“Power tends to corrupt and absolute power corrupts absolutely.” (Lord Acton in 1887)
[Analysis by Jim Dickinson] Suppose you had a way of speeding your product through FDA’s approval process while slowing down your competitor’s similar product. That’s what some medical device companies – like some drug companies in FDA’s transformative Generic Drug Scandal 35 years ago – are allegedly doing at CDRH with the active help of 13-year director Jeffrey E. Shuren. Their trick? They hire an unscrupulous outside intermediary who promotes a cozy relationship with the affable and gregarious Shuren as well as with his subordinates who do the real work of reviewing devices.
That’s the way it worked in the Generic Drug Scandal too, the difference being that the drug reviewers did it for money and illegal perks directly trafficked by principals of the generic drug manufacturer without going through an intermediary. Inappropriate coziness between regulator and regulatee, however fueled, is a driving feature in both examples. Corruption isn’t only bribery -- it can be all kinds of complicit shenanigans in high places, such as dishonesty, lucrative employment prospects after leaving FDA, breach of trust by someone in public authority for the concealed benefit of a private third party, and language alterations to achieve scientifically unsupportable ends, all of which apply to CDRH’s handling of 510(k) review documents.
The 510(k) “generic device” parallel is just as much as, if not more in need of statutory reform than the generic drug review process was when the Generic Drug Scandal brought statutory reforms after Congress got involved in 1989. In 2011, the Institute of Medicine recommended total replacement of the 510(k) program. Then 35 years old, the program was showing its age and needed modernization or replacement, neither of which happened. Shuren promptly opposed the idea. As in the case of abbreviated marketing applications for generic drugs (copies of innovator drugs), Section 510(k) of the Food, Drug and Cosmetic Act had been enacted to provide a simpler and easier regulatory route to market for second and subsequent copies of innovator devices -- but now it was being abused.
Power of personal relations
The key ingredient in the corruption of the generic drug office back then and CDRH now is how “personal” a person-to-person relationship is allowed to get between regulator and regulatee. As a client’s testimonial on one FDA-focused attorney’s Web site puts it: “We will not make a submission or have any type of communication with the FDA without first getting [the attorney’s] insights and recommendations. His understanding of FDA policy and all its nuances, and even more importantly, his personal relationships with the senior folks at the FDA make him an indispensable member of our extended regulatory team.”
It was his personal relationships with outside industry that got FDA chief of generic drugs Marvin Seife, MD, into federal prison in the wake of the Generic Drug Scandal; he received no bribes, just an occasional meal with industry “friends.” And then he lied about it. Seife, like Shuren, had an infectious camaraderie about him – probably not a good thing in a leader whose job has responsibility for enforcing discipline in those below. It was impossible to dislike Seife. He loved everybody and everybody loved him. He had a distinctive gait that was caused by a foot condition that required him to wear special boots. When he went to prison those boots were taken away and he was forced to wear prison footwear which ultimately caused gangrene requiring amputation of his left lower leg and right big toe.
While FDA employees have more restrictive rules governing their outside relationships than do attorneys and consultants working for regulated industry, the government-wide sanctions against “even the appearance of a conflict of interest” require respect on both sides of the regulatory fence. Even commissioners can get too cozy with regulated companies – Lester M. Crawford was compelled to resign after only two months in office in 2005 because he had undisclosed conflicts of interest.
But that was then, this is now. Today’s FDA – at least, Shuren’s corner of it – welcomes coziness with industry and user fees unintentionally encourage that. Shuren has even been publicly challenged about alleged conflict of interest involving his lawyer-wife’s device industry clients. A current case study involves – ironically – the original designer of the modern spring-powered mechanical infusion pump, Andy Sealfon, of Chester, NY, who first got into Shuren’s gunsights in 2015 with an earlier version of his 510(k)-cleared device, the Freedom60 pump. Back then, as we exclusively reported, Minneapolis attorney Mark DuVal and Shuren pulled levers inside FDA to inject a false trade complaint into a routine agency inspection then in process at Sealfon’s RMS Medical Products in Chester. The complaint had been generated by a DuVal client, EMED Technologies, that was then and still is a competitor of Sealfon.
Relations with CDRH ruined
DuVal’s high-speed intrusion warped the inspection, ruined Sealfon’s relationship with CDRH and had “devastating” effects for his company in the infusion pump marketplace, eventually leading an activist investor in RMS to force him out of the company he founded. The RMS name was later changed to Koru Medical Systems and moved to Mahwah, NJ. Koru continued his product line through the Freedom60. Sealfon, an engineer who earned 55 patents in fluidics and infusion products dating back to 1968, later formed a new partnership with Innovative Health Sciences (IHS). There, he revamped his pump design in a new 510(k) citing the Freedom60 as predicate, thus projecting three companies into imminent competition, two of them (Koru and IHS) with different versions of the same original Sealfon design.
