Extent of Extrapolation

The uncertainties in extrapolated data should increase as the extent of extrapolation increases. Since equation (1.15) does not always give this result, manual adjustment is sometime required in this range. 

Extent of Extrapolation For a regression-like QSAR, a simple measure of a chemical being too far from the applicability domain of the model is its leverage, hi [36], which is defined as... 

Product to be Stored at Room Temperature For products to be stored at room temperatures the extent of extrapolation will depend on the evaluation of the results from the short-term, intermediate, and long-term data. If no significant change at the accelerated (short-term) condition is observed, then the long-term data are evaluated for change and variability ... 

Data for which statistical analysis does apply When statistical analysis is applicable to the long-term data and it is not performed, a justification is required and the extent of extrapolation is the same as when statistical analysis does not apply. If the statistical analysis is performed, the extent of extrapolation can be up to twice but not more than 12 months beyond the period covered in the stability study for the long-term data. 

If significant change at any time during the 6-month period of the accelerated (shortterm) condition is observed, then stability testing at an intermediate condition is required and the extent of extrapolation will depend on the outcome of both the long-term and intermediate conditions ... 

Data with No Significant Change at Intermediate Conditions In this case extrapolation of the long-term data can be proposed and the extent of extrapolation will depend on whether a statistical analysis is applicable. 

Product to be Stored In a Refrigerator For products to be stored in a refrigerator the same data evaluation approach but with more restrictive extent of extrapolation as for product to be stored at room temperature will be used unless otherwise specified below. 

Long-Term Data with Change over Time and Variability The extent of extrapolation depends on whether statistical analysis is apphcable to the stabihty data ... 

If long term data are amenable to statistical analysis but no analysis is performed, the extent of extrapolation should be the same as when data are not amenable to statistical analysis. 

If there is no significant change at the intermediate condition, extrapolation beyond the period covered by long term data can be proposed however, the extent of extrapolation would depend on whether long term data for the attribute arc amenable to statistical analysis. 

When the long term data for an attribute are amenable to statistical analysis but no analysis is performed, the extent of extrapolation should be the same as when data are not amenable to statistical analysis. However, if a statistical analysis is performed, the proposed shelf life can be up to one-and-half times, but should not be more than 6 months beyond, the period covered by long term data, when backed by statistical analysis and relevant supporting data. 

Even though each step of the corrosion test procedure is important, if not properly evaluated the results can be misused or can lead to erroneous conclusions. It is important to remember that the corrosion rate is based on the extrapolation of results from short-term tests. For example, corrosion rates may be reported in the terms of millimeters per year (mmpy) on the basis of results obtained from exposure of one month, one week, or one day. The degree of extrapolation rcmges finm a multiple of 12 to 365. Since corrosion is rcirely Unear, with initial rates being extremely high, it is importcmt that the extent of extrapolation be considered in the evaluation phase of the test program. 

---