One of the more productive experiments in government at the USPTO in the past decade has been the Central Reexamination Unit (CRU). Although the fruits of the experiment have not always been apparent, the history preceding the formation of the CRU confirms the logic underlying its existence, and the history subsequent to its formation confirms its productive impact. This article examines the history and structure of the CRU, as well as recent reforms that have led to decrease in the pendency of inter partes reexamination proceedings.

Genesis of the CRU5

Prior to 2005, there was no central administration of reexamination proceedings. Reexaminations would be handled alongside applications according to technology. Often, the same Examiner who examined the original application would be assigned the reexamination,6 to be conducted along with his or her regular docket of applications. Reexaminations, being far fewer in number than applications, were oddball proceedings that were difficult to docket and difficult to account for in terms of productivity. Special Programs Examiners (“SPRE’s”) were required to review papers and attempt to ensure that reexamination-specific rules were followed. Because of the large number of Examiners relative to the number of reexaminations, no core group of Examiners was able to develop a body of experience that would allow for identification of flaws in the system and incremental improvement. This made the job of the Office of Patent Legal Administration, the de facto organizationproviders7 of reexamination proceedings, significantly more difficult than it otherwise could have been.

The CRU was formed in July 2005 to address these problems.8 At that time, the vast majority of new requests were assigned to Examiners within the CRU. Non-CRU examiners were permitted to keep cases that they wanted to keep and they had already done extensive work on. Today, the Examiners of the CRU handle approximately 2000 reexamination cases pending at any one time,9 with approximately 80 cases remaining in various technology centers.10

Present Status And Organization of the CRU

The CRU is more an organizational structure than a place. Although there is a nucleus of CRU management, attorneys and support staff housed in one location in the Madison building of the USPTO in Alexandria, CRU Examiners are either housed in their former Technology Centers, or work outside of the Office as part of the PTO’s hoteling program.

The nucleus housed in the Madison Building includes the Director of the CRU, and Supervisory Examiners, and is immediately adjacent the principal members of the Office of Patent Legal Administration. This allows for members to draw on mutual experience, and makes significant decisions a far more collaborative effort than might be suspected by decision signatures alone.

The organizational structure that more defines the CRU than its location is organic in appearance, but has been cultivated according to the collective experience that was largely lost during the first two-and-a-half decades of reexamination. Five supervisory Examiners oversee 59 Examiners, all of whom are Primary Examiners, and all of whom are paid at one grade higher (GS-15) than Primary Examiners in other units. Each Supervisory Examiner supervises a group of Examiners within a broadly defined technical field, with three of the five Supervisory Examiners dealing with electronics, software and business method patents, one Supervisory Examiner dealing with chemical patents, and one Supervisory Examiner dealing with mechanical and consumer goods patents.

Importantly, the CRU has a number of paralegals who conduct initial reviews of ex parte and inter partes requests to determine formal compliance with the reexamination rules. If their review shows non-compliance with the rules governing the content of a request, they will send out a “Notice of Incomplete Request”. Such notices are issued frequently.11 The notice has the effect of resetting the PTO’s three month clock for issuing an order granting or denying reexamination. The clock does not start ticking again until all formal deficiencies are corrected. Since reexaminations are often requested in response to ongoing or threatened litigation with a goal of achieving a rapid rejection of the patent claims, it is often important for practitioners to meticulously comply with the formal requirements for requests. Otherwise, they will encounter delays12 that may diminish the strategic value of requesting reexamination.

Once a new request has successfully through formal compliance review by the paralegals,13 it is assigned to a single Examiner by the relevant Supervisory Examiner. There are few formal criteria for doing so, rather the Supervisory Examiner makes a judgment as to who has the most capacity and most familiarity with the subject matter. Multiple co-pending cases on the same family of patents tend to be assigned to the same Examiner, and indeed some Examiners spend a significant percentage of their time dealing with requests related to a single licensing effort. Where a prior reexamination has been completed, however, a new reexamination request will not be assigned to the same CRU Examiner.

Administration in the CRU is perhaps surprisingly flexible, when compared to the rest of the PTO. The USPTO as a whole is an organization built to handle massive amounts of throughput. Its processes and information systems are built around the need to keep track of a very large number of active applications, and are of necessity highly automated. In contrast, the CRU handles only around 2000 proceedings at a time, including, currently, about 900 new requests each year. This allows the CRU the luxury of human decision-making on the administrative aspects of individual cases.

One dramatic consequence of this flexibility is the lack of a count system within the CRU. Examiner productivity is not bound to such metrics, because a much more detailed approach is possible to evaluate quantity and quality of work product. Examiners are accountable to their supervisors in terms of quantity, certainly, but the time allotted for any particular case can be tailored to the circumstances of that case. The “allotting” is done by the Examiner, so long as the Examiner can justify the time spent on the case. Supervisors watch for cases that are taking particularly long to advance, or Examiners who have not issued actions in a long period of time, and intervene on an as-needed basis. Quality is reviewed as a part of the process— every single action is reviewed by a supervisor and additional conferee,14 so random quality checks are unnecessary. This kind of time-intensive review is commensurate with the importance of reexamination in the overall schema of PTO operations.

