Biostatistics Analysis Services

MDS’s Biostatistical Analysis Division provides statistical consulting services designed to take complex statistical analyses and translate their results into clear and meaningful conclusions for research and biotechnology institutions as well as drug development companies. We provide statistical services to clients for a range of studies including toxicity, tumorgenicity, and immunogenicity.

We can assist with all stages of the clinical study including start-up, data analysis, and summary reports during our statistical methods. More specifically, we offer the following services for each study phase:

Start-Up

• Formulating Hypotheses or Study Aims
• Defining Appropriate Predictor, Response, and Confounding Variables
• Devising Appropriate Study Designs
  • Calculation of Sample Size and Power
  • Randomization and Sampling Schemes
• Planning the Data Collection and Management System

Data Analysis

• Statistical Analysis Plan
• Statistical Programming Using SAS or R
• Descriptive, Univariate, and Multivariate Methods (see below)
• Unscheduled Interim Analyses
• Post-Hoc and Exploratory Analyses
• Formal Interim and Final Analyses

Summary Reports

• Informal Data Summaries
  • Statistical Tables and Figures
  • Informal Reports
• Interpretation of Results
• Formal Statistical Reports
• Project Specific Documentation

Although we are happy to provide statistical consulting advice on specific stages of a clinical study, involving MDS in all stages assures consistency and clarity. Moreover, we provide the following as part of our services.

Statistical Analysis Plan
MDS will prepare a complete Statistical Analysis Plan that includes a detailed description of all statistical analyses to be carried out on the data collected during the clinical investigation. Mock-ups of tables, listings, and figures will also be part of the Statistical Analysis Plan.

Statistical Programming
All work will be done using industry-standard software, either SAS or R. The code used to conduct all statistical analyses will be thoroughly-annotated and subjected to validation through independent review by a statistician different from the one who generated the original programs.

Statistical Report
A statistical report will be produced in conjunction with the preparation of a full clinical report. The statistical report will provide a description of the findings of the statistical analyses for the clinical investigation and will confirm that the analyses were conducted per the statistical analysis plan. Deviations from the statistical analysis plan or additional unplanned analyses will be fully described in the statistical report.

The statistical methods used to analyze the data will vary from study to study but may include one or more of the following:

Statistical Methods

• Descriptive Statistics
  • Central Tendency (Mean, Median, etc.)
  • Variability (Variance, Standard Error, Confidence Intervals, etc.)
• Sample Size and Power Calculation
• Categorical Analyses
  • Chi-Square
  • Contingency Tables
  • Binomial Test
• Analysis of Variance (ANOVA)
  • Basic
  • Repeated Measures ANOVA
  • MANOVA
  • ANCOVA
• Correlation Analysis
• Regression Analysis
  • Multiple Regression
  • Non-Linear Regression
  • Logistic Regression
• Multivariate Statistics
  • Cluster Analysis
  • Ordination/Principal Components Analysis
  • Discriminant Analysis
  • Canonical Correlation Analysis
  • Factor Analysis
• Interim Analyses and Stopping Rules
• Survival Analysis
• Equivalence Testing
• Odds Ratios
• Log Likelihood Analysis
• Non-Parametric Analogs to the Methods Listed Above

For example, the following statistical analyses might be applied to a toxicity study, while an immunogenicity ELISA would demand an entirely different set of analyses.