Decisions by the Court as an Institution; or by the Judge as a Human?

by Dennis Crouch

In Novartis Pharm. Corp. v. Accord Healthcare, Inc., 38 F.4th 1013 (Fed. Cir. 2022), the Federal Circuit made a procedurally embarrassing rehearing decision.  In its original decision the appellate court sided with the patentee on validity–affirming that the claims satisfied the written description requirement. Novartis Pharm. Corp. v. Accord Healthcare, Inc., 21 F.4th 1362 (Fed. Cir. 2022).  That original opinion had been authored by Judge O’Malley with Chief Judge Moore in dissent.  When O’Malley retired, the court silently replaced O’Malley with Judge Hughes and issued a new opinion on rehearing flipping the outcome.  In the new decision, Judge Moore’s dissent became the majority (joined by Judge Hughes) with Judge Linn moving to the dissent.

Now, Novartis is seeking a rehearing — arguing (1) that the original decision was correct; and (2) that the judicial replacement to alter the outcome was improper.  I previously wrote about the petition by Novartis and the amicus briefs in support.  The defendant Accord has now filed its responsive brief in opposition.

The brief makes three arguments to explain why it was proper to appoint a new judge on panel rehearing in a way that flipped the outcome.

  1. First, the dissent (Judge Linn) “did not [publicly] raise any objection to the procedure employed.” Note here that Judge Linn has senior status and now generally an outsider to the courts internal procedure.
  2. Second, Courts largely have authority to make their own rules and procedures, but Congress has declared that an appellate judicial panel must be at least three judges.
  3. Third, the petition grossly overstated common practice when it argued that  the operating procedure of “virtually every circuit” would prohibit this outcome.

In my mind, none of these arguments are particularly compelling.  But, I expect that the court will choose to simply deny the rehearing without further comment.  If the court forces itself to make an express decision, it will only draw further criticism.

It is nice to think of the Federal Government in terms of its various institutions of government, such as the Federal Circuit Court of Appeals or the US Patent & Trademark Office.  That framework provides a sense of stability and coherence.  But, perhaps I should take this case as a reminder that the constitutional structure of our government is not based upon those institutions.  Rather, it centers on the individual humans charged with making the decisions.

Prior Writings on the case:

= = =

Accord’s brief was filed by Paul Skiermont (SkiermontDerby) along with his team of Sarah Spires and Mieke Malmberg.  The Novartis brief was filed by Jane Love (Gibson Dunn). That brief was signed by a host of attorneys, including Robert Trenchard, Thomas Hungar, and Jacob Spencer (all at Gibson Dunn); Deanne Maynerd and Seth Lloyd (both from MoFo); and Willy Jay, David Zimmer, and Edwina Clarke (Goodwin Procter).

42 thoughts on “Decisions by the Court as an Institution; or by the Judge as a Human?

  1. 11

    This is not “a procedurally embarrassing” issue for the Federal Circuit. The court should deny rehearing, as the petition does not warrant a response.

    Courts flip decisions all the time upon a change in the judicial assignment, because judges are human beings and a new judge can of course take a different view of the case. This is far more common in district courts because cases often remain pending for years before a single judge and thus are more susceptible to judicial reassignment. Either way, procedurally speaking, so long as the earlier panel decision here hadn’t reached final judgment (it hadn’t, the mandate hadn’t issued), the court was permitted to revise the decision and change the outcome.

    The judicial reassignment here was caused by Judge O’Malley’s retirement, so this was not a “judicial replacement to alter the outcome,” as Dennis seems to suggest. Once Judge Hughes was assigned to the case, and determined that the original panel decision was wrongly decided, he had two options: (1) go along with the original panel decision out of deference to the original panel, even if he thought it was erroneous, or (2) allow the court to issue a new decision adopting what he thought was the correct outcome. There are principled arguments in support of both approaches, and it’s possible that Judge Hughes may have gone down path (2) if the original panel decision was not precedential. But it was, and in either case, Judge Hughes had no obligation to circle the wagons around a decision he thought was erroneous.

    Whenever there’s a flip in outcome of a pending case, of course, it raises short term eyebrows, but in most cases, the embarrassment of those those eyebrows is preferable to upholding an erroneous decision when there’s an opportunity to correct it. The new panel here obviously thought that short term eyebrows from people like Dennis were insignificant compared to the long term consequences of the original panel precedential decision, on future cases.

    1. 11.1

      LR, you err — explicitly — with an attempted comparison to a single judge. Also, this is not merely a change of “judge” during a case prior to a decision, but quite clearly the losing judge involved in manipulating the system to gain the opposite results — on all of the SAME facts already adjudicated.

