The stabilization period is over, and most of the industry is now inside the enhanced DSCSA requirements.

Manufacturers, repackagers, wholesale distributors, and dispensers with 26 or more full-time pharmacy employees have now passed the expiration dates for FDA’s temporary exemptions from certain enhanced tracing requirements. Small dispensers (defined for this exemption based on the company that owned the dispenser having 25 or fewer full-time pharmacists and qualified pharmacy technicians as of November 27, 2024) remain exempt from those specified requirements until November 27, 2026.

(Those exemptions don’t suspend the separate DSCSA requirements for handling suspect and illegitimate product.)

Most of the attention we’ve seen here has gone to serialization and interoperable data exchange. That is the visible, expensive, IT-heavy part, and it deserves the attention it gets. The part that gets less of it is what a trading partner actually has to do when a product looks wrong. That’s an investigation, and it looks a lot more like GMP work than distribution work.

We broke down the verification-systems guidance when it was finalized way back in 2023. Today, we wanted to talk about the operational reality underneath it since it’s come up in multiple conversations we’ve had with experts and industry teams.

What the law asks you to do with a suspect product

Under section 582 of the FD&C Act, manufacturers, repackagers, wholesale distributors, and dispensers must have systems to identify suspect product and handle it. A suspect product is one for which there is reason to believe it may be counterfeit, diverted, stolen, the subject of a fraudulent transaction, intentionally adulterated in a way that would cause serious adverse health consequences or death, or otherwise unfit for distribution with similarly serious consequences.

The FDA spells out how it reads each of those terms in its Definitions of Suspect Product and Illegitimate Product guidance.

When something meets that bar, the steps are specific.

1. Quarantine it from distribution using physical separation, effective electronic controls, or both.

2. Investigate in coordination with your trading partners, which includes validating any applicable transaction history and transaction information in the trading partner’s possession and conducting the required product-identifier verification.

3. Then decide.

The workflow is laid out in FDA’s Verification Systems guidance and summarized cleanly in the Partnership for DSCSA Governance’s report on suspect-product investigations.

Two outcomes branch from there:

Why this is GMP muscle

You might notice that this reads more like a deviation investigation than typical distribution work.

Distribution and dispensing operations are built to move product accurately and fast. When a case comes up short or a scan won’t reconcile, the reflex is to fix the count and keep the line moving. That reflex is efficient, but under DSCSA it’s also the wrong one, because a short case or a mismatched identifier can be the first sign of theft, diversion, or a counterfeit finding its way into the chain.

The discipline the law is asking for here already exists on the manufacturing side. It lives in 21 CFR 211.192. That rule tells a quality unit:

• that any unexplained discrepancy or specification failure has to be thoroughly investigated, whether or not the batch already shipped,

• that the investigation has to extend to other batches or products that might be affected, and

• that a written record has to capture the conclusions and follow-up. In practice, that means establishing the cause, defining the scope, documenting the conclusions, and completing appropriate follow-up.

A GMP investigator does this without thinking about it. That’s the posture DSCSA requires, and that the enhanced system makes increasingly difficult to avoid.

When investigations get treated as paperwork instead of real work, root causes get named and closed without support, and the record falls apart under scrutiny. We wrote about how that plays out under 211.192: teams that don’t take clear notes, discrepancies miscategorized at the start, root causes labeled “to be determined” and never resolved.

Where the gap shows up

Here are a few of the places this tends to break (and therefore where you should devote your attention):

• No written definition of suspect versus illegitimate that a floor supervisor can actually apply, so the judgment call gets made inconsistently or not at all.

• No named owner for the 24-hour notification, so the clock is already running before anyone realizes a 3911 has to go out.

• Investigations that record what happened but not why, with no follow-up. That is the classic 211.192 finding, moved into a new setting.

• Quarantine that isn’t immediate, or isn’t controlled tightly enough to