Fig. 5.

Bootstrap estimates of stability slopes. (A) Potency over time: The X-axis represents the testing time points in months post-manufacture and the Y-axis represents the difference between the upper 99% confidence interval of “current” bootstrapped stability slopes and the lower 99% confidence interval of “cumulative” bootstrapped stability slopes. The green and red lines represent the values of difference at each time point for the first (#11-69F-003) and the second (#1975) lots, respectively. The red horizontal dotted line at 0 represents the specification for acceptance that this difference should not be less than 0. (B) Lot-to-lot comparison: Bootstrap estimates of stability slopes for the first (#11-69F-003), second (#1975) and simulated sub-potent lots. The histogram shows the distribution of 10,000 bootstrap replications of slopes from the first lot (#11-69F-003, in purple), the second lot (#1975, in green), and a simulated sub-potent lot (in yellow). The X-axis represents estimated stability slope from probit regression model, and the Y-axis represents the number of stability slope that falls into the corresponding intervals set by X-axis. The three vertical dotted lines, from left to right, represent the estimated mean stability slopes for the simulated sub-potent, the first (#11-69F-003) and the second (#1975) lots, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)