HIV transmission.

Cross-border paid plasma donation among injection drug users in two Mexico-U.S. border cities.

OBJECTIVE: Paid plasma donation has contributed to HIV epidemics in many countries. Eleven million liters of plasma are fractionated annually in the U.S., mainly from paid donors. Deferral of high-risk donors such as injection drug users (IDUs) is required for paid donations. We studied circumstances surrounding paid plasma donation among IDUs in two Mexico-U.S. border cities. METHODS: In 2005, IDUs > or = 18 years old in Tijuana (N=222) and Cd. Juarez (N=206) who injected in the last month were recruited through respondent-driven sampling. Subjects underwent antibody testing for HIV and HCV and an interviewer-administered survey including questions on donating and selling whole blood and plasma. RESULTS: Of 428 IDUs, HIV and HCV prevalence were 3% and 96%, respectively; 75 (17.5%) reported ever having donated/sold their blood or plasma, of whom 28 (37%) had sold their plasma for an average of $16 USD. The majority of IDUs selling plasma were residents of Ciudad Juarez (82%); 93% had sold their plasma only in the U.S. The last time they sold their plasma, 65% of IDUs had been asked if they injected drugs. Although the median time since last selling plasma was 13 years ago, 3 had done so within the prior 2 years, one within the prior 6 months; of these 3 IDUs, 2 were from Cd. Juarez, one from Tijuana; all 3 had only sold their plasma in the U.S. CONCLUSIONS: Although selling plasma appears uncommon among IDUs in these two Mexican border cities, the majority sold plasma in the U.S. and only one-third were deferred as high-risk donors. Paying donors for plasma should be a matter of public inquiry to encourage strict compliance with regulations. Plasma clinics should defer donors not only on behavioral risks, but should specifically inspect for injection stigmata.

Shooting gallery attendance among IDUs in Tijuana and Ciudad Juarez, Mexico: correlates, prevention opportunities, and the role of the environment.

We identified factors associated with shooting gallery attendance among injection drug users (IDUs) in two Mexico-US border cities. IDUs in Tijuana (n=222) and Ciudad Juarez (n=205), Mexico, who were >or=18 years and injected illicit drugs in the last month were recruited using respondent-driven sampling (RDS). An interviewer-administered survey collected sociodemographic and behavioral data. Logistic regression was used to examine correlates of shooting gallery attendance in each of the two cities. Homelessness and being arrested for syringe possession--both structural level factors--were associated with shooting gallery use in both cities. In Ciudad Juarez, younger age and having overdosed were also associated with shooting gallery use. Our study highlights the need for structural interventions that mitigate homelessness among IDUs and facilitate changes in law enforcement practices associated with shooting gallery use. Harm reduction interventions based within shooting galleries should also be considered to prevent transmission of blood-borne pathogens among IDUs.

Hepatitis C virus transmission among oral crack users: viral detection on crack paraphernalia.

OBJECTIVE: Epidemiological studies present oral crack use as a potential independent risk factor for hepatitis C virus (HCV) status, yet actual HCV transmission pathways via crack use have not been evidenced. To this end, this exploratory study sought to detect HCV on crack-use paraphernalia used by street crack users. METHODS: Crack-use paraphernalia within 60 min of use was collected from 51 (N) street-crack users. HCV RNA detection was conducted through eluate sampling and manual RNA extraction. Participants provided a saliva sample to test for HCV antibody, and had a digital photograph taken of their oral cavities, to assess the presence of oral sores as a possible risk factor for oral HCV transmission. RESULTS: About 43.1% (n=22) of the study participants were HCV-antibody positive. One (2.0%) of the 51 pipes tested positive. A minority of the participants presented oral sores. The pipe on which HCV was detected was made from a glass stem; its owner was HCV-antibody positive, and there was full rater agreement on the presence of oral sores in the pipe owner's oral cavity. CONCLUSIONS: HCV transmission from an infected host onto paraphernalia as a precondition of HCV host-to-host transmission via shared crack paraphernalia use seems possible, with oral sores and paraphernalia condition constituting possible risk modifiers. Larger-scale studies with crack users are needed to corroborate our findings.

At the borders, on the edge: use of injected methamphetamine in Tijuana and Ciudad Juarez, Mexico.

