OxyContin's FDA approval was granted with zero addiction or abuse liability testing — Purdue's only evidence was an untested hypothesis that its controlled-release formulation would prevent euphoric peaks.
2
Curtis Wright, the FDA medical examiner who approved OxyContin, left the agency one year later to work at Purdue Pharma for three times his government salary — the federal prosecutor who built the biggest case against Purdue said on the record: 'I think a deal was made.'
3
Purdue Pharma teams physically traveled to Maryland to spend three days helping Curtis Wright write his own FDA reviews of their drug — emails proving this were later reported 'lost or destroyed' by the FDA.
4
The 'revolving door' between FDA reviewers and pharma industry creates what Attia's guest calls 'soft corruption' — not an explicit quid pro quo but a structural incentive that biases approval decisions without ever requiring a spoken deal.
Protocols
Concrete recipes — what, when, how much, and why
5 items
Read the package insert as a marketing document, not a neutral safety summary
WhatWhen evaluating a drug's safety profile, read the package insert with the assumption that every claim — especially hedged language like 'is believed to' or 'may reduce' — was negotiated between the drug company and the FDA, not independently verified. Pay particular attention to abuse liability claims for controlled substances.
WhenWhenever reviewing a new prescription for a controlled substance, evaluating a drug's addiction potential, or advising patients on opioid or high-abuse-potential medications.
For whomClinicians prescribing controlled substances, patients evaluating opioid prescriptions, policy researchers assessing drug approval quality.
WhyRichard Sackler explicitly called the package insert 'our most potent selling instrument.' Marketing language can be embedded in what looks like regulatory fine print. The OxyContin package insert contained an unsubstantiated abuse liability claim that drove a decade of prescribing behavior.
The practical heuristic: when a package insert contains a comparative safety claim — especially one framed with epistemic hedges like 'is believed to,' 'may reduce,' or 'has not been shown to cause' — ask: what study is cited in support of that claim? If there is no citation to a specific trial, the claim likely reflects a negotiated hypothesis rather than demonstrated evidence. The OxyContin package insert's claim that the controlled-release 'seal is believed to reduce the abuse liability of the drug' had no supporting study. It was Purdue's unverified hypothesis, drafted with FDA collaboration, and it became the cornerstone of a decade of aggressive marketing to physicians.
there's a line in the original package insert for oxycontin that says the content seal is believed to reduce the abuse liability of the drug and that's a weird thing to have in a package insert
Apply heightened scrutiny to marketing claims that lack cited studies
WhatFor any drug claim — especially comparative safety claims about addiction or abuse potential — require a specific cited study before accepting the claim as evidence-based. Treat unsupported comparative claims ('safer than,' 'less prone to abuse than,' 'lower addiction potential than') as marketing hypotheses until proven.
WhenAt the point of prescription, when assessing a new drug being introduced into clinical practice, when a sales representative cites the package insert as evidence.
DoseThis is a permanent epistemic standard, not a one-time check.
For whomClinicians, pharmacists, policy analysts, journalists, and patients evaluating drug safety claims.
WhyOxyContin's entire abuse liability narrative was built on an untested conjecture. Zero addiction or abuse liability studies were conducted before approval. The claim made it into the package insert without a scientific basis, and physicians who trusted the insert as an authoritative document prescribed accordingly.
CaveatsSome legitimate safety claims are extrapolated from mechanism or from related compounds where direct testing would be unethical. The standard is not 'is there a randomized trial' but 'is there a stated evidentiary basis for this specific claim?'
The abuse liability claim in OxyContin's package insert was particularly dangerous because it directly addressed the primary clinical concern physicians would have about prescribing a new opioid — and it resolved that concern with false confidence. Physicians who were appropriately cautious about opioid addiction risk read the package insert and found what appeared to be an FDA-endorsed safety claim. The claim was not only unsubstantiated but ultimately backwards: crushing the controlled-release tablet to defeat the time-release mechanism produced exactly the peak concentration the claim said was impossible.
what was the basis of that belief the scientific basis was there a study that was cited was there an experiment that was done i mean it wouldn't be hard to have studied these things there wasn't one no
Treat the revolving door as a structural risk signal when evaluating regulatory decisions
WhatWhen assessing the reliability of an FDA (or other regulatory agency) approval decision, check whether the primary reviewer moved to the pharmaceutical company being regulated within one to three years of the decision. This is a systemic risk signal for 'soft corruption' — not proof of wrongdoing, but a reason to examine the approval record more carefully.
