|
|
|
|
|
|
|
data is incorporated in papers that are rapidly put together. All too often industry waits for an investigator to come forward and publish a small excerpt of the database in which they were involved. This can give rise to a misleading impression of the data or the publication of data that does not have the power to stand statistical scrutiny. I have heard the FDA saying that the published data is unimportant and has no relation to the NDA they have just reviewed and, in fact, the published data is of little use. I find that abhorrent. It should be that the published data would be consonant with that in the NDA, and that the FDA, as physicians who are going to use the drug, could look at information in the public sector, evaluate it, and make a judgment. The scientific data in the public domain should be complete, accurate, and informative. That is not to say that the NDA would not be far more complex and thorough than a publication. That is to be expected, but the implication that the information is misleading or vastly incomplete is a considerable disservice to the scientific community and to the physicians who will use the drug clinically. This is especially important in terms of secondary claims or off label uses of a new drug that often will never undergo adequate study. Indeed, if the medical literature were more complete, the FDA could more readily rely on this information and permit its dissemination to physicians with the understanding that the information is not definitive and not the subject of FDA critical review. |
|
|
|
|
|
|
|
|
While there have been a number of attempts to fully automate the NDA process with electronic data entry and electronic NDA submission, there have been limitations to most systems that have been undertaken. However, this has been tested only a very short period and certainly over the next few years electronic data entry, special data-handling techniques, automated quality assurance data handling, and NDA automated submissions will be the rule, not the exception. |
|
|
|
|
|
|
|
|
The FDA reviewers will require considerable computer assistance. They will be able to rapidly cross-check the data bases to ensure data integrity, completeness, as well as establish a system that will permit checking parts of the data analysis and summaries for validity without going through a line-by-line analysis. Once this is done, the work of the FDA reviewer will be vastly simplified and he will be able to work with his superiors on-line through E-mail systems and through automated data-review algorithms that will permit a timely and much more thorough and complete evaluation of the database. Having the reviewers recreate the NDA piece by piece, make their own reviews and own analysis, and having the NDA broken up and reviewed individually in different sections of the FDA is a very laborious, time-consuming process that is at the heart of the slow-moving NDA review process. This could be expedited in the information age and this will be a great contribution to speeding up the process as well as ensuring the validity of the review. |
|
|
|
|
|
|
|
|
At the completion of the submission of the NDA, one has to continue to consider a number of aspects in drug development. There will be ongoing |
|
|
|
|
|