This presentation will give an overview of these new tools and techniques for the non-statistician/epidemiologist. It is vital that a pharmaceutical manufacturer be able to quickly assess these issues on a sound statistical basis, in a way that eliminates bias due to confounding factors as much as possible.
In a number of recent, high-profile situations, urgent drug-safety issues have been brought to the attention of regulators and pharmaceutical manufacturers in a way that requires a rapid assessment of available data. Other safety issues arise on a day-by-day basis, as a result of ongoing spontaneous reports. It is vital that a pharmaceutical manufacturer be able to quickly assess these issues on a sound statistical basis, in a way that eliminates bias due to confounding factors as much as possible. Recently, databases and software tools for investigating the relationship of drugs to safety issues have become available, which utilize large-scale, anonymous electronic-health-record (EHR) and medical-claims data. The large number of patients in these databases facilitates the use of statistical methods such as propensity score adjustment to reduce bias. This presentation will give an overview of these new tools and techniques for the non-statistician/epidemiologist.
Can you describe propensity score matching to a non-statistician, and explain how it reduces bias?
What are the capabilities and limitations of this approach to safety studies?
What are the next steps after analyzing a safety issue using these databases?
How to I manage the resulting information?
What procedures and policies are required to implement the analysis of these databases?
What does the future hold for the use of these databases in drug safety?
In a number of recent, high-profile situations, urgent drug-safety issues have been brought to the attention of regulators and pharmaceutical manufacturers in a way that requires a rapid assessment of available data. Other safety issues arise on a day-by-day basis, as a result of ongoing spontaneous reports. It is vital that a pharmaceutical manufacturer be able to quickly assess these issues on a sound statistical basis, in a way that eliminates bias due to confounding factors as much as possible. Recently, databases and software tools for investigating the relationship of drugs to safety issues have become available, which utilize large-scale, anonymous electronic-health-record (EHR) and medical-claims data. The large number of patients in these databases facilitates the use of statistical methods such as propensity score adjustment to reduce bias. This presentation will give an overview of these new tools and techniques for the non-statistician/epidemiologist.
Areas Covered in the seminar:
What safety information is contained in EHR databases, and how is it extracted?Can you describe propensity score matching to a non-statistician, and explain how it reduces bias?
What are the capabilities and limitations of this approach to safety studies?
What are the next steps after analyzing a safety issue using these databases?
How to I manage the resulting information?
What procedures and policies are required to implement the analysis of these databases?
What does the future hold for the use of these databases in drug safety?
Who Will Benefit:
- Safety and regulatory professionals who work with statisticians and epidemiologists in the analysis of safety signals.
- Executive decision-makers who are considering whether or not to incorporate E.H.R. and claims data as a component of pharmacovigilance, and if so, how.
- Safety consultants who currently rely primarily on tradtional medical review, and who are interested in adding the capability of analyzing E.H.R. and claims data.
Course Provider
Alan Hochberg,
