Biostatistics Core

点击翻译按钮获取中文摘要

The overarching goals of the Biostatistics Core (BC), RC-2, are to provide design and analytical services to OAIC investigators conducting studies of multifactorial geriatric conditions; to develop and disseminate new design and analytical techniques for conducting studies with older persons; and to train a cadre of clinical investigators, biostatisticians, and epidemiologists in the skills necessary to design, conduct, and analyze gerontologic studies. The BC provides state-of-the-art research design and biostatistical expertise for the Yale OAIC with the following specific aims: 1) to collaborate with Pepper Scholars and other REC trainees, PES investigators, Operations Core (RC-1) colleagues, and investigators of External Projects (EPs) to achieve the mission of the Yale OAIC in promoting the understanding and management of multifactorial geriatric conditions; 2) to develop new and rigorous methods that address the challenges of designing and analyzing studies in basic, translational, and clinical geriatric/aging research, with special emphasis on introducing rigorous statistical methods; and 3) to train the next generation of gerontologic biostatisticians, epidemiologists, and early-stage investigators in Gerontologic Biostatistics research methods specific to studies of multifactorial geriatric conditions, e.g., methods for handling missing data due to attrition. The BC’s standard operating procedures address: 1) missing data, 2) reproducibility, 3) effect modification and mediation, 4) design of experiments, 5) conduct of trials, 6) multiplicity, 7) consideration of sex as a biological variable, 8) assignments and resource allocation to Yale OAIC projects, 9) mentoring, 10) documentation of data and analytical procedures, and 11) publication of statistical results. The BC’s collaboration and services to the OAIC and EPs include: 1) design for grant preparation of clinical trials and basic, translational, and epidemiological studies, including sample size estimation, randomization, and strategies to prevent or reduce missing data; 2) consultation on measurement instruments; 3) quality control procedures for ensuring rigor and reproducible analyses; 4) statistical analyses; 5) interpretation of study results in statistically responsible and clinically meaningful ways for abstracts, manuscripts, and presentations, often with specialized figures; 6) methodological development and novel application of methods from other fields relevant to the study of multifactorial geriatric conditions; and 7) education, mentoring, and training of junior investigators in Gerontologic Biostatistics. The proposed Biostatistics Core’s Development Project (DP-2), which focuses on the application of time-varying effects modeling (TVEM) in aging research, is innovative because (TVEM) has rarely been used in aging research. These longitudinal TVEM models can make important contributions to understanding the health of older adults by numerically and graphically showing how the effects of patient or clinical characteristics vary intensely over time. DP-2 will also provide new insights about how rigorous missing data methods can be used with TVEM.