• Overview of B&RD


    The goal of B&RD is to provide quality statistical support to further the endeavors of RCMI researchers by establishing viable working relationships with the Biostatistical Cores at each RCMI institution. Responsibilities include maintaining working relationships with RCMI investigators; producing statistical outcomes; consulting on clinical study design and statistical and regulatory issues; and providing expertise in statistical methodology applied to clinical studies. B&RD has supported more than 15 research projects for developing proposals/protocols (study design, sample size calculation, and statistical analysis plan), collecting data and preparing publications. The division has also developed and implemented various programs to increase the Network’s biostatistical capacities, such as conducting RTRN Biostatsitical Capacities and Needs Assessmenst, providing webinars on biostatistical seminars, workshops, on- and off-site consultations, and preparing RTRN biostatistician databases. The division has provided services for Network investigators to increase academic outcomes through the large secondary data including the Jackson Heart Study Vanguard Center data and various national survey datasets (NHANES, BRFSS, CPES, YRBS, etc).


    The DCC Biostatistics and Research Development Division not only continues supporting RTRN investigators-initiated research projects but also involves in developing multi-site research projects to provide junior scientists with opportunities to be involved in well-developed projects. The Division plays role as research idea bank as well as research supporting unit. The DCC Biostatistics Division as an organizer of RTRN biostatistical personnel resources provides various programs to reinforce working relationship to RTRN biostatisticians and graduate students through the Adjunct Scientist/Biostatistician appointment and Internship program. These personnel resources are being utilized for the research support and/or publication development. divider_pattern
  • Staff and Clinical Study Capabilities

    The current staff of the DCC Biostatistics Division is consisted of three senior members. They have 50 years of combined statistical experience and led/coauthored over 160 peer-reviewed scientific articles. They have supported various research projects including animal experimental study, genetics, clinical trials and epidemiology/community study. In addition to core members, the DCC B&RD has a strong working relationship with two outstanding biostatistical consultants who will start providing their biostatistical consultation : Dr. Chul Ahn, University of Texas Southwestern Medical Center and Dr. John Chen, University of Hawaii at Manoa.

    Their expertise is in study design, parametric and nonparametric longitudinal data analysis, latent variable analysis, outcome measurements, medical decision analysis, Rasch psychometric analysis, statistical methods for assessing gene-environment interaction, clinical trial design, computing power and sample size for correlated samples, multiplicity adjustment for confidence intervals, risk behavior (HIV, drug abusing, smoking, etc), cardiovascular disease, quality of life research, and the evaluation of intervention programs, cardiovascular disease epidemiology, substance abuse and HIV/AIDS, and pharmacoeconomics and outcomes research. divider_pattern

    Services

    •  Sample size calculations & power analysis

    •  Assisting with proposal development

    •  Protocol review

    •  Statistical Analysis Plan

    •  Randomization Scheme or matching scheme

    •  Data Analysis (interim & final)

    •  Preparing Clinical Study Report

    •  Statistical Contribution for Publications & Presentations

    •  Statistical Quality Control

    •  Genetic data analysis including Admixture Analysis

    •   Linkage Analysis

    •   Admixture Analysis

    •   Gene Expression Analysis

    •   Gene Environment Interaction analysis

    •   Geospatial Statistical Analysis

    •   Services for Large Dataset (Jackson Heart Study, NHANES, BRFSS, CPES, etc): Preliminary Analysis for Feasibility Tests, Develop Manuscript Proposals as needed, Develop Ancillary Studies, Provide Aggregate Data, Statistical Analysis, Assist preparing Publications & Presentations

  • divider_pattern

    Statistical assistance in developing scientific publications

    DCC B&RD has helped you to create manuscript development by conducting data analyses, creating tables and figures, and providing descriptions for methods, results and discussion sections.

    DCC B&RD has been involved in developing more than 30 scientific manuscripts and 70 peer-reviewed presentations for last 5 years.

