We offer fully Bayesian, adaptive dose-escalation for single agent or combination treatments incorporating pre-clinical or existing historical information. The models we use have been shown repeatedly to work in practice.
The use of Bayesian statistics for medical devices has been particularly endorsed by the FDA. Especially in situations where historical information is available, the FDA may accept the use of this information when there is sufficient evidence to assume exchangeability.
Bayesian approaches are of special interest in the field of Diagnostics since they allow accounting for uncertainty of underlying parameters. We offer support to incorporate existing information through the use of Bayesian statistics and evidence synthesis.
Cogitars provides statistical support for innovative development of drugs, medical devices and diagnostics.
The company's co-founder have gained specific expertise in implementation of international, fully Bayesian adaptive Phase I/II trials during their career at Novartis, the leading company with regards to Bayesian adaptive doseescalation. For later phases of development and for special fields such as orphan diseases or biosimilars, Cogitars provides a competitive advantage through the use of methods for evidence synthesis. Furthermore, the co-founders' in-depth understanding of business needs, their proven track record in the pharmaceutical industry and their solid scientific background make Cogitars a best-in-class partner to support your development. Cogitars GmbH is privately owned.
Cogitars was honored at the HAE accelerator program.
Here is an article on the website of BioPro, the life science association in Baden-Württemberg. (german)
For the english version follow this link.
Cogitars offers tailored support, consulting & coaching as well as education services. With its services, Cogitars can offer specific solutions in order to optimally meet the needs of its customers.
Specific needs usually differ between projects and studies. Therefore, we evaluate our customers' needs first and
then offer solutions tailored to their needs. Examples include support for initial study planning, protocol writing
and final reporting for all Bayesian analyses of phase I/II studies and full services ranging from planning, statistical
analyses, to interpretation and reporting of projects relying on evidence synthesis such as mixed treatment
Our support ranges from highly specialized technical solutions to working with diverse teams to help them implementing adaptive, Bayesian trials.
Many pharmaceutical companies nowadays would like to take advantage of Bayesian methods for drug development. However, they often lack internal practical experience and thus may not be able to reliably quantify feasibility e.g. with regards to regulatory requirements. In these situations, based on our proven track record, we offer specific consulting to assess feasibility and to optimally prepare for potential challenges. Furthermore, we help companies to build know-how in-house by coaching individuals and teams in implementing these methods.
We provide specific on-site trainings for companies depending on their needs. Our trainings are based on the experience
that real world examples, understanding fundamental concepts and – where appropriate – using homework
dramatically increases the learning capability of the participants.
We offer specific trainings for statisticians and non-statisticians. For statisticians, these consist of e.g. fundamentals of Bayesian statistics, Bayesian adaptive dose-escalation including interactions and negotiations with internal and external stakeholders, evidence synthesis including mixed treatment comparison and prediction. For nonstatisticians, we offer e.g. introduction to Bayesian dose-escalation and fundamentals of evidence synthesis.
Our specific expertise includes Bayesian statistics, evidence synthesis, interaction with regulators, managing transition activities for compounds and strategic support in development of compounds. Statistical expertise on its own is insufficient to best support drug development. Only if statistical expertise is combined with operational know-how, a good understanding of business needs and the regulatory environment it becomes a powerful tool. That is what you find at Cogitars.
Transition of compounds from early development to late development (for approval) has changed dramatically during the past few years. Nowadays, transition activities often start already in phase I and do not end with phase II. We have specific expertise in managing transition activities for compounds with multiple on-going trials at different stages and in different indications. We can help in prospectively plan for these activities and in designing trials which allow for quicker, more cost-efficient decision making.
Cogitars offers support for Health Technology Assessment through the use of standard (such as meta-analysis) as well as advanced (e.g. mixed treatment comparisons) methods for evidence synthesis. Furthermore, we are experts in using the Bayesian approach which allows to formally incorporate information from different sources (e.g., not necessarily randomized studies only) to build more advanced models integrating evidence from different sources, especially powerful for predictions.
Especially in early drug development, Bayesian statistics can play a very powerful role by improved, model-based dose-escalation. We offer fully Bayesian, adaptive dose-escalation for single agent or combination treatments incorporating pre-clinical or existing historical information. The models we use have been shown repeatedly to work in practice. Furthermore, we provide implementation of successfully used designs for phase II trials which allow reduced sample sizes through the use of historical information, either in single arm or in randomized trials. These methods may also be used in phase IV, single-arm or randomized/cluster-randomized trials.
In development of biosimilars, where the working hypothesis of similarity is the base for any development, hierarchical models can become very important in order to combine evidence from multiple trials (e.g. in different indication). This can speed up the clinical development of biosimilars substantially. For such approaches, the Bayesian approach is particularly useful and based on previous work we can offer different solutions.
The use of Bayesian statistics for medical devices has been particularly endorsed by the FDA. Especially in situations where historical information is available, the FDA may accept the use of this information when there is sufficient evidence to assume exchangeability. Another important approach is the use of putative placebo where we offer special support relying on methods for mixed treatment comparisons.
Bayesian approaches are of special interest in the field of Diagnostics since they allow accounting for uncertainty of underlying parameters (e.g., prevalence or incidence). For diagnostic tests, this becomes especially important when no gold standard is available, or when multiple tests are available against which one needs to compare. For these situations, we offer support to incorporate existing information through the use of Bayesian statistics and evidence synthesis.
Nowadays, clinical trial simulation based on multiple sources of information such as PK, PD and historical data has become more and more important. However, statistical teams confronted with this task often face challenges. Based on our extensive experience in clinical trial simulation to assess operating characteristics, we offer strategic as well as technical support to avoid and overcome challenges.