  2. 10

    This is why for years I asked Dennis to list the members of each panel in each decision he summarizes (a request to which he appears to have recently acceded). Anyone who follows the CAFC knows it’s not a monolith and that outcomes are often panel-dependent.

  3. 9

    Would not en banc reconsideration rather than a panel rehearing moot the whole argument here?

    1. 9.1

      The “whole argument” here is both an Ends and a Means thing Paul.

      An en banc hearing would resolve the Ends – but do you think such would resolve the Means? (as in, remove the appearance of impropriety?)

      1. 9.1.1

        An en banc decision would resolve the strange title question Dennis raised here as to: “Decisions by the Court as an Institution; or by the Judge as a Human?” and render moot any effect of changing one panel judge and the decision in a rehearing.


          Only if they address the procedural aspects (and not JUST the merits of this case).

          As I stated, there is an Ends AND a Means at issue here.

  4. 7

    An appropriately programmed and configured AI system, has potential to possibly eliminate Judges, as the AI has been used to replace ppl who operate deep-fryers to make french fries, progress in tech development can possibly eliminate the need for much of the Judiciary, it rests on the Inventors to show, and claim the method for. In the 100% efficient system, an inventor files electronically and the AI delivers a response within five minutes, of all possible litigation, opposition, etc. scenarios and lets you know whether you’ve won or lost your patent case, before a human examiner ever lays eyes on it !!!!! Provided, IF, the AI is “good enough”. So convenient that would be, to have the results from 10T simulations of scanrios, kicked right back at ya five minutes after you “file” the spec. Trust in AI. Sci-fi is fun !

  5. 6

    “Rather, it centers on the individual humans charged with making the decisions.”

    Lelz, if you think that individual examiners (even with sig authority) are just let loose to make their own decisions you’re obviously wrong.

  6. 5

    “Rather, it centers on the individual humans charged with making the decisions”

    ….which in turn depends on the conditioning the minds have been subjected to (sometimes maybe money too, lol). The concept of “operation of law” is a misnomer since laws can never operate by themselves and it ties in to the thesis that all battles are battles for the mind.

      1. 5.1.1

        Yeah, make a list of all the “minds” involved, in a typical case, the target audience to satisfy includes:

        your boss if you have one
        patent examiner(s)
        Board of Appeals possibly
        typical opposing Counsel of ordinary skill, who’s gonna come after you
        one or more Judges possibly
        possibly Jurists selected from the general public

        Many different minds involved, each having their own biases. Most folk recognize this, maybe some don’t, I just think its “cool” to think about the many diff. minds involved, and their biases regardless of instructions, they’re still present. Cool stuff !

        If you had a choice to write but for a single one of the audience in the list above towards “satisfying” them, which in the list would you focus on ? Does the nature of the tech involved influence your decision ? I don’t know, my typing fingers were bored is all perhaps. haha, hava good weekend.


          Chrissy / this is actually a very intelligent post.

          Nicely done (maybe this type of boredom more often…)


            floof, the wind is all yours, junya, it is nice to know basis your past replies, that I have a personal +roll !! I’m loved !! Awww.. Not sure whether it was Zimmerman or Cohen who wrote it, but “ya gots to serve somebody” the song went. So who do you write for ? Certainly not the Inventor(s), and certainly not the Examiner(s)… yet, they’re part of the subset. haha, I’m not a teacher but it seems a good question – Who are you really writing for when drafting a patent ? I’d reckon that some in the past at least have run into problems, merely because they wrote for the “wrong” audience.

            That’s a good lead for maybe the good writer “Ellen MacDermott” to pick up on…. “ya gots to serve somebody” when writing a patent application, as (I think it was she) who found interest re my comment on Steely Dan and Scotch Whiskey.

            Tell me, Anon, guru, who do you write for ? 🙂 btzch


              All of the above, taking care to both weave a story (for any variety of jury-folk) with enough technical grit for any opposing team (along with projected back-up embodiments).

              Of course, there are baseline versions (and Cadillac versions) depending on the needs and desires of the client.

              I would also point out that even for the same client, I will see a spectrum of such needs and desires, so the drafting is highly specialized per individual application.


                Yeah, write for everybody, but if you try to please everybody, it is possible to please nobody ! Or so I heard….

                One more to add to the list might be:

                “A CAFC, or other like-tribunal in the future, 5 – 15 years down the road”

                One has to write for a hypothetical court that won’t exist until 10-15 yrs or so after the patent is issued !! And one must project what they think about how the law will be applied 15 yrs down the road or whatever. Tough task ! Maybe…

                One more for the list too, might be:

                Persons of ordinary skill in the art.

                That’s since the spec must enable such POS in the art
                to make, use, etc. So, one writes for them also !!

  7. 2

    My sense is that the rehearing opp. has to be right.