    DCC B&RD has applied cutting-edge techniques to handle research data such as generalized mixed model, individual latent growth curve model, latent class transition model, Bayesian methods, advanced techniques to handle missing data, etc.

    divider_pattern

    Supporting Research Implementation

    The DCC B&RD will perform various efforts to prepare valid data for statistical analyses from the pre-stage of database building to the post-data collection session. Before building the database, critical variables will be decided and the annotated CRFs, data validation plan and database training plan will be reviewed and approved. During the research implementation, statistical quality control activities will be conducted by applying high level statistical methods to the accumulating data. The tools to be used are, but are not limited to, descriptive statistics, Shewart charts, plots, breakpoint regression, and recursive residuals with CUSUM and V-charts, which will be applied to measurement and event data. SAS macros will allow us to examine the quality of data through graphical outputs and various statistical outputs. The DCC statistician will play a leading role on the quality control committee (if applicable). All procedures for statistical quality control will be governed by the DCC SOP B-606 Statistical Quality Control. Before conducting data analyses, the entire collected data will be reviewed using the methods described above for the data locking and freezing processes. The reviewed data will be delivered to the PI(s) to obtain their final approval. Only data approved by the PI(s) will be analyzed for the final analyses.

    divider_pattern

    Training Programs in statistics and study design

    The DCC Biostatistical Mobile Clinic Collaborative Exchange (BMCCE) program, which was offered at six RTRN sites (Howard University, University of Puerto Rico, University of Central Caribbean, Ponce School of Medicine and Howard University) on May-June, 2011, was a good opportunity for DCC and sites to mutually understand the needs of RCMI investigators and to develop potential collaboration opportunities for future studies.

    divider_pattern

    Large Secondary Data Services

    The DCC procured various large datasets to support studies for generating hypothesis and resolving issues during the conduct of trials. The large datasets that the DCC has utilized include the Jackson Heart Study (Link to Vanguard Center) which is the largest single-site cohort study to prospectively investigate the determinants of cardiovascular disease among African Americans; the National Health and Nutrition Examination Survey (NHANES) which is a program of studies designed to assess the health and nutritional status of adults and children in the United States where interviews and physical examinations are combined; and the Behavioral Risk Factor Surveillance System (BRFSS) which is the world’s largest, on-going telephone health survey system, tracking health conditions and risk behaviors in the United States yearly since 1984. The DCC has also experienced in Youth Risk Behavior Surveillance data and Collaborative Psychiatric Epidemiology Surveys (CPES) which require advanced statistical skills to incorporate the complex sampling structure. The DCC has routinely utilized these datasets to address study questions. For example, the JHS data were utilized for the Minority Health Genomics and Translational Research Bio-Repository Database Network (MH-GRID) case-control study to evaluate how relaxing inclusion and exclusion criteria would impact enrollment and statistical power. Another example is the use of the NHANES data for the RTRN Vitamin D3 Pilot study in order to resolve the issue of whether the changed gender ratio for study sample recruitment had a potential effect on the study objective. The NHANES data have also been utilized for generating trial hypotheses or as preliminary results for grant proposals.

    The DCC B&RD has assisted RTRN researchers to generate scientific output by the use of secondary data and the provision of customized services. The customized services included: preliminary analyses for feasibility, data-mining, data workshops and idea sharing meetings, profiling potential investigators for team building, providing data information and query tools through the website, helping develop manuscript proposals, and providing data analyses for publication.

    divider_pattern

    Statistical Support for Genotype Endpoints

    The DCC will support statistical genetics and bioinformatics research with services ranging from primer design to phylogenetic analysis (e.g., migration, selection, and recombination). Analyses also include quantitative genetics modeling (e.g., linkage analysis, segregation analysis, identity-by-descent, additive polygenic models, linkage disequilibrium, and haplotype phase prediction and inference). Additionally, the DCC will assist in preparation of Illumina SAM/BAM files for GATK analysis (e.g., queue script for Next Generation Sequencing (NGS) processing, Iindel cleaning, duplicate marking and base score recalibration), generate volcano plots to identify under/over represented pathways, analyze copy number and loss of heterozygosity, perform multidimensional scaling for ordering markers within linkage groups, determine haplotype diversity, and create kernel density MA plots for microarray data. Statistical analysis includes basic inferential statistics, multivariate analysis, binary and Poisson regression models, non-parametric methods, group sequential boundary points, permutation-based micro-array analysis, and K-means clustering.

Untitled Document