That’s when the present war between them broke out. Already on the market, Koru’s and EMED’s competing pumps slid through Shuren’s 510(k) review machine with their latest upgrades and label changes in 261 and 90 days respectively, while Sealfon’s new Insignis version was kept out the market for 570-plus days and counting under the obstructions of four CDRH reviewers; it had won Europe’s counterpart CE Mark (license to market) within 60 days. According to a 2021 EpsteinBecker Green analysis of FDA statistics, the average review time for all 510(k)s submitted over a 12-year period was 176.5 days, ranging from 50 to over 500 days. Sealfon says a market source told IHS that “Koru is telling customers they are working on building up something similar to us and that we are never going to get ours out of FDA, so just wait and we’ll bring it to market. Koru’s plan was to delay us until we ran out of money and create the time for them to copy us. So far they have succeeded on both counts.”
Beginning 11/16/22 the Insignis reviewing team in CDRH mysteriously ”went dark” for the next 37 days, compounded by Thanksgiving and suspending all communications with IHS which had been encouraged to think clearance was imminent. On 12/5/22, Sealfon contacted CDRH’s Office of the Ombudsman to ask for help in reaching out to the unresponsive Insignis reviewers. Weeks later, deputy ombudsman Ken Skodacek reported back that the matter had been “kicked upstairs” and now was out of the reviewers’ hands. “They are working on a response shortly,” Sealfon remembers being told of Skodacek saying, “which will not be good news, but they will offer you a way out to finally obtain a clearance.” Who “they” were was not explained, nor was the reason for the reversed prospects of imminent clearance.
False basis for clearance denial
The reason came two weeks later. On 12/21/22, IHS was blown out of the water by the arrival of a three-page Not Substantially Equivalent (NSE) letter from Office of GastroRenal, ObGyn, General Hospital and Urology Devices director Courtney H. Lias in which she alleged data validity and integrity issues in the 510(k)’s submitted performance data from two independent tests. These included “several instances of identical results” that Lias wrote were considered “improbable,” but she did not elaborate. Sealfon says this was a disagreement that flowed from the use of his previously published data that he believed was adequately explained and corrected. However, since FDA has never been known to admit error, the NSE letter itself stood despite its misunderstood basis as CDRH deployed alternative reasons for rejecting Insignis. What its language meant without saying so, as later events would demonstrate, was that IHS was suspected of committing data fraud in its 510(k) documentation.
Under the guise of an honest appeals process, CDRH’s entrapment of IHS now began. This took the form of two doomed appeals and two fruitless meetings. During these exchanges, IHS says CDRH personnel changed their justifications for the NSE decision numerous times, waxing and waning between different performance tests, environments of use, technical characteristics and comparisons with other pumps including the predicate. Outside the clinical trial setting, FDA has never defined data integrity and validity for medical devices, as it has for drugs and biologics.
IHS pointed out that CDRH’s review team on the Insignis had until 11/16/22 virtually declared the Insignis system Substantially Equivalent to its predicates when it wrote in a decision document: “The technology comparison resulted in the conclusion that the subject device technology differs from the predicate device. Specifically, the identified differences between the subject device and the predicate devices included: environmental of use, type of tubing (variable and fixed rate tubing vs fixed rate only), needle set configurations, and residual volume left in the infusion sets. The review team determined that the differences in technologies as compared to predicate devices did not raise different questions of safety and effectiveness.”
All to no avail. On 5/23 this year, Shuren delivered in 15 pages just what IHS feared: a convoluted endorsement of the 12/21/22 NSE letter imbedded in a morass of what had gone before, even the same inconsistencies and contradictions. Significantly, it did not materially advance any data integrity issues, resting instead on performance test data. Sealfon saw in the letter that CDRH basically did not understand his device. He counted 29 falsehoods the company had already pointed out in previous answers to reviewers’ questions. In short, Shuren’s denial recycled 23 contentious issues that Sealfon says attempted to apply performance standards that don’t fit the Insignis device and six differences of opinion, all adding up to 29 corrupt reasons to block Insignis’s market entry.