WhenWhen evaluating the evidentiary basis for a regulatory approval, particularly for high-commercial-value drugs in chronic-use categories (pain management, psychiatry, metabolic disease).
For whomClinicians, policy researchers, journalists, patients evaluating whether to trust regulatory decisions, and policymakers designing regulatory reform.
WhyCurtis Wright approved OxyContin, then joined Purdue Pharma one year later at three times his government salary. The western-district federal prosecutor stated on the record: 'I think a deal was made.' The revolving door does not require an explicit arrangement — the ambient awareness of future employment opportunities shapes behavior even without a spoken agreement.
The 'soft corruption' dynamic is structural rather than individual: FDA reviewers are expert professionals making government salaries while overseeing an industry that actively recruits them at multiples of their agency pay. No explicit deal needs to be made — the mere awareness that favorable decisions can lead to industry careers creates an incentive gradient. This is why Attia's guest argues that Curtis Wright is 'a very extreme example' of a dynamic that 'pervades the whole process' rather than a rogue individual. The policy question — how long should cooling-off periods be, how do you retain talent without accepting structural capture — has no clean answer, but naming the mechanism is the prerequisite for addressing it.
a year after leaving the fda curtis wright is now employed by purdue pharma isn't the only obvious explanation here that there was a quid pro quo a very clear and direct quid pro quo made between purdue and curtis wright
Also said
“i think a deal was made he said i can't prove it just my opinion but i think a deal was made when you look at the totality of the evidence”— The lead federal prosecutor's on-the-record assessment — careful language from someone who reviewed millions of documents.
Use FOIA requests and federal court orders as investigative tools for regulatory transparency
WhatWhen investigating a regulatory decision that may have been compromised, file Freedom of Information Act requests for the reviewing official's correspondence. If the agency delays (common), be prepared to file in federal court to compel document production. Document preservation failures — especially for officials already under DOJ investigation — are themselves evidentiary.
WhenIn investigative journalism, academic research, legal proceedings, or policy review involving FDA, DEA, or other regulatory approval decisions for drugs with serious public health implications.
For whomInvestigative journalists, academic researchers, public health lawyers, congressional oversight staff, patient advocacy organizations.
WhyThe author obtained a New York federal judge's order compelling the FDA to produce thousands of pages of documents, including Curtis Wright's files. The FDA's subsequent claim that his emails were 'lost or destroyed' — for an official already flagged by DOJ — is itself a data point about institutional behavior that would not have been discoverable without the FOIA litigation.
CaveatsFOIA litigation is slow, expensive, and often produces redacted documents. It is most productive when used to establish what the agency is unwilling to disclose — the shape of the absence can be as informative as the documents themselves.
The investigative sequence described: (1) phone call to Curtis Wright — immediate hang-up; (2) FOIA request to FDA — administrative delay; (3) federal court lawsuit — judge ordered production; (4) thousands of pages of documents produced; (5) FDA's claim that Wright's specific files were 'lost or destroyed' — a claim that, given his pre-existing DOJ scrutiny, implies either negligent or deliberate document destruction. The email chain that did emerge — showing a Purdue executive reporting that Wright 'indicated that he would be open to other such informal exchanges in the future' — came not from the FDA but from federal investigations of Purdue itself, illustrating the value of triangulating across multiple FOIA targets.
i filed a freedom of information act request with the fda which they they dragged their feet in the way that the federal bureaucracies always do so i actually sued them in federal court to compel them to turn over documents to me and i got a new york federal judge to order them to do document production to me
Recognize informal 'sidebar' communications as the mechanism of regulatory capture
WhatIn any regulated industry, the formal documented process (submissions, official meetings, written reviews) is where accountability lives. Informal side channels — conference hallway conversations, home-delivered correspondence, private webcam connections, hotel-room work sessions — are where influence happens outside the official record. Treat any informal channel between regulator and regulated party as a regulatory integrity risk.
WhenWhen designing regulatory processes, auditing institutional behavior, investigating approval decisions, or managing compliance programs in regulated industries.
WhyThe OxyContin approval was corrupted through exactly these informal channels: a webcam enabling direct private communication, correspondence routed to Wright's home rather than his FDA office, a sidebar conference conversation that Wright said he was 'open to repeating informally,' and finally the hotel-room co-drafting session. Each step moved the relationship outside the formal record and therefore outside any accountability mechanism.
