Certificates of correction corrected: their history and retroactive application.

• Author: Priest, Chelsea A.

INTRODUCTION I. CERTIFICATES OF CORRECTION A. History B. Certificates of Correction Today II. THE FEDERAL CIRCUIT'S CURRENT APPROACH TO CERTIFICATES OF CORRECTION III. PROBLEMS WITH THE FEDERAL CIRCUIT'S CURRENT APPROACH A. Direct Effects on Patentees' Ability to Pursue Claims 1. Rely on the uncorrected patent 2. Special concerns raised by reissues and reexaminations 3. Only pursue claims for post-certificate infringement B. Indirect Effects on Patentees' and Competitors' Primary Conduct 1. Patentees 2. Alleged infringers IV. AVENUES FOR RELIEF A. Retroactivity with Intervening Rights 1. Equivalence to reissued patents a. History of reissued patents b. Implications for certificates of correction 2. Read in light of the entire patent statute B. Judicial Correction C. Amend [section] 254. CONCLUSION INTRODUCTION

Over the last ten years, the U.S. Patent and Trademark Office (PTO) has issued an average of 230,540 patents per year. (1) On top of that, the PTO has also issued an average of 659 reissued patents and 595 reexamination certificates each year. (2) And these numbers have only been increasing. (3) As with any bureaucracy, especially one charged with printing such large numbers of publications each year, mistakes happen--a word misspelled or misplaced, a figure left out, an old version published rather than the new, or sometimes even an entire claim or limitation missing from the patent. Luckily, Congress has provided for just such occasions. Section 254 of Title 35 of the U.S. Code allows the PTO to issue "certificates of correction" when, through the fault of the PTO, a mistake appears in a published patent. (4) In total, the PTO issued 25,088 certificates of correction in fiscal year 2014. (5)

A problem arises, however, when the patentee seeks to assert her corrected patent in a court of law. From which date can the patentee make an infringement claim? The date the original patent issued? The date the reissued patent or reexamination certificate issued, if applicable? The date the PTO became aware of the error? Or should courts use some other metric entirely?

The retroactivity problem with respect to certificates of correction is particularly acute in cases in which the error first appears in a reissued patent or reexamination certificate. Normally, a patentee is allowed to sue for infringement predating the issuance of a reissued patent or reexamination certificate as long as the new and old claims are substantially the same. (6) But the Federal Circuit held in Southwest Software, Inc. v. Harlequin Inc. that a certificate of correction is a bright line, barring all claims for infringement predating its issuance. (7) In that case, an appendix had been accidentally omitted from the official patent but later was added back in through a certificate of correction. The Federal Circuit, however, held that the appendix could not be considered in determining the patent's validity in a suit for damages predating the issuance of the certificate. (8) If the same rule applies to certificates correcting reissued patents or reexamination certificates, it could wipe out years of infringement through no fault of the patentee.

Despite the importance of these issues, there is a dearth of legal research on the topic. In fact, this Note provides the first in-depth examination of the history of certificates of correction and finds extensive legislative history that has thus far been ignored by the courts. Using this research as a springboard, this Note then argues that Southwest Software should be reconsidered, or at the very least not extended to certificates of correction as applied to reissued patents or reexamination certificates. (9) Instead, courts should apply the same standard that applies to reissued or reexamined patents: certificates of correction should have retroactive application so long as the corrected patent is "substantially the same" as the uncorrected patent. Alternatively, this Note proposes that courts embrace judicial correction of errors in patents (and the attendant retroactivity) in an attempt to lessen the harshness of a bright-line rule prohibiting any retroactive effect.

