I'm exploring JMP Platform options for analyzing accelerate reagent stability studies. A typicall study would consist of the following. Store reagents at 4 or 5 different temperatures (e.g. 15 degC, 20 degC, 30 degC, 35 degC, 40 degC). The "activity" of of reagent (Concentration of a molecule, activity of an enzyme, etc.) would then be measured at 4-5 time points (e.g. 1 day, 4 days, 7 days, 10 days, 20 days). The stability of reagent samples that are stored at lower temperatures typically are measured over a longer period of time. The actual reagent would typically be stored refrigerated at -20 degC or -80 degC. At thise temperatures, the reagent would typically last for 12-36 months. However, during new product development, we can't afford to wait 12 months to get an initial esitmate of stability, so we want to use Accelerated stability testing.
Once the study is compleed and the data is available, the goal is to use the Arrhenius model where temperature is the accelaration factor to predict product stability. I'm trying to figure out which platform in JMP would be the best choice to analyze this type of data. I'm currently reviewing the Reliability and Survival > Degradation platform. It looks like the Stabiltiy Test Tab would be used for Real-Time stabiltiy studies, but not for accelerated stabilty studies. I would like to calcluate is the Q10 value and the 95% confidence interval on the Q10 value. This Q10 value would then be used for future Stabiltiy Studies for the same or similar reagent to predict reagent stability.
If anyone on this forum has experince using JMP to analyze accelerated reagent stability studies for reagents, your guidance would be appreciated.
It generally sounds like you are headed in the correct direction. Are you following Q1E / Q1A guidance or CLSI protocols?
The Degradation platform offers a special stability analysis that follows the guidance above. It assumes the default alpha = 0.25, not 0.05, for poolability. This assumes that you are determining expiry from data collected at recommended storage. I would use the regular degradation analysis to model the acceleration factor instead of this stability analysis.
Q10 is not an unusual statistic but you might want to use the Life Distribution platform instead. Use the Degradation platform first to estimate the crossing times (i.e., expiry) at recommended storage and then save the crossing times. Now use the Life Distribution platform to fit a distribution model to the crossing times from which you can estimate the point or the interval for Q10.
Please follow Ian's advice and look at the presentation.