    It’s not obvious what should happen when a panel issues a 2-1 decision and then one of the 2 fully leaves the bench (or dies or recuses or whatever) before a rehearing petition comes in. One rule favors the status quo: the judge who left is treated the same as if s/he had stayed on the bench but voted to reject rehearing. That’s what Novartis wants and is what some courts do. The other rule favors the petitioner a bit: the judge who left is replaced with a new judge. That’s what Accord wants and what some other courts do. I think courts could rationally do it either way. I somewhat favor the way the Federal Circuit’s done it here (Accord’s way), and don’t buy the argument that the Federal Circuit’s somehow committed to do it Novartis’s way.

    1. 2.1

      At the CFAC, the ‘tradition’ is that someone from the Majority must vote to allow the case to be reheard.

      That simply did not happen here.

      Like it or not, the Means matter.

      PS: the easy way to see this is to see what Malcolm wants and generally go to the opposite.

      1. 2.1.1

        I’m not aware of this “tradition” and am skeptical it exists. If there are no retirements/recusals/deaths, then that’s just how the math works for a 3-judge panel. If you have a 2-1 decision and no one from the majority votes for rehearing, then there can be no rehearing.

        idk who malcolm is


          Come now dcl, you want to be clue1ess about both the immediate case AND Patently-O history?

          How new are you? Still wet from the boat splash?

          As to:
          If you have a 2-1 decision and no one from the majority votes for rehearing, then there can be no rehearing.

          Unless you have the facts of this case. I won’t spoil your discovery when you read into the facts of what happened here.

          idk who malcolm is

          S I G H
          “The Prophet” is merely the latest incarnation of Malcolm. (Started out as Malcolm Mooney, was “mm” for quite a while at the turn of his fifteen year anniversary here, before he disappeared for about 14 months or so.

          His moniker may be “relatively” new, but there is nothing new about his script.


            I may have misunderstood what you meant by tradition. *If* there’s no change in panel composition (e.g., death, retirement, recusal), then someone from the majority has to vote for rehearing…

            I’m not aware of a “tradition” where the court issues a 2-1 decision, one of the 2 leaves, and no rehearing can happen without the remaining member of the majority … if that’s what you meant. Has this ever happened?

            I remember MM; wasn’t aware of the reincarnation.


                Yes, has this ever happened at the Federal Circuit before this case?

                If not, what’s this “tradition” you’re referring to?

                1. One last try because I can’t tell if you’re being deliberately obtuse or I’m just not explaining myself well.

                  You say this: the ‘tradition’ is that someone from the Majority must vote to allow the case to be reheard.

                  1. If you’re referring to ordinary appeals where there are no deaths, retirements, or recusals between the opinion and considering the rehearing decision–that’s not a “tradition.” It’s just math. If there’s a 2-1 decision, there can be no rehearing unless one of the 2 votes for it. Not because of any tradition, but because 2 votes out of 3 against rehearing means no rehearing.

                  2. You seemingly don’t mean that there’s a “tradition” at the Federal Circuit that applies where one judge drops off of a three-judge panel, because you have no examples of that ever happening.

                  3. In each of my previous comments, when I make either of points 1 and 2, you point to the other one.

                  4. This seems to be a weird, one-off case that seems never to have come up at the Federal Circuit before. There’s no relevant “tradition.” Unless a rule, a statute, or binding precedent answers this question, then the court probably has discretion to do whatever it thinks makes sense. Both parties suggest rational answers. I don’t think it was irrational for the court to do what it did. Rule 47.11 refers to substituting a judge in analogous circumstances. Novartis says that Rule 47.11 doesn’t apply here, but doesn’t (in my view) show that some other law does apply that dictates the result it wants.

                2. You are the one being deliberately obtuse.

                  You just don’t get the fact here that not a single one of a majority moved to allow this.

                  Either and both paths to get to that tradition ARE the tradition.

                  The fact of the matter is that this [the facts of this case, and — importantly — no facts of any other case] has brought about the instant result.

                  At a case with this posture.
                  At the CAFC.

    2. 2.2

      The “ethical” solution is a tradition that Judges shouldn’t retire in the middle of cases. The fact that the Fed Cir allowed this fiasco to occur is sad.

      Mechanically, the solution is simple; Judges just stop sitting on panels one ‘cycle’ before they actually turn in their keys/badge. This just leaves the *very* rare situation where a sitting judge unexpectedly dies, and even then, wrt the few cases in flight. Given this is a one-a-century event, handle that small handful of cases en banc.