Injection drug use is of increasing concern along the US-Mexico border where Tijuana and Ciudad (Cd.) Juarez are located. Methamphetamine has long been manufactured and trafficked through Mexico, with low rates of use within Mexico. With methamphetamine use now considered epidemic in the United States, and with associated individual and community harms such as HIV, STDs, domestic violence and crime, there is concern that rates of methamphetamine in the Northwestern border regions of Mexico may be rising. We conducted a qualitative study to explore the context of injection drug use in Tijuana and Cd. Juarez and included questions about methamphetamine. Guided in-depth interviews were conducted with 10 male and 10 female injection drug users (IDUs) in Tijuana and 15 male and 8 female IDUs in Cd. Juarez (total N = 43). Topics included types of drug used, injection settings, access to sterile needles and environmental influences. Interviews were taped, transcribed verbatim and translated. Content analysis was conducted to identify themes. The median age of injectors in both cities was 30. Methamphetamine was injected, either alone or in combination with other drugs by injectors in both Tijuana (85%) and Cd. Juarez (17%) in the 6 months previous to interview. Several important themes emerged with respect to methamphetamine use in both cities. IDUs in both cities considered methamphetamine to be widely used in Tijuana and infrequently used in Cd. Juarez, while the converse was true for cocaine. In both cities, stimulant (either cocaine or methamphetamine) use was widespread, with 85% in Tijuana and 83% in Cd. Juarez reporting current use of a stimulant, most often used in combination with heroin. Some injectors reported knowledge of local manufacturing and one had direct experience in making methamphetamine; some cross-border use and trafficking was reported. Injectors reported concerns or experience with serious health effects of methamphetamine such as abscesses or tuberculosis. Our study suggests that injected methamphetamine is entrenched in Tijuana and that Cd. Juarez may experience a methamphetamine outbreak in the future. Robust targeted interventions for both injected and non-injected methamphetamine should be a public health priority in both cities.

Prevalence of hepatitis C virus and HIV infection among injection drug users in two Mexican cities bordering the U. S

OBJECTIVE: To estimate the prevalence of the hepatitis C virus (HCV) and HIV infection and associated risk behaviors among injection drug users (IDUs) in two northern Mexican cities. MATERIAL AND METHODS: Between February and April 2005, IDUs were recruited in Tijuana (N=222) and Ciudad Juarez (N=206) using respondent-driven sampling (RDS), a chain referral sampling approach. Interviewer-administered questionnaires assessed drug-using behaviors during the prior six months. Venous blood was collected for immunoassays to detect HIV and HCV antibodies. For HIV, Western blot or immunofluorescence assay was used for confirmatory testing. Final HCV antibody prevalence was estimated using RDS adjustments. RESULTS: Overall, HCV and HIV prevalence was 96.0 percent and 2.8 percent, respectively, and was similar in both cities. Most IDUs (87.5 percent) reported passing on their used injection equipment to others, and 85.9 percent had received used equipment from others. CONCLUSIONS: HIV prevalence was relatively high given the prevalence of HIV in the general population, and HCV prevalence was extremely high among IDUs in Tijuana and Ciudad Juarez. Frequent sharing practices indicate a high potential for continued transmission for both infections. HCV counseling and testing for IDUs in Mexico and interventions to reduce sharing of injection equipment are needed.

OBJETIVO: Estimar las prevalencias de los virus de hepatitis C (VHC) y de VIH y los comportamientos de riesgo asociados con ellos, entre usuarios de drogas inyectables (UDI) en dos ciudades del norte de México. MATERIAL Y MÉTODOS: Entre febrero y abril de 2005, se reclutaron UDIs en Tijuana (N=222) y en Ciudad Juárez (N=206), mediante un método de muestreo llamado en inglés "respondent-driven sampling" (RDS), lo cual es un sistema basado en cadenas de referencia. Los participantes contestaron una encuesta aplicada por entrevista, la cual indagó acerca de los comportamientos en el uso de drogas durante los seis meses previos. Una muestra de sangre venosa fue colectada de cada individuo, para determinar la presencia de anticuerpos contra VIH y VHC mediante técnicas inmunoenzimáticas. En el caso del VIH la técnica de "Western blot" se aplicó con fines de confirmación. La prevalencia final de anticuerpos contra VHC se hizo mediante un cálculo ajustado, que empleó un estimador poblacional del RDS. RESULTADOS: Las seroprevalencias globales de VHC y VIH, fueron 96 por ciento y 2.8 por ciento, respectivamente. Estas frecuencias fueron similares entre las muestras de ambas ciudades. La gran mayoría de los UDI (87.5 por ciento) manifestó haber transferido a otros sus equipos de inyección usados y a su vez 85.9 por ciento de los participantes declaró haber recibido equipos usados de otros. CONCLUSIONES: La seroprevalencia encontrada de VIH fue relativamente alta dada la prevalencia de VIH en la población general y la de VHC fue extremadamente alta entre los UDI estudiados en Tijuana y en Ciudad Juárez. Las prácticas frecuentes de compartimiento de equipo señalan hacia un alto potencial que favorece la transmisión de ambas infecciones investigadas. Por tanto, son necesarias actividades de consejería y pruebas de laboratorio para VHC dirigidas a UDI en México y asimismo intervenciones para reducir el uso compartido de equipos de inyección.

Prevalence of hepatitis C virus and HIV infection among injection drug users in two Mexican cities bordering the U.S.

