The principal aim of this project is to further develop and implement new epidemiologic methods developed by and author for analyzing observational data bases and randomized trials of HIV-infected persons. The proposed approach is to a large extent based on the estimation of the parameters of a new class of causal models, the structural nested models using a new class of estimators - the G-estimators. The new methods are fundamentally """"""""epidemiologic"""""""" in that they require data on time-independent and time- dependent confounding factors - that is, risk factors for the outcome of interest that also predict subsequent treatment with or exposure to the drug or co-factor under study. The proposed methods of analysis will improve upon previous methods in the following ways. First, the new methods are the only methods available to estimate the effect of a treatment (e.g., mortality, time to AIDS, or CD4-count level) from data available in an observational data base, whenever symptoms of HIV disease (e.g., thrush, fever) are simultaneously confounders and intermediate variables on the causal pathway from the treatment or co- factor under study to the outcome of interest. We shall use the new methods of analyze the effect of marijuana, alcohol, cocaine, cigarette smoking, continued high-risk sexual behavior, aerosolized pentamidine, and AZT on the evolution of CD4-counts and on time to progression of HIV- disease, AIDS and death among the HIV-infected subjects in the San Francisco Men's Health Study (SFMHS), an observational longitudinal study of HIV-infected gay men. Results will be compared with those obtained using standard methods. Second, the new methods are the only methods available that can appropriately adjust for the concurrent effect of additional non-randomized treatments in randomized clinical trials. For example, in the AIDS Clinical Trial Unit (ACTU) trial 002 of the effect of high dose versus low- dose AZT on the survival of AIDS patients, patients in the low-dose arm had improved survival, but they also took more aerosolized pentamidine (a non- randomized treatment). We shall use our methods to adjust for the differential pentamidine usage. The new methods are the only methods available that efficiently incorporate information on surrogate markers (e.g., CD4-count) in order to stop, at the earliest possible moment, randomized trials of the effects a treatment on a survival time outcome (e.g., time to AIDS). Specifically, these methods allow one to construct a valid alpha-level test of the null hypothesis of no effect of treatment on survival that incorporates data on the evolution of the surrogate marker, e.g., CD4-count whose power may greatly exceed that of the log-rank test. We shall reanalyze ACTU randomized trials 019 and 016 to determine whether the proposed methods could have led to earlier stoppage of either trial.
| Richardson, Thomas S; Robins, James M; Wang, Linbo (2018) Discussion of ""Data-driven confounder selection via Markov and Bayesian networks"" by Häggström. Biometrics 74:403-406 |
| Wang, Linbo; Richardson, Thomas S; Zhou, Xiao-Hua (2017) Causal analysis of ordinal treatments and binary outcomes under truncation by death. J R Stat Soc Series B Stat Methodol 79:719-735 |
| Evans, R J; Forcina, A (2013) Two algorithms for fitting constrained marginal models. Comput Stat Data Anal 66:1-7 |
| Evans, Robin J; Richardson, Thomas S (2013) Marginal log-linear parameters for graphical Markov models. J R Stat Soc Series B Stat Methodol 75:743-768 |
| Page, John H; Rotnitzky, Andrea (2009) Estimation of the disease-specific diagnostic marker distribution under verification bias. Comput Stat Data Anal 53:707-717 |
| Moodie, Erica E M; Richardson, Thomas S (2009) Estimating Optimal Dynamic Regimes: Correcting Bias under the Null: [Optimal dynamic regimes: bias correction]. Scand Stat Theory Appl 37:126-146 |
| Shardell, Michelle; Scharfstein, Daniel O; Vlahov, David et al. (2008) Inference for cumulative incidence functions with informatively coarsened discrete event-time data. Stat Med 27:5861-79 |
| Moodie, Erica E M; Richardson, Thomas S; Stephens, David A (2007) Demystifying optimal dynamic treatment regimes. Biometrics 63:447-55 |
| Scharfstein, Daniel O; Halloran, M Elizabeth; Chu, Haitao et al. (2006) On estimation of vaccine efficacy using validation samples with selection bias. Biostatistics 7:615-29 |
| Hernan, Miguel A; Lanoy, Emilie; Costagliola, Dominique et al. (2006) Comparison of dynamic treatment regimes via inverse probability weighting. Basic Clin Pharmacol Toxicol 98:237-42 |
Showing the most recent 10 out of 25 publications
