Biostatistics
Biostatistics is the toolset to be used to generate scientific knowledge and to enable decision-making along with product optimization using methods of statistics. In clinical trials Numerics is evaluating the effectiveness of drugs, medical devices (implants), or treatment regimens. An optimal study design is a key determinant for the success of a study. Starting from a hypothesis we work together with our customers to develop an adequate study design, which has to answer the study objectives. This includes, inter alia, the sample size estimation (so that the study has adequate statistical power), the randomization of subjects in comparative studies, and finally, the statistical modeling of the study endpoints. Numerics uses multivariate methods in order to obtain clear conclusions about the effectiveness with minimal patient or experimental units. In the same way, by using statistical methods in observational (epidemiological) studies, risk factors such as environmental influences for certain diseases are also being evaluated.
- Clinical and Experimental Studies
- Epidemiological / Explorative Studies
- Sample Estimation / Randomization
- Statistical Consulting
Numerics supports companies in the pharmaceutical industry, medical device area, and the health care industry in general in the study phase of the registration as well as post-marketing surveillance. We develop study plans, the study design and statistical modeling, and accompany your projects through to editorial work in study reports.The statistical design of scientific experiments is dealing with questions such as: Which test units are to be chosen? Can treatments be “blocked” within experimental units? Can certain factors be examined in combination (so-called “Factorials”)? A well-planned experiment saves costs because with minimum effort a maximum amount of information is gained.We always carry out thorough sample size calculations to estimate the number of observations needed. By means of an optimal experimental design, the variability can be reduced, whereby the sensitivity for detecting effects is increased accordingly.
By means of epidemiological studies, health risks can be assessed with respect to the population. Based on various specific study designs (cohort studies, case-control studies, cross-sectional studies), the causality of possible causes (environment, occupation, lifestyle, etc.) can be tested quantitatively. Hereby confounding factors such as age and sex distribution as well as possible interactions have to be accounted for.
In purely exploratory studies, there is no specific target variable with an appropriate working hypothesis. However, an attempt is made to reveal “structures” and to find patterns in the data, usually with graphical methods (e.g., correspondence analysis).
To test a scientific hypothesis it has to be ensured that the study has sufficient power and that the study sample being selected is representative of the underlying population. An optimal sampling design (either complete random, stratified, fixed or sequential) is guided by the research question. This is done to prevent a situation in which not enough data is collected resulting in an otherwise reasonable hypothesis being unable to be evaluated successfully. At the same time it should to be avoided that too many cases are being collected. This is why sample size calculation saves time and money and ensures that a sound statement about the study question can be made. By means of randomization lists it is ensured that the allocation to the treatment groups is made randomly.
As the first service that was provided by our company, Numerics has it’s roots in statistical consulting Working together with you to discuss questions, evaluate appropriate working methods, and then analyze the data in a way that the results generate real added value – these are tasks that fascinate us over and over again. Statistics being an interface between mathematics and science commits somehow to draw up a clear hypothesis, to design testable models and fosters objective thinking. We have a broad experience in various bio-medical disciplines (Pharmacy, Prostheses, Implantology) to draw upon in order to find solutions for your studies.