They all knew DuVal
Twice in his letter, Shuren explicitly asserted that each of FDA’s decisions on the Insignis had been “made independently and without outside influence, undue interference, or delay,” as if tacitly recognizing IHS’s frequent allegations about what Sealfon saw as DuVal’s continuing role against him, resumed from the 2015 interference in the RMS inspection. Sealfon remembers Skodacek, a recent recruit from industry, cautioning IHS not to mention DuVal in an upcoming appeal of the NSE decision “and then we found out that he (Skodacek) was working for DuVal presenting at Duval’s meetings on video and in person.” Skodacek’s boss, ombudsman Abiy Desta, “volunteered to us at the second appeal that he approved Skodacek to join DuVal’s presentations to industry to drum up new business.” Sealfon said. “They all knew DuVal. Shuren in the past also attended these meetings with DuVal.”
Skodacek’s Linked In Web page says: “As a medical device professional, I am an enthusiastic results-driven team leader with ~30 years of experience working with implantable medical devices including regulating devices at FDA and numerous positions in the medical device industry including senior management, clinical studies, field-based operations, project management, marketing, and engineering.
“As a professional coach, I partner with people in a thought-provoking and creative process that inspires others to maximize their personal and professional potential. I create awareness, develop structures, and encourage actions in alignment with their goals. I tell them what they don't want to hear and have them see what they don't want to see in order to allow them to achieve goals, which they previously thought were impossible.”
IHS president and CEO Phil Daetwyler remembers that Skodacek “deliberately did not inform us that he had a pre-existing relationship with DuVal and had been at DuVal-sponsored marketing events put on by DuVal in Europe and in the USA.” Daetwyler says he made a second complaint about DuVal’s conflict of interest as an attorney representing an IHS competitor (EMED) while simultaneously working with Desta who was counseling IHS in its post-NSE problems at CDRH. “This conversation is on the record,” Daetwyler told us. “Mr. Desta stated that he had run the marketing events ‘up the flagpole’ and ‘no one at the FDA has a problem with them’.”
Perhaps because of all this personal chumminess between DuVal and the CDRH hierarchy IHS could find no sign that the competitive EMED and Koru devices had been subjected to the same tests and standards as were inflicted on the fundamentally equivalent but more advanced Insignis, even though Koru’s pump and IHS’s had essentially the same predicate in common, each invented by Sealfon.
Dirty tricks OK at CDRH
“They know there is no one we can turn to for a scientific review and the CDRH appeals are all kangaroo courts,” Sealfon told us. “There is no oversight and they can do as they please.” Doing as they please apparently includes entrapment by dirty tricks, something CDRH attempted on the same day (12/21/22) that Lias’s NSE letter to IHS went out. At 9:22 that evening, one of her subordinates, biomedical engineer Farzaneh Akhavannik, using a different name (Farzaneh Mariani) and a private email address, wrote IHS’s info@ address: “I have learned about you [sic] Insignis Infusion Pump from your website through a google search. I was wondering if you currently sell the product within the US? If so, please indicate how I may put in an order for the device.” It was an obvious if clumsy attempt to entrap IHS into committing a felony by accepting an order for an illegal purchase of an uncleared device. Outraged, IHS president Daetwyler emailed CDRH chief medical officer William Maisel asking if this was acceptable FDA behavior. Maisel answered: “Yes … In some cases, those investigations include inquiries from non-FDA email addresses.”
3 worst falsehoods
Three of the Shuren appeal denial letter’s most egregious falsehoods, Sealfon told us, were:
Shuren: “You state that your device has a feature that, ‘To ensure safety, the system will automatically respond to increasing resistance at the infusion site, immediately decreasing the flow rate for patients experiencing tissue saturation. For intravenous applications, the system’s ability to respond to site pressure minimizes infiltration and can therefore prevent overflow or overdose.’ (Draft Labeling, page 1). However, your submission does not describe device sensors or another way the device could detect or manage these issues.
“I have decided to uphold both the March Appeal decision, and FDA’s NSE decision, because the Insignis has not been demonstrated to be substantially equivalent to the predicate device identified in your 510(k) submission.”
Sealfon: “FDA requested testing our system with TIR101 [a privately developed AAMI standard for infusion pump performance] which tests how accurately an electronic pump using constant flow rate can maintain flow with increasing back pressure. This always results in higher pressures being created which will harm patients. My designs do the exact opposite, they reduce flow rate and pressure immediately and in proportion to any back pressure developed in the delivery catheters or inside the patient.