OBJECTIVE: To estimate the prevalence of the hepatitis C virus (HCV) and HIV infection and associated risk behaviors among injection drug users (IDUs) in two northern Mexican cities. MATERIAL AND METHODS: Between February and April 2005, IDUs were recruited in Tijuana (N=222) and Ciudad Juarez (N=206) using respondent-driven sampling (RDS), a chain referral sampling approach. Interviewer-administered questionnaires assessed drug-using behaviors during the prior six months. Venous blood was collected for immunoassays to detect HIV and HCV antibodies. For HIV, Western blot or immunofluorescence assay was used for confirmatory testing. Final HCV antibody prevalence was estimated using RDS adjustments. RESULTS: Overall, HCV and HIV prevalence was 96.0% and 2.8%, respectively, and was similar in both cities. Most IDUs (87.5%) reported passing on their used injection equipment to others, and 85.9% had received used equipment from others. CONCLUSIONS: HIV prevalence was relatively high given the prevalence of HIV in the general population, and HCV prevalence was extremely high among IDUs in Tijuana and Ciudad Juarez. Frequent sharing practices indicate a high potential for continued transmission for both infections. HCV counseling and testing for IDUs in Mexico and interventions to reduce sharing of injection equipment are needed.

Prevalence and associated factors of hepatitis C infection (HCV) in a multi-site Canadian population of illicit opioid and other drug users (OPICAN).

BACKGROUND: Hepatitis C virus (HCV) infection is highly prevalent in illicit drug user populations, with three in four new HCV infections related to this risk behaviour and a growing HCV disease burden in Canada. Using data from a multi-site cohort study of illicit opioid users in five Canadian cities (OPICAN), this paper explores the prevalence and predictors of HCV status in this high-risk population. METHODS: HCV status of cohort participants was assessed by salivary antibody test. Univariate relationships of HCV status with select variables were examined on the basis of cohort baseline data, and subsequently multivariate models using logistic regression to determine independent predictors of HCV status were generated. RESULTS: 54.6% of the analysis sample (n=482) was HCV positive. Significant differences in terms of HCV prevalence existed across the sites. Significant variables in the final stepwise logistic regression model included age, site (Toronto), unprotected sex, injecting drug use, drug treatment and incarceration in past year, in addition to opioid use in combination with non-opioids. DISCUSSION: Besides drug injecting, various other socio-behavioural factors were associated with HCV status in our cohort. On this basis, interventions focusing solely on injection risks are overly limited in scope to prevent HCV transmission in the high-risk population of illicit drug users and need to be broadened. Prevention efforts should also target young injectors as a priority.

Respondent-driven sampling of injection drug users in two U.S.-Mexico border cities: recruitment dynamics and impact on estimates of HIV and syphilis prevalence.

Respondent-driven sampling (RDS), a chain referral sampling approach, is increasingly used to recruit participants from hard-to-reach populations, such as injection drug users (IDUs). Using RDS, we recruited IDUs in Tijuana and Ciudad (Cd.) Juárez, two Mexican cities bordering San Diego, CA and El Paso, TX, respectively, and compared recruitment dynamics, reported network size, and estimates of HIV and syphilis prevalence. Between February and April 2005, we used RDS to recruit IDUs in Tijuana (15 seeds, 207 recruits) and Cd. Juárez (9 seeds, 197 recruits), Mexico for a cross-sectional study of behavioral and contextual factors associated with HIV, HCV and syphilis infections. All subjects provided informed consent, an anonymous interview, and a venous blood sample for serologic testing of HIV, HCV, HBV (Cd. Juárez only) and syphilis antibody. Log-linear models were used to analyze the association between the state of the recruiter and that of the recruitee in the referral chains, and population estimates of the presence of syphilis antibody were obtained, correcting for biased sampling using RDS-based estimators. Sampling of the targeted 200 recruits per city was achieved rapidly (2 months in Tijuana, 2 weeks in Cd. Juárez). After excluding seeds and missing data, the sample prevalence of HCV, HIV and syphilis were 96.6, 1.9 and 13.5% respectively in Tijuana, and 95.3, 4.1, and 2.7% respectively in Cd. Juárez (where HBV prevalence was 84.7%). Syphilis cases were clustered in recruitment trees. RDS-corrected estimates of syphilis antibody prevalence ranged from 12.8 to 26.8% in Tijuana and from 2.9 to 15.6% in Ciudad Juárez, depending on how recruitment patterns were modeled, and assumptions about how network size affected an individual's probability of being included in the sample. RDS was an effective method to rapidly recruit IDUs in these cities. Although the frequency of HIV was low, syphilis prevalence was high, particularly in Tijuana. RDS-corrected estimates of syphilis prevalence were sensitive to model assumptions, suggesting that further validation of RDS is necessary.