      1. 2.2.1

        I don’t think your proposal would have worked here, given the timeline of this case:

        July 8, 2021 – oral argument

        July 27, 2021 – Judge O’Malley announces forthcoming retirement in March 2022 – eight months later – i.e., plenty of time for this case to have been finalized

        Jan. 3, 2022 – opinion issues (note this is 6 months after oral argument, which is on the long side; this length of time for the opinion to issue can’t be blamed on Judge O’Malley alone — Judge Moore could have delayed writing the dissent). The petition for rehearing is due in 30 days, which could still have allowed this case to be finalized before Judge O’Malley’s retirement, except:

        Jan 21, 2022 (a Friday) – motion to extend time to file petition for rehearing until 2/23/2022 (appellants had contacted Novartis’s counsel re consent but had not heard back before filing the motion); extension granted on Monday, Jan. 24.

        Feb. 23, 2022 – petition for rehearing filed

        Feb. 28, 2022 – court invited response to petition, due March 14 (any of the judges on the panel could have requested this; my money is on Judge Moore)

        March 4, 2022 – Novartis response filed (10 days early)

        March 11, 2022 – Judge O’Malley retires.


          Would be interesting with a FOIA request to see if it was indeed Moore that pushed delays.

          Which of course, only makes that tradition — of only proceeding if one of the judges in the majority permit it — even more significant.


            Yes. It’s too bad FOIA only applies to the Executive branch.

            Regarding the point that “tradition” would require one of the majority judges to support rehearing, Federal Circuit Rule 47.11 “Quorum” is interesting — unlike with ordinary case panels, where the panels are supposed to be randomly composed, and cases randomly assigned (see IOP#3), when a judge can no longer participate because of retirement, the Chief Judge gets to select the replacement judge:

            “If a judge of a panel that has heard oral argument or
            taken under submission any appeal, petition, or motion is unable to continue with consideration of the matter because of death, illness, resignation, incapacity, or recusal, the remaining judges will determine the matter if they are in agreement and no remaining judge requests the designation of another judge. If the remaining judges are not in agreement or if any remaining judge requests the designation of another judge, the remaining judges will promptly advise the chief judge who will secure another judge to sit with the panel.”

            I couldn’t find anything in the Rules or IOPs creating an exception to the Chief Judge’s authority to appoint the replacement judge when the Chief Judge herself is on the panel where the judges are not in agreement.

            Sadly, and regardless of the merits, in my humble opinion, this whole situation is not a good look for the court.


              “tradition”. always conjers up the Erie RR case whe(rei)n it was writ maybe, that stare decices is merely an option, at discretion. Since then, perception of new possibilities might make even a non-rabid animal, to froth !!


          This seems right. Some cases just drag on for years for reasons beyond the control of individual judges. Also, the proposed solution doesn’t account for all the different ways a judge might leave a panel. Justice KBJ probably left some pending cases behind at the DC Circuit. There are also deaths, recusals, and other circumstances judges can’t fully plan for.

      2. 2.2.2

        > The “ethical” solution is a tradition that Judges shouldn’t retire in
        > the middle of cases. The fact that the Fed Cir allowed this
        > fiasco to occur is sad

        I don’t think you can begrudge Judge O’Malley for deciding this case, as it was properly assigned to her and she remained an Article III judge until the day she was officially retired. You can argue that, given that there was already a split among the other two judges, she could have simply stepped back and not weighed in on the decision, to let it get decided by a panel that includes the replacement judge. But there are no real “ethical” issues here in Judge O’Malley deciding to instead stay on and author the majority opinion. There are plenty of examples of judges going out of their way to issue “parting shot” decisions on very important issues of great concern, on their way out the door to retirement.

    3. 2.3

      > It’s not obvious what should happen when a panel issues a 2-1
      > decision and then one of the 2 fully leaves the bench (or dies or
      > recuses or whatever) before a rehearing petition comes in.

      This pretty much nails the problem, and the problem here is amplified by the fact that one of the two judges in the original majority — in fact the lead author of the opinion — has retired. Most of the time when you have a judicial reassignment between the original panel decision and rehearing, it doesn’t flip the outcome because the initial decision was either 3-0 or 2-1 with the dissenting judge being the one leaving the bench.

      The particular fact pattern here really tests the wisdom of the so-called “tradition” that Novartis argues about, because if you accept their view, no rehearing should be granted even though only one of the judges currently assigned to the case actually believes it was correctly decided.

  8. 1

    “ Now, Novartis is seeking a rehearing — arguing (1) that the original decision was correct”

    It wasn’t correct. It was a travesty.

    1. 1.1

      As I wrote in another thread, you have an obligation to let us know your real name and the law firm that you are with, as well as telling each of your IP clients of your disdain for patents, since otherwise you are committing fraud. This information should be made known to all of your firm’s clients to warn them that they are not getting the service they expect to receive due to your hatred of the patent system

Comments are closed.