The pattern has a logic: formalized processes create paper trails, paper trails create accountability, accountability limits what either party can do. The informal channel systematically dismantles each link. The webcam allowed real-time voice communication with no written record. Home-address delivery bypassed official intake logging. The conference sidebar was explicitly described by the Purdue executive as 'not formal.' The hotel stay produced co-authored documents that neither party could be held accountable for individually. By the end, even authorship of a specific sentence in the package insert — 'is believed to reduce the abuse liability' — was disputed between the two parties, with each pointing at the other. The formal accountability structure had been so thoroughly eroded that the specific origin of a key regulatory claim could not be established.
the purdue guy sends back to headquarters where he says you know i got talking with curtis wright was great we talked about the whole oxycontin thing he indicated that he would be open to other such informal exchanges in the future
Also said
“the idea that the fda is working so closely with a drug company that literally you don't know where one ends and the other begins and there's a line that ends up in the package insert claiming that this drug is safer than other alternatives and nobody can tell you who wrote it”— The end-state of the informal channel strategy — complete dissolution of the boundary between regulator and regulated.
What's new
Personal practice updates, fresh positions, predictions
5 items
The package insert is the drug company's most potent selling instrument
~early segment
Richard Sackler, second-generation head of Purdue Pharma, explicitly described the package insert inside a bottle of pills as the company's single most powerful marketing instrument — not advertising, not sales reps, but the supposedly neutral regulatory document.
Why this matters: Most patients and even many physicians treat the package insert as neutral regulatory fine print. The Sacklers understood it as a weaponizable marketing asset, and that framing drove every decision in how they sought FDA approval.
Background
The FDA approval process has two components: safety/efficacy review AND authorization of specific marketing claims. Purdue understood that getting favorable language into the package insert would allow their sales force to quote a regulatory document rather than an advertisement.
The insight is structural: a drug company cannot say whatever it wants in an advertisement, but it can quote its own package insert verbatim to physicians. The package insert therefore becomes the foundation of the entire sales narrative. Getting the FDA to approve language like 'is believed to reduce abuse liability' — even if unsubstantiated — gives salespeople a government-stamped claim they can repeat in every detail call without legal risk. Richard Sackler's description of the package insert as 'our most potent selling instrument' was not a cynical aside — it was a strategic orientation that shaped everything from how they framed the drug's mechanism to how they cultivated Curtis Wright.
richard sackler this second generation sackler was probably the most intimately acquainted with or involved with with oxycontin um there's a point where he talks about how the package insert in in a you know bottle of pills is our most potent selling instrument
Purdue conducted zero abuse liability or addiction tests before OxyContin's approval
~mid segment
Purdue Pharma did not perform any studies on OxyContin's addictiveness or abuse liability before seeking FDA approval. Their entire safety case rested on a theoretical conjecture: that the controlled-release formulation would produce no euphoric peaks, therefore the drug would not be addictive.
Why this matters: An opioid — in the same drug class as heroin and morphine — was approved for widespread clinical use based on a hypothesis rather than data. The absence of abuse liability testing was not a regulatory oversight; it was a deliberate strategy that the FDA allowed.
Background
Opioid drugs were already known to carry addiction potential before OxyContin's mid-1990s approval. The question was whether a controlled-release formulation would change that pharmacological reality.
The conjecture was not inherently unreasonable as a hypothesis: if peak plasma concentrations drive the euphoria that conditions the addiction loop, a formulation that flattened those peaks might genuinely reduce abuse potential. The problem was that Purdue treated this hypothesis as established fact, put it in the package insert as a marketing claim, and never tested it. The controlled-release mechanism also had a critical flaw that turned the hypothesis backwards: abusers quickly discovered they could crush the pill to defeat the time-release coating, releasing the full opioid dose at once — producing precisely the peak concentration Purdue claimed their formulation prevented.
there's they didn't do any tests on the addictiveness of the drug they didn't do any tests on the abuse liability of the drug but there was a kind of hypothesis that they had at purdue which was because of the content system you wouldn't have the kind of peaks and troughs uh of you know a sort of wave of euphoria
Also said
“in this case what it was was it was this conjecture that because of the content system you're not getting the high highs and the low lows and therefore it would it would make it uh you know less prone to abuse than other drugs”— Confirms the abuse liability claim was conjecture, not evidence — the specific word used in the original context.
Purdue employees spent three days at a hotel helping Curtis Wright write his own FDA reviews
~mid segment
A team from Purdue Pharma traveled to Maryland, checked into a hotel, and spent three days with FDA medical examiner Curtis Wright helping him draft the FDA's official review documents for their OxyContin application. Email evidence confirmed this happened.
Why this matters: The FDA reviewer is supposed to be an adversarial check on the drug company — the independent expert who validates or rejects the company's safety claims. Having the company's own team co-author the reviewer's official documents destroys the independence the entire approval process depends on.
