For Industry
Appendix B: Supporting Information for Monte Carlo Analysis
The results of the linear regression analysis that was performed on the seven model inputs are presented in charts and table in Figures B1 through B7. (Note: The FY 2002 5-year average used by FDA covers four years and not five years due to the lack of data available for FY 1997.)
 |
 |
Figure B1 – Regression Analysis on Labeling Supplements
Figure B1: 508-Compliant Narrative
|
|
|
Figure B2 - Regression Analysis on Annual Reports
Figure B2: 508-Compliant Narrative
|
|
|
Figure B3 – Regression Analysis on NDA/BLA Meetings Scheduled
Figure B3: 508-Compliant Narrative
|
|
|
Figure B4 - Regression Analysis on NDA/BLA Applications
Figure B4: 508-Compliant Narrative
 |
 |
Figure B5 - Regression Analysis on SPAs
Figure B5: 508-Compliant Narrative
 |
 |
Figure B6 - Regression Analysis on IND Meetings Scheduled
Figure B6: 508-Compliant Narrative
 |
 |
Figure B7 - Regression Analysis on IND Applications
Figure B7: 508-Compliant Narrative
The detailed results of the Monte Carlo simulations for both the baseline model (Figures B8 and B9) and alternative model (Figures B10 and B11) are presented below:
 |
Figure B8 – Histogram of Monte Carlo Analysis Results for Baseline Model
Figure B8: 508-Compliant Narrative
 |
Figure B9 – Details of Monte Carlo Analysis Results for Baseline Model
Figure B9: 508-Compliant Narrative
 |
Figure B10 – Histogram of Monte Carlo Analysis Results for Alternative Model
Figure B10: 508-Compliant Narrative
 |
Figure B11 – Details of Monte Carlo Analysis Results for Alternative Model
Figure B11: 508-Compliant Narrative
Figure B1 – Regression Analysis on Labeling Supplements
The formula for the best linear fit for Labeling Supplements was y = (19.2790 x -37,780) and the r2 value was 0.9357. The linear regression analysis for Labeling Supplements resulted in a mean difference of -17.9093 and a standard deviation of 13.6393.
Figure B2 - Regression Analysis on Annual Reports
The formula for the best linear fit for Annual Reports was y = (12.0430 x -21,336) and the r2 value was 0.4804. The linear regression analysis for labeling supplements resulted in a mean difference of -0.2436 and a standard deviation of 27.0577.
Figure B3 – Regression Analysis on NDA/BLA Meetings Scheduled
The formula for the best linear fit for NDA/BLA Meetings Scheduled was y = (2.5821 x 5562.6000) and the r2 value was 0.0644. The linear regression analysis for NDA/BLA Meetings Scheduled resulted in a mean difference of -0.0752 and a standard deviation of 21.2625.
Figure B4 - Regression Analysis on Number of NDA/BLA Applications
The formula for the best linear fit for NDA/BLA Applications was y = (1.4714 x -2829.5000) and the r2 value was -0.5561. The linear regression analysis for NDA/BLA Applications resulted in a mean difference of 0.0859 and a standard deviation of 2.8400.
Figure B5 - Regression Analysis on SPAs
The formula for the best linear fit for SPAs was y = (49.3920 x -98,757) and the r2 value was 0.9866. The linear regression analysis for SPAs resulted in a mean difference of -0.9171 and a standard deviation of 12.4202.
Figure B6 - Regression Analysis on IND Meetings Scheduled
The formula for the best linear fit for IND Meetings Scheduled was y = (163.2900 x -326,076) and the r2 value was 0.9855. The linear regression analysis for IND Meetings Scheduled resulted in a mean difference of 9.3857 and a standard deviation of 42.7210.
Figure B7 - Regression Analysis on IND Applications
The formula for the best linear fit for IND Applications was y = (202.2200 x -400,125) and the r2 value was 0.9931. The linear regression analysis for IND Applications resulted in a mean difference of 2.5571 and a standard deviation of 36.3143.
Figure B8 – Histogram of Monte Carlo Analysis Results for Baseline Model
The histogram displays the results of the Monte Carlo analysis for the baseline model that are shown in Figure C9.
Figure B9 – Details of Monte Carlo Analysis Results for Baseline Model
This table shows the results of the Monte Carlo analysis for the baseline model. The average workload adjuster value for the baseline model observed after the 10,000 runs performed by the analysis was 3.13%.
Figure B10 – Histogram of Monte Carlo Analysis Results for Alternative Model
The histogram displays the results of the Monte Carlo analysis for the alternative model that are shown in Figure C11.
Figure B11 – Details of Monte Carlo Analysis Results for Alternative Model
This table shows the results of the Monte Carlo analysis for the alternative model. The average workload adjuster value for the alternative model observed after the 10,000 runs performed by the analysis was 3.14%.