CRU Performance In Inter Partes Reexamination

If the pre-CRU level of assembled experience was low with respect to ex parte reexamination, it was nearly non-existent with respect to inter partes reexamination. In the years prior to FY2005, only 53 requests for inter partes reexamination had been filed. When the proceeding was implemented in 1999, no one could be certain how the new rules would play out. Due to the low number of requests, this level of uncertainty was inherited by the CRU when cases were transferred in early 2006.

Pendency Of Inter Partes Reexamination

Inter partes reexamination, then, came of age in the CRU. Its adolescence has not been without its share of stumbles, of course, many of which have been the subject of recent legislative proposals. One of the primary criticisms15 has been directed toward pendency. While official PTO statistical summaries show a respectable pendency of some 32-35 months to a reexamination certificate,16 these summaries are inappropriate tools for predicting the length of a fully contested proceeding, because they by definition only include those proceedings that are fast enough to have concluded. Studies that focus on predicting the length of a fully contested proceeding, including one conducted by one of the authors,17 showed that significantly longer periods could be expected. Earlier this year, three of the authors reported evidence that much of the delay in inter partes cases appeared unrelated to Examiner workload issues.18

The results of a more recent study by the authors, however, tends to show that the CRU has made much progress against Examiner-level pendency issues. Part of the solution has come by way of improved procedural tracking and management within the CRU. The CRU uses the same underlying IT infrastructure as the rest of the PTO – it is supplied to the Unit with no particular optimization for reexamination proceedings. Over time, however, CRU management has adapted the systems, adapted to the systems, or created its own ad hoc systems in a way that allows more effective management and tracking of individual cases.

Better facilities for tracking cases, together with more aggressive docket management and a number of procedural tweaks implemented in 2008, have led to a truly remarkable decrease in delay in areas largely uncorrelated to Examiner workload. These areas include the time to publish a reexamination certificate after all decisions on the merits have been made, the issuance of formal deficiency notices, the issuance of actions after the Patent Owner fails to respond, and the termination of proceedings after a substantive action giving grounds for termination has occurred.

In areas directly related to Examiner workload, however, the CRU has also made significant progress, especially in the past two years. The authors studied a few metrics that appear to show significant progress in addressing Examiner-level pendency issues.

In the first study, the authors studied the time to issuance of the order on the request for reexamination in inter partes reexamination. The PTO is constrained by statute to issue the order within three months19 of the accorded20 filing date. In early cases, the PTO required, on average, nearly 80 days to issue orders.21 More recently, however, the PTO appears to have shaved several weeks off the average time to order, lowering it to the two-month mark. The attached graphic shows the approximate trend:

Click here to view graph

The graphic above was prepared by the following methodology. First, the request closest to the beginning of each calendar quarter was chosen. The average time to order in days was calculated for the 30 requests surrounding that order, i.e. +/- 15 requests on each side. The results were plotted according to quarter number above, using circles to mark data points. A least-squares-fit trend line, coincident at mid- 2003, was also plotted.

For 2003 and 2004, there were so few requests on file that a window of +/- 15 requests could not accurately represent a single quarter’s worth of data. For 2008 and 2009, a 30-request window usually represents fewer than all the requests filed in a quarter. The entire-year average for 2008 and the available data average for 2009, however, were similar at about 67 and 58 days, respectively. Among the more notable features of Fig. 1 is the apparent sharp drop in days to order that accompanied the assignment of new requests to CRU Examiners and transfer of other cases into the CRU in 2005-06.

The authors also studied the time to the Right of Appeal Notice in inter partes reexamination over the past decade. The Right of Appeal Notice is the final action issued by an Examiner in inter partes reexamination, and reflects the conclusion of Examiner-level prosecution before the Office.22 Thus, the time to the Right of Appeal Notice reflects the time taken by the CRU to process cases.

Because it is difficult to estimate the average time to the Right of Appeal Notice for recent cases that may not have reached that stage, the authors instead reviewed the percentage of cases that reached the Right of Appeal Notice within 18 months of the accorded filing date, studying cases with accorded filing dates through May, 2008 (or 18 months prior to the study date). In the years prior to 2005, around 10% of all cases reached a Right of Appeal Notice within 18 months. Over time, and particularly within the last two years, this proportion has tripled to about 1/3 of all cases, as shown in Fig. 2: Click here to view

Figure 2 was created by counting the number of cases with accorded filing dates in each half-year, for example, Jan 1, 2004 – June 30, 2004, excluding cases where reexamination was denied or procedurally terminated, and then counting the number that had issued Right of Appeal Notices within 18 months.

These statistics are viewed from the perspective of users of the system, who do not have a view into the machine that produces such output. Two important levers on that machine might also be examined, however, if one wanted to see how efficiently the system is performing. These are the number of Examiners available to examine inter partes reexaminations, and the total number of requests being filed. Both numbers are increasing. In PTO Fiscal Year 2005, there were 20 CRU Examiners and 59 requests. In PTO Fiscal Year 2009, there were 55 CRU Examiners23 and 258 requests.

The authors computed an ad hoc efficiency metric by multiplying the values shown in Fig. 2, above by the number of requests used to compute the value (resulting in the number of requests reaching a Right of Appeal Notice within 18 months again), and then dividing by the number of Examiners handling the cases (which can be different from the total number of Examiners in the CRU). The results, scaled to an efficiency of 1.0 for the first half-year of 2004, are shown in Fig. 3: Click here to view

The authors caution that the methodology is too crude24 to draw precise conclusions, such as that Examiners are now six times more efficient than five years ago. However, one probably can draw the conclusion that the Examiner-level process is getting faster, and the increase in speed is due at least in part to better management of the resources allocated to reexamination and improved efficiency on the part of individual Examiners. Although the pendency levels are significantly higher than the aggressive goals set by the current Directorship, there is evidence that positive change has been occurring. Steady, positive change is an indicator of management that is both aware of the issue and making progress against it. Good management in turn engenders the hope that progress will continue until inter partes reexamination is significantly faster than district court litigation, at current levels of quality.

Variability In Inter Partes Reexamination

One of the benefits of the CRU contemplated by its founders was an increase in the predictability of the proceeding. By bringing together specialist Examiners under a single management unit, it was hoped that the application of the rules by Examiners would become more consistent (both between different Examiners and between different cases assigned to the same Examiner). It was also hoped that quality would increase overall.

The authors have reviewed a number of inter partes cases for indications of inconsistency between Examiners in a few key areas, including rejection rates, claim construction, and the use of evidence. Although the authors are aware of numerous differences between individual Examiners, the results, speaking subjectively, do not show a lack of consistency that raises any concerns about the integrity of the process.

For rejection rates among Examiners, the authors looked at Examiner treatment of claims in initial office actions by subject matter area, owing to different levels of predictability in different fields. The results for Examiners that have handled a sufficient number of electronics and software cases to be at least somewhat meaningful, is shown in Fig. 4: Click here to view

The results show a relatively consistent high rejection rate in electronics and software cases. The authors caution that, while Examiners with few mailed office actions were filtered out,25 a number of the Examiner results shown above rest on the examination of only a few dozen claims, while others rest on thousands. Nor is it entirely clear that, given a sufficiently large number of cases, that the initial rejection rate should approach a consistent mean.

Examiners’ handling of claim interpretation was not particularly susceptible to numerical analysis, and is thus presented here by way of the authors’ subjective conclusions. Subjectively, there was indeed some variation between individual Examiners. There was natural variation in the legal interpretation of the “broadest reasonable interpretation” of the claims, which is to be expected given the slippery nature of that standard, as well as differences between subject matter. Probably the greatest individual variation, however, was the willingness of Examiners to discuss their claim interpretations on the record and provide reasons therefor. Some Examiners give detailed reasons for accepting a claim interpretation put forward in a request, some give detailed reasons for rejecting interpretations put forward in a request, but adopt alternate interpretations only implicitly, some adopt alternate interpretations expressly but reject proferred interpretations implicitly, some Examiners comment extensively on both, some do neither, and some vary the approach depending on the situation. To some extent, these varying approaches can be used by Requesters to put claim interpretations in front of Examiners where the Examiner’s interpretation is important to the outcome of litigation.

Examination of evidence, like claim interpretation, is also not particularly susceptible to statistical presentation. Subjectively speaking, the Examiners’ approach to evidence is remarkably consistent, although individual decisions vary on the merits.

As for declaration evidence, recent requests filed with declarations were identified through keyword searching. Declarations that go to pure facts are generally accepted until challenged. These range from the truly mundane, such as declarations presented to prove publication dates, to the more substantive, such as declarations intended to show inherent features. The legal conclusions to be drawn from these facts, however, were still open to Examiner interpretation. Declarations that make legal conclusions or factual conclusions based on the interpretation of express subject matter in a reference, however, are less likely to be accepted. These sorts of issues appear to be consistently decided based on their own merits, largely independent of the declarant’s opinion. Although the authors are aware of a number of cases where Examiners noted that such declarations were persuasive, the authors are also aware of cases where such declarations have been found unpersuasive. The authors have not located a single case where the Examiner found on the record that the issue was unclear, and that the declaration provided the deciding factor.

Treatment of one form of evidence – litigation submissions, does appear to vary somewhat by Examiner. Some Examiners state expressly that such evidence is not considered, usually where claim interpretation is concerned. Other Examiners appear to be more comfortable reviewing litigation documents and applying the evidence (usually evidence of a party’s belief as to the proper interpretation of the claims) during reexamination.


Recent data from the inter partes reexamination cases tends to support a decrease in Examiner-level pendency. The authors believe that this decrease in pendency is attributable to the formation of the CRU, and process changes made that have been made over the term of CRU history to increase performance. The authors express their hope that currently contemplated rule changes will further improve quality while decreasing pendency, and in particular pendency before the Board of Patent Appeals and Interferences.