Data collected by the Multicenter AIDS Cohort Study (MACS) are extremely valuable for the description of the natural history of HIV-1 infection. The profile of clinical disease is expected to change with increasing use of therapies. Appropriate methods are needed to determine characteristics that explain the variability of AIDS-free time with emphasis in long term survivorship. The MACS provides the means of assessing the changing epidemic and its effects on the clinical, psychosocial and financial burden on HIV-1 infected individuals. The data coordinating center will continue to provide statistical and methodological expertise for the interpretation of the MACS data and for the planning of future MACS studies. In particular, methods will be provided to: l) plan, design and coordinate data acquisition; 2) standardize procedures for transferring, editing, merging, storing and backing up the data; 3) implement appropriate data analytical procedures; 4) develop statistical methods pertinent to research questions for which the MACS data is a valuable resource; 5) procure and develop hardware and software to carry out the coordination, management and analysis of the data; and 6) participate in study working groups and writing committees. Research coordination will be continued through our established collaborative activities with investigators and through the direction of the Data Working Group. Data will be organized in an ORACLE database using a SUN 69OMP network with SUN SPARCstations. Software is available to download data files for analysis using a variety of statistical packages including S, SAS and EGRET. Analytical methods include those of longitudinal data (e.g., trajectories of CD4), survival analysis (e.g., variability of AIDS-free time), categorical data analysis (e.g., conditional logistic regression for the analysis of nested case-control studies of malignancies). Expertise in each of the above areas will allow MACS to continue its contributions in the study of HIV infection: an epidemic of major public health concern.
Showing the most recent 10 out of 434 publications