Background
Curtis Wright was the FDA's 'grand inquisitor' on the OxyContin application — the person Purdue had to satisfy both for drug approval and for the specific marketing claims they wanted in the package insert.
The collaboration was not limited to the hotel stay. Earlier in the approval process, Purdue sent Curtis Wright a webcam — unusual in 1995 — so he could communicate directly with Purdue officials in Stamford. Wright also began requesting that correspondence be sent to his home address rather than his FDA office, a departure from standard procedure that created less of an official record. These steps, taken together, indicate a systematic erosion of the arm's-length relationship the FDA process requires: first the direct line of communication outside official channels, then the request for private delivery, then the joint document-drafting session. The hotel stay was the culmination, not an isolated incident.
a team of people from purdue travel to maryland and get a hotel room and they spend three days with curtis wright helping him write the his reviews of their studies
Also said
“the company sends curtis wright this is in the 1990s 95. the company sends him a webcam this is like a very early webcam where he can talk directly with folks in stamford curtis wright starts requesting that they send things not to his fda office but to his house”— Documents the escalating pattern of improper contact that preceded the hotel stay — the webcam and home-delivery requests preceded the direct co-drafting.
FDA reported Curtis Wright's emails 'lost or destroyed' after DOJ had already flagged him
~later segment
When the author filed a FOIA request and then successfully sued the FDA in federal court for Curtis Wright's correspondence, the FDA reported that all of Wright's emails and documents had been 'lost or destroyed.' This occurred despite the fact that federal DOJ investigators had already been scrutinizing Wright since 2002-2003.
Why this matters: Document preservation is a legal requirement when a federal employee is under investigation. The FDA's claim that documents were 'lost' for an individual already flagged by multiple federal agencies is, at minimum, an extraordinary institutional failure — and at worst, active obstruction.
Background
The author filed a Freedom of Information Act request, the FDA dragged its feet, and the author ultimately obtained a New York federal judge's order compelling the FDA to produce documents. The email destruction is what was revealed through that process.
The author drew the comparison explicitly: this is not a routine document-retention purge where old records are deleted after a standard retention period. Curtis Wright had been 'red flagged early by other federal agencies' starting around 2002-2003. An employee under active DOJ scrutiny has an affirmative legal obligation to preserve documents, and the agency managing that employee has the same obligation. The FDA's response — essentially 'we happened not to have those' — for the very emails that would be the core evidence in any investigation is, as the author put it, 'the equivalent of saying like there was a very small warehouse fire it's just that box.'
the fda came back and told me that uh all of his correspondents and emails have either been lost or destroyed yeah which again you couldn't make that level of of of malfeasance up
Also said
“this was somebody who was red flagged early by other other federal agencies the notion that um that they would today say uh you know it's the equivalent of saying like there was a very small warehouse fire it's just that box”— The author's framing of why the 'lost' claim is uniquely suspicious given the pre-existing DOJ scrutiny.
The 'revolving door' creates 'soft corruption' that doesn't require an explicit deal
~later segment
The structural incentive problem: FDA reviewers are highly educated experts earning government salaries while overseeing a multi-billion dollar industry that actively recruits former regulators at multiples of their agency pay. Curtis Wright tripled his salary by moving to Purdue one year after approving OxyContin. The author argues this creates a form of corruption that operates without any explicit agreement — just ambient awareness on both sides that favorable regulatory decisions can lead to industry careers.
Why this matters: Explicit quid pro quo corruption is prosecutable and rare. This 'soft corruption' dynamic is structural, legal, extremely common, and far more consequential for public health. The opioid crisis is partly a product of a regulatory incentive structure that was never illegal.
Background
The western-district-of-Virginia federal prosecutor who built the largest case against Purdue reviewed millions of documents and concluded 'I think a deal was made' — but could not prove it, suggesting the evidence supported both the explicit-deal theory and the soft-corruption theory.
The policy dilemma is genuinely hard: you cannot ban former FDA officials from working in the pharmaceutical industry without either accepting that you will lose talented regulators who refuse to be career-limited, or accepting regulatory capture from a workforce that knows its future employers are the firms it regulates. Cooling-off periods (one year, three years) create a delay but not a structural fix — the awareness of potential future employment still shapes decisions during the regulatory tenure. The author does not propose a clean solution but names the problem precisely: it is 'soft corruption' that 'pervades the whole process' and 'is not unique to Curtis Wright — he's a very extreme example.'
i think there's probably a great deal of interest in having the fda officials that you're talking to know that someday there might be a job for them and that's this hard thing because it's not corruption in a in an explicit quid pro quo sense
Also said
“what i would call a kind of soft corruption that pervades this whole process and i don't think it's unique to curtis wright he's a very extreme example”— The author explicitly frames Curtis Wright as the acute extreme of a chronic structural problem — not an anomaly.
