Analysis of COA Data
The nature of COA data analysis can be wide-ranging and may encompass:
At The Psychometrics Team, we have been analyzing data for over 25 years. That is a lot of experience. That teaches you how things break, why they break, and how to avoid or diagnose such events. It also teaches where the pitfalls are and where the bodies are buried. That enables us to be both precise and fast. Over this time, we have cultivated and codified sets of protocols that guarantee the safety of your data, the precision of your results, and the fidelity to the true patient voice hiding in the data. Our expertise, cultivated over this time includes the following areas useful in analysis of post-approval data:
Conditional & Marginal Models
Mixed Model Extensions
Pattern Mixture Models
Mediation and Moderation and both
Mediation
Mediated Moderation
Moderated Mediation
Instrumental Variable Analysis
Multi-Trait Multi-Method Analysis
Latent Mixture Models
Single Terminal Event
Recurrent Event
Multi-State Transition Models
Joint Models (both terminal and recurrent or both)
Experience with the disease area, or an ability to gain understanding of the disease area is essential. We know this from experience.
The porous transition boundaries in Alzheimer’s Dementia and how this can make effect detection perilous.
Whether it is the fluctuating nature of migraine attacks and how these can contribute to effect suppression and placebo effects.
Consideration of the disease area and careful application of the corresponding first principles is essential to yield sound results.
Without an understanding of these components, analysis will go nowhere, or nowhere good.
The influence of recurrent deterioration and missing data for PROs in oncology and how these must be accounted for to detect long-undiscovered effects.
Or validating a legacy instrument previously rejected by DCOA as a fit-for-purpose endpoint and getting the endpoint into the label.
Have a chat with us
Got any questions? Contact us and we'll be happy to help.