Josh Malone
Updated October 22, 2020

This study utilizes queries in Docket Navigator (docketnavigator.com), a service that reviews and classifies court decisions for various parameters. This service is widely used and cited by professionals in the patent industry. In this case I used the Patent Library and the Determinations code. Docket Navigator codes PTAB Final Written Decisions as either “Not Unpatentable” or “Unpatentable/Cancelled”. I only looked at Final Written Decisions.

How Many Issued Patents Have Been Challenged at the PTAB?

First, I ran a query to find that 6,663 (UPDATE 7,615) patents have been involved in cases at the Patent Trial and Appeal Board.

How Many Issued Patents Have Been Fully Reviewed by the PTAB?

Then I narrowed those results to the 2,534 (UPDATE 2,925) patents that had one of the codes for a Final Written Decision – either “Not Unpatentable” or “Unpatentable/Cancelled”.  

What Can We Concluded About Patents Not Fully Reviewed by PTAB?

There is much debate over interpreting the remaining 4,129 patents that began, but did not complete a PTAB review. The USPTO issues a quarterly report with a “waterfall” chart tracking the number of petitions at each stage of the proceeding.

While this report tracks PETITIONS rather than PATENTS, it provides some insight into what likely happened to the 4,129 patents that were challenged but did not reach a Final Written Decision. Here is the summary:

12%        1,266     Pending Petitions (665 + 601)
31%        3,157     Undetermined Petitions (1,409 + 115 + 42 + 1,197 + 72 + 322)
28%        2,926     Denied Institution
29%        2,955     Final Written Decision
100%     10,314 Unique Petitions (10,819 Petitions – 515 Joined)

Again, these are PETITIONS, merely a proxy for what happened to the 4,129 PATENTS that did not reach a final determination. There is a great deal of controversy on how to interpret these results. Some analysts count settlements and denials as a win for inventors, but I disagree. First of all, many patents in those categories are later invalidated in a parallel or subsequent trial. Secondly, settlements usually mean the challenger gets to keep infringing with little or no compensation paid to the inventor. Finally, a denied petition does not provide any assurance that the patent would be upheld in a subsequent challenge, thus should not be included in the denominator when calculating the invalidation rate. A patent denied institution is almost the same as a patent that has never been challenged – at substantial risk of future invalidation by the PTAB.

The bottom line is that we know very little about the 4,129 patents that did not reach a final decision. They are pretty much on the same footing as the 2.5 million patents that have never been challenged.

What is the Invalidation Rate on Patents Fully Reviewed by the PTAB?

Accordingly, I am focusing on the 2,534 (UPDATE 2,925) patents that have been subject to a PTAB Final Written Decision. Of those patents, we find that 2,138 (UPDATE 2,469) of them have been determined to be “Unpatentable/Cancelled”.

The net result is 84% of patents that have been fully reviewed by the PTAB are found to be invalid. The USPTO data on petitions indicates that a fraction of these patents would have some of the claims invalidated, while a majority of them would have all challenged claims invalidated.

Are Things Getting Better at the PTAB?

The trend on patent validity rates is beyond the scope of this study as it is not readily accessible on Docket Navigator. One relevant data point that is available is the petition institution rate published by the USPTO.

According to the latest USPTO report things are not getting better, as the institution rate has settled in at 60-63% (disregarding initial FY20 for small sample size).

Conclusion

This report demonstrates one methodology for assessing the invalidation rates of the PTAB. According this method 2,138 of the 2,534 patents that have been fully reviewed by the PTAB are determined to be invalid, resulting in an invalidation rate of 84%. Questions or corrections should be addressed to josh@usinventor.org.