“Thus this was proving a negative—that this test does not apply to us at all, and CDRH is left defending this wrong request and saying they need it as part of their substantially equivalent findings. While TIR101 is not appropriate for constant pressure infusion devices, it does totally demonstrate the theoretical mechanism of how the system slows down with increasing back pressure. They also strongly advised we do this test at least three times, but again, this data was provided for theoretical discussion, it was for educational use to understand why TIR101 is the wrong test for this device and not intended for a non-substantially equivalence (NSE) finding. The NSE confirmation is especially egregious: I am the inventor of both the FREEDOM60 and Insignis (as well as their predicates), TIR101 confirms they both have the same mechanism of operation, the FREEDOM60 predicate contained the same language as was used for Insignis, and the entire discussion of this mechanism was included in the predicate submission—which was the very data that was used by Shuren to claim ‘Data Integrity’ violation. (They read the predicate only to find the reference for Data Integrity but not to understand substantial equivalence?)
However, after the fact and after we repeated this test on the Insignis, it confirms that the Insignis and Freedom60 have exactly the same performance with respect to back pressure. Identical in fact, which demonstrates the same claim for Insignis as the Freedom60 had, which was that the flow rate decreases when back pressure increases.”
Shuren: “The Monte Carlo evaluation [used by IHS] utilizes a modeling methodology which is not sufficient to demonstrate performance of the proposed infusion system. Although HIS provided information that purports to validate free flow, air emboli, and fluid ingress, the methodology that IHS’s submission described was not conducted on the complete Insignis infusion system.”
CDRH incompetence
Sealfon: “This is just CDRH incompetence in not understanding how manufacturing uses statistical sampling to maintain accuracy. I don’t know any way to quote accuracy for any system that relies on a tolerance of multiple components to achieve a certain accuracy. The choice is to perform millions of tests on combinations of parts, use statistics, or fake it by assembling known performing parts. So if the pump pressure has a tolerance of 10%, and uses two tubings of tolerance of 10%, plus the fluid viscosity has a tolerance of 10% and the friction of the syringe has a tolerance of 15%, then the total tolerance if you add it all up is 55%. But the probability would be extremely rare that the worst-case scenarios could all happen at one time. They never asked about this.”
Shuren: “In its presentation, IHS noted that FDA should have addressed any data integrity issues with IHS while evaluating its 510(k) submission. IHS claimed that had FDA done so, IHS would have had an opportunity to address any data integrity issues during the review process.
“FDA followed its internal processes to evaluate IHS’s submission. For example, FDA utilized the communication processes described in its publicly available communication guidance document to interact with IHS. FDA informed IHS of the 510(k) submission deficiencies on August 2, 2022. In that letter, FDA also clearly articulated IHS’s response timeframe, so that FDA could make a timely decision on IHS’s submission. Therefore, I find that FDA utilized its internal processes to conduct its 510(k) review, and appropriately gave IHS the opportunity to address the deficiencies set forth in FDA’s August 2, 2022 letter.”
Sealfon: “There are many false statements in these paragraphs. FDA did not follow its internal processes on ‘Data Integrity’. FDA did not inform us of ‘Data Integrity’ on Aug. 2, 2022 but in the termination of our 510(k) submission on December 21, 2021. They did not follow ANY part of their published procedure. Everything in these paragraphs is false.”
A rogue process ‘harms patients’
These and myriad lesser examples display a hidden rogue medical device review process at FDA that has not only been corrupted by unwholesome outside influences that continue unchecked but has avoidably deprived patients of at least one superior device and probably others. In their place are new EMED and Koru infusion pumps that Sealfon says can hurt patients by being less patient-friendly, less adjustable than and lack the drug flow accuracy of IHS’s Insignis.
Those clearances, he told us, “are very likely to harm patients. The actual patient risks are infusions that are too fast or too slow. The bigger risk is infusing too fast as that leads to painful site reactions, and can lead to large inflamed bumps, oozing, and infection. While going too slow seems harmless enough, it becomes a quality of life issue, and patients are loath to remain hooked up for hours to get this medication. So if these infusions hurt too much or take too long, patient compliance goes down—they don’t do the infusions—and that result is serious systemic infection, which can require hospitalization and can sometimes be fatal.
“So if these competitors provide systems that have widely different flow rates, and both EMED and Koru products have been measured at +/- 40% of stated flow rate, then there is no way for the patient to find and use the flow rate that produces the best balance between site reactions and length of infusion. So if a patient is complaining about a flow rate prescribed that is creating these painful site reactions, the clinician prescribes a slower flow rate. But the new supposedly slower flow rate turns out to be faster, because the tolerance is so broad. So the pain gets worse. On the next go-around they have no idea what flow rate to pick. This is a patient’s and clinician’s nightmare. Nothing seems to work.”
Nothing, that is, except the corruption in Shuren’s 510(k) review machine.
Sent advance copies of this report with an invitation to comment, all FDA named persons in it, as well as commissioner Robert Califf and assistant commissioner for media affairs Michael Felberbaum twice declined to respond.
