Consulting Services
Kelly Statistical Consulting has extensive experience in the application of sound statistical methodology to diverse application areas. We can assist you in designing your study to ensure that the appropriate variables are measured in the most precise and accurate manner and that the best, most powerful statistical analysis will be used to improve your results and save you time and money.
Our philosophy is to build long-term relationships by ensuring client satisfaction. Our clients depend on us for accurate and insightful advice in a timely manner. Frequent and effective communication is our practical approach to ensuring that we are meeting our clients’ needs. Our clients have trusted us with their consulting needs for the following key reasons:
- Quick response times.
- Results explained in non-technical, understandable language.
- Personalized approach with on-going communication.
- Solutions specific to the problems at hand.
- Compliance with FDA and ISO standards.
- Extensive theoretical research and applied industry experience.
Medical devices
Medical device trials can have unique statistical methodological issues. Whether the goal is to show the device is as good as its predicate (for a 510(k) submission) or to show that the device outperforms a performance goal, we have expertise in the design and analysis of animal and clinical studies for medical devices. In clinical trials, appropriate treatment of missing data is especially important to provide unbiased estimates of treatment effects. We assist medical device companies with:
- Meta-analyses to establish performance goals
- Designing powerful and efficient clinical trials, either single arm or randomized multiple arm studies (with interim analyses) to show non-inferiority or superiority
- Designing post-market observational studies that account for potential confounding variables
- Design and analysis of cost-effectiveness studies
- Analyzing pre-clinical, clinical trial, and quality control data
- Writing study protocols, Statistical Analysis Plans, and Clinical Study Reports
- Assisting in the submittal of pre-market approvals (PMAs), 510(k), HUD, and product development protocols (PDPs) to the FDA
Example Projects:
How can an ANSI standard for assessing accuracy be extended to a new type of device?
Diagnostics
We have expertise in developing biomarker signatures for diagnosis, as well as designing validation studies for more traditional blood or urine assays and medical device diagnostics. Developing gene expression or other biomarker diagnostic signatures can be problematic due to the large number of potential predictors and the relatively small number of subjects. Appropriate statistical methods for this type of “big data” are required to avoid overfitting, so that the signature accurately diagnoses future patients. Analysis of blood and urine assays, as well as medical device diagnostics, can also be tricky, particularly for continuous measurement applications. We assist biotechnology diagnostic companies in:
- Developing appropriate statistical analysis plans for diagnostics.
- Precision analyses of qualitative and quantitative diagnostics, linearity, stability interference, limits of detection and quantitation analyses using CLSI standards
- Method comparison, sensitivity, specificity and accuracy analyses
- Quality control, acceptance sampling, tolerance interval design
- Writing study protocols, Statistical Analysis Plans, and Clinical Study Reports
Example Projects:
How can an ANSI standard for assessing accuracy be extended to a new type of device?
Pharmaceuticals and Supplements
A good study design for a drug trial is key to ensuring that the most accurate and powerful information is gathered, while exposing the fewest subjects to potentially harmful effects. We are accomplished researchers and practitioners in the design and analysis of in-vitro, in-vivo, animal, and phase I-IV clinical studies, including dose-ranging, dose-response, mode-of-action assessment, pharmacokinetic, pharmacodynamic, drug interactions, safety and efficacy, equivalence and bioequivalence studies. In clinical trials, appropriate treatment of missing data is especially important to provide unbiased estimates of treatment effects. We assist companies developing pharmaceutical and supplement products in:
- Designing powerful and effective clinical trials, using completely randomized, block randomized, or group sequential designs (with interim analyses)
- Writing study protocols, Statistical Analysis Plans, and Clinical Study Reports
- Analyzing pre-clinical, clinical trial, and quality control data
- Assisting in the submittal of new drug applications (NDAs), post-marketing studies, or cost-effectiveness studies to the FDA and other regulatory bodies
Example Projects:
Are two drugs better than one?
Is the New Zealand strain-specific meningococcal disease vaccine effective?
Conservation and ecological investigations
Proper statistical design and analysis of conservation and ecological studies contributes scientific rigor and credibility to their results. Kelly Statistical Consulting has experience in estimating population sizes and assessing the factors that affect the health of endangered species. In addition, we have assisted with studies to assess dangers to survival, population demographics, genetic variability, phylogenic relationships and preferred habitat and food in species such as the humpback whale, black rails, California sheephead, junko (a rare sparrow), warblers, buckwheat, vole, baleen whales, skinks, New Zealand crabs, otters, fungi, the Maud Island frog, and others.
Example Projects:
How many bat and bird fatalities are caused by wind turbines?
