We propose a two-phase study to assess the prevalence of latent, active and MDR-TB, and the factors related to this illness among drug users not- in-treatment in the San Juan metropolitan area (Phase I). We also propose to elucidate the factors that facilitate and impede the initiation, continuation and completion of TB treatment and compare drug users to non- drug users in terms of TB treatment use (Phase II). Study Phase I. This study is based on the assumption that the high prevalence of HIV/AlDS among out-of-treatment drug users (34%) place these individuals at great risk of acquiring TB. The proposed community sampling plan described in Appendix 2 establishes subject recruitment procedures that allow generalization to a larger population, and the measurement of trends in risk behaviors, HIV seropositivity and TB infection. The proposed procedures aim to reduce selection biases to the greatest extent possible given the sampling limitations associated with this study population. Our sampling procedures require four basic steps: identify copping areas (sites where drugs can be purchased), randomly select a number of copping areas and recruitment days and times of each month, count area users on randomly selected day and time, and select users at random from each selected copping area and day and time. The project will work through assessment centers strategically situated in catchment areas of metropolitan San Juan, the area with the highest prevalence of drug use and AIDS in Puerto Rico. Household members of drug abusers diagnosed with a TB condition will also be invited to participate and tested for TB infection. Specific statistical analysis have been planned for each one of the phases of the project. Data from the community based survey will be used to derive descriptive epidemiological rates and estimates. Co- morbidity of TB with HIV and risk factors associated with TB will also be studied utilizing these data. Prevalence rates and confidence intervals will be calculated for the whole sample as well as for HIV seropositives and seronegatives, and age, sex, and main drug of abuse sub-groups. With the aid of Chi-Square tests, differences of proportions between the sub- groups will be examined. Trends in TB prevalence rates over the 18 month period of community surveillance will be examined. In addition, follow-up measures taken from latent TB cases and anergic drug abusers will allow for the estimation of active TB incidence rates. Confidence intervals on incidence rates will be estimated using a Poisson distribution. The analytical procedures to assess TB prevalence among the members of the households of drug abusers will be similar to those of drug abusers except that several potential factors affecting TB infection among household members will be explored using multiple logistic regression. Study Phase II. Latent and active TB cases identified in the community will be combined with latent and active TB cases identified through the participating TB control units of the Puerto Rico Health Department and followed-up with a three-wave longitudinal panel design in order to study the utilization of TB treatment. These data will be used to estimate the likelihood of latent and active TB cases entering, continuing and completing the prescribed therapeutic treatments. Comparability between the samples will be established through contingency analysis and Chi- Square tests to measure differences in proportions. Multiple logistic regression analysis will be used to assess the effects of the variables specified in the theoretical model on each stage of TB treatment utilization. Stratified analyses will help determine significant differences in the factors affecting treatment utilization between drug abusers and non-drug abusers. Drug abuse research using longitudinal designs must deal with some proportion of incomplete data due to attrition and non-response. We will employ a threefold strategy to reduce the validity concerns due to incomplete data: 1) Design of a systematic follow-up and relocation plan, 2) Continuous monitoring of a sample of sound-recorded interviews, and 3) Exploratory analysis to discover systematic differences between respondents with complete and incomplete data.
| Robles, R R; Reyes, J C; Colon, H M et al. (1999) Prevalence and correlates of anergy among drug users in Puerto Rico. Int J Epidemiol 28:509-13 |
| Robles, R R; Marrero, C A; Reyes, J C et al. (1998) Risk behaviors, HIV seropositivity, and tuberculosis infection in injecting drug users who operate shooting galleries in Puerto Rico. J Acquir Immune Defic Syndr Hum Retrovirol 17:477-83 |
| Reyes, J C; Robles, R R; Colon, H M et al. (1996) Mycobacterium tuberculosis infection among crack and injection drug users in San Juan, Puerto Rico. P R Health Sci J 15:233-6 |