Recommendations
Products, supplements, and tools mentioned in the episode
3 items
Empire of Pain by Patrick Radden Keefe
Book
The guest references 'the book' repeatedly throughout the conversation as the evidentiary source for the OxyContin story — the webcam gift, the hotel stay, and other specifics are cited as documented in the book. Keefe's 2021 work is the definitive account of the Sackler family and Purdue Pharma.
The reference appears as 'i don't even think this is in the book' when describing the webcam incident, implying the guest has information that goes beyond Keefe's reporting — obtained through his own FOIA litigation and federal document production. The book is the baseline account; the FOIA documents fill in what the book does not cover.
the company said i don't even think this is in the book but the company sends curtis wright this is in the 1990s 95. the company sends him a webcam
Freedom of Information Act (FOIA) federal court litigation strategy
Tool
When FDA dragged its feet on a FOIA request, the guest sued in federal court and obtained a judge's order requiring document production — resulting in thousands of pages of documents that established the facts of the Curtis Wright case.
The FOIA litigation strategy has a specific sequence: file the initial administrative request, wait for the agency's delay (expected), file in federal court to compel production, obtain a judge's order. The guest notes the FDA's claim that Wright's emails were 'lost or destroyed' is itself a product of this process — without the litigation, that disclosure never would have been made. The shape of what was withheld or destroyed is evidence.
i actually sued them in federal court to compel them to turn over documents to me and i got a new york federal judge to order them to do document production to me thousands and thousands of pages of documents
Structural scrutiny: check reviewer-to-industry employment pipeline for major drug approvals
Practice
The guest identifies the post-approval employment pattern (FDA reviewer joins the company whose drug they approved, at 3x salary, within one year) as the clearest structural signal of potential regulatory compromise — even if no explicit deal can be proven.
This practice is applicable beyond the OxyContin case: any major drug approval where the primary FDA reviewer moves to the pharmaceutical company within a short window warrants additional scrutiny of the approval record. The mechanism does not require corruption in the legal sense — the 'soft corruption' of ambient career incentives shapes decisions without any explicit agreement. Cooling-off periods in the existing regulatory structure are one to two years; the guest implies this may be insufficient.
given that a year after leaving the fda curtis wright is now employed by purdue pharma isn't the only obvious explanation here that there was a quid pro quo a very clear and direct quid pro quo made between purdue and curtis wright
Lines worth pulling out — contrarian, specific, or perfectly phrased
5 items
richard sackler this second generation sackler was probably the most intimately acquainted with or involved with with oxycontin um there's a point where he talks about how the package insert in in a you know bottle of pills is our most potent selling instrument
Reveals the strategic framing inside Purdue — the package insert was never conceived as a neutral safety document but as the company's primary marketing weapon.
a team of people from purdue travel to maryland and get a hotel room and they spend three days with curtis wright helping him write the his reviews of their studies
The single most concrete illustration of how thoroughly the FDA's independence in this approval was compromised — a drug company co-authoring its own regulator's review documents.
i think a deal was made he said i can't prove it just my opinion but i think a deal was made when you look at the totality of the evidence
The lead federal prosecutor who spent years reviewing the case — a man described as 'pretty careful about what he says' — stating on the record that he believes a corrupt arrangement existed.
what i would call a kind of soft corruption that pervades this whole process and i don't think it's unique to curtis wright he's a very extreme example
The conceptual framing that makes this episode relevant beyond its specific case: the structural incentive problem is systemic, not individual — Curtis Wright is the extreme end of a spectrum.
the idea that the fda is working so closely with a drug company that literally you don't know where one ends and the other begins and there's a line that ends up in the package insert claiming that this drug is safer than other alternatives and nobody can tell you who wrote it
The final image that captures what regulatory capture looks like in practice: not bribery, not a smoking gun, but a complete dissolution of the institutional boundary between regulator and regulated.
Sign in to share feedback
Tell us if this brief hit the mark or missed it — feedback feeds back into the next iteration of the prompt.
Reading is free for everyone. A free account adds the personal layer: save protocols, follow experts, and see how the other experts weigh in on this same topic.
Educational summary of the cited expert source — not medical advice. Open the source recording linked above and consult a qualified physician before acting on any protocol.
