This week our Data Science Team decided to go one step further and review the Institution Denial rate for patents challenged from 2012 through 2015. We learned that similar to case-level analysis we performed last week, the Institution Denial Rate at the patent-level has also been rising year-over-year. However, it is generally lower than the case-level rate and in fact, the overall number is closer to 33 percent compared to 38 percent we reported for the case-level analysis we performed last week.

Last week we released the Institution Denial Rate for all Inter-Partes Review (IPR) challenges filed from September 2012 through December 2015. We analyzed over 3,954 cases and compared that with the AIA Trial Statistics released by the USPTO. We concluded that depending on how we look at this data, the actual institution denial rate is closer to 38 percent or in fact 10 percent lower than 48 percent reported by the USPTO in their October stats.

However, with the patent-level study we performed this week, it seems that the institution denial rate is even lower and out of every three unique patents challenged from 2012 through 2015, two of them have been affected negatively by the IPR process (e.g., invalidated, instituted (pending) or settled out of weakness).

The above chart compares the institution denial rate at the patent and case levels. The reason that the case-level rates are higher than the patent-level is because in some cases, a single patent may be challenged multiple times and for example, while it survives the first and the second IPRs, some or all of its claims are invalidated by the third IPR. In other words, we only need one successful IPR to invalidate one or more claims of a patent regardless of how many IPRs are filed. Because of this, the case-level institution denial rate is generally higher than the patent-level.

To calculate and illustrate the patent-level institution denial rate, we reviewed all the cases from 2012 through the end of 2015. This included total of 3,954 cases. However, these cases only challenged 2,495 unique patents. In other words, about 37 percent of IPR challenges were filed to challenge an already challenged patent. In some cases, they were filed by multiple petitioners while in other cases, the same petitioner might have filed multiple IPRs due to different reasons such as the page limitation imposed by the PTAB.

To find the status for each patent, we used the status of their related cases. For example for those patents that had only one related case, the same status was used for the patent. However, if there were more than one case related to a patent, we had to decide which status properly describes the status of the patent. The following table shows the logic we used in selecting the patent status. For example, if there were two cases for a single patent, and the first case had reached the Final Written Decision (i.e., “FWD Entered” status) and the second case was “Terminated-Denied”, we selected the status of the first case (i.e., FWD Entered) as the status for the patent since most likely one or more claims of the patent have been invalidated.

Status of Case 1	Status of Case 2	Patent Status
Terminated-Denied	Terminated-Denied	Terminated-Denied
Terminated-Settled	Terminated-Denied	Terminated-Settled
Terminated-Dismissed	Terminated-Denied	Terminated-Denied
FWD Entered	Terminated-Denied	FWD Entered
Instituted	Terminated-Denied	Instituted
PO Response Filed	Terminated-Denied	PO Response Filed

The following chart compares the ratio of those patents with the Terminated-Denied status versus all other patents calculated for each year separately. As you can see, there is a gradual rise in the institution denial rate.

To calculate the overall institution Denial Rate, we reviewed all the 2,495 unique patents challenged in 3,954 IPR cases together and using the above mapping table assigned the equivalent patent status to each patent. We learned that only about 33 percent or one-third of the patents have not been affected by IPR challenges.

By reviewing the IPR challenges at the patent-level, we removed those cases that had artificially inflated the institution denial rate. This usually happens when multiple IPRs are filed to challenge a single patent and for example only one is instituted while the rest are denied. We learned that compared to the case-level, the institution denial rate is lower by a few percentage for each year and in total it is about 33 percent meaning that out of every three patents, at least two of them have been negatively affected by IPR.

In the coming weeks, our Data Science Team is going to expand this study and look at the IPR cases at the claim-level. This can potentially show the true invalidation rate (claims listed in the petition vs. those invalidated). As you may have noticed, we are gradually adding the challenged, instituted and invalidated claim numbers for each of the cases to Patexia Lawsuits and PTAB Database. This is performed through manual and automated review of PTAB documents including Institution and Final Written Decision documents. We plan to release a comprehensive report for all IPR cases which includes review and ranking of judges, law firms and attorneys based on both quantity and quality of their work at the case, patent and claim levels.