Part I traces the long history of certificates of correction, providing the first scholarly analysis of where today's [section] 254 began. Part II reviews how courts so far have approached certificates of correction, specifically in the context of retroactivity. Part III explains why the current approach is problematic even now and how it could be even more problematic in the case of certificates of correction issued to correct a reissued patent or reexamination certificate. Part IV proposes a multipart solution: Subpart A argues that, in light of the newly analyzed history of certificates of correction, courts should allow such certificates to function retroactively, particularly in the context of reissued patents or reexamination certificates, while using the intervening rights doctrine as a safeguard. But even if a court were to bar all certificates of correction from operating retroactively, Subpart B argues that courts should embrace judicial correction of errors in patents, thereby providing another avenue by which patentees can recover for preissuance infringement. Finally, and as a last resort if courts are unwilling to act, Subpart C urges Congress to amend [section] 254 to include the language that the reissue and reexamination statutes have that allows reissued patents and reexamination certificates to have retroactive effect in appropriate cases.

Whatever the approach, the problem is clear: patentees are forfeiting years of infringement claims through no fault of their own, and the threat of such forfeiture can impact other behaviors, such as decreasing a patentee's willingness to draw errors to the PTO's attention for correction or encouraging patentees to pursue more costly alternatives to certificates of correction, such as complete reissues. Armed with a new and fuller understanding of the history of certifi of correction, courts should take this opportunity to allow patentees to claim what is rightfully theirs.

1. CERTIFICATES OF CORRECTION

   1. History

   Mistakes in issued patents have been a problem from the Patent Office's inception. (10) At least as early as 1892, the Patent Office made it a habit to issue certificates of correction when an error in a patent was identified. (11) In doing so, however, the Office was acting without authorization. As a result, in Edison Electric Light Co. v. U.S. Electric Lighting Co., the Second Circuit held that a certificate of correction was entirely irrelevant because "the statute providing for no subsequent alteration [of a regularly issued patent] by the patent office, ... the action of the commissioner in indorsing [the patent] with an attempted 'correction' was without jurisdiction, and wholly void." (12) The court concluded that a mistake in an issued patent could only be corrected through reissuance. (13)

   Nevertheless, the Patent Office continued making corrections to issued patents. At a 1912 House committee hearing on a bill to authorize the issuance of certificates of correction, the Commissioner of Patents testified on the need for such a bill. He explained that the Office had made a practice, for at least the previous thirty or forty years, of correcting errors in patents by issuing certificates of correction. (14) As the Commissioner testified, "[frequently we find wrong words used by the inventor, and also errors made by the office and by the Public Printer.... All of which we issue certificates of correction for." (15) The Assistant Commissioner further testified that, at the time, the Office was finding errors in about 10 patents (out of around 700 printed) each week. (16) The congressmen attending the hearing questioned the need for a statute authorizing such corrections if the Office was already making them, to which the witnesses expressed a concern that a court would find such corrections invalid, citing the Edison Electric case as evidence. (17)

   The Patent Office did admit that its power to correct patents was somewhat limited, though. The Assistant Commissioner, for example, testified that the PTO

   correct[s] errors which occur in patents merely in the phraseology of the specifications and some, excepting material errors, in the claims. We very seldom correct an error in a claim for the reason that is the effective part of the patent and not subject to correction after the patent has once been granted. (18) The Office acknowledged that errors in the claims were just as likely to happen as errors elsewhere, and that "[i]t sometimes happens... that the printers will pull the type or mix up words of the claim so it is utterly unintelligible, or so the claim would have a different meaning." (19) In those circumstances, the Office insisted the patentee's "way of correction is by way of a reissue." (20)

   It was clear throughout the hearing that the Patent Office's main goal was to minimize the costs of errors. The witnesses again and again emphasized the enormous burden placed on the Office of having to identify all errors or reprint entire patents, some of which could span hundreds of pages. (21) And while some congressmen seemed to think the problem of errors could be solved by adding additional proofreaders or requiring a patentee to give a final sign-off on the patent, the witnesses were insistent that such alternatives were too costly and time-consuming. (22)

   Despite these efficiency concerns, even in 1912, some congressmen and witnesses had concerns about giving notice to the public about a change in the patent. The hearing's attendees recognized that even a minor typographical error can "change[] the entire scope and meaning of that patent--absolutely and utterly." (23) To ameliorate these concerns, W.W. Dodge, a witness representing the Patent Law Association of Washington, suggested "that before any correction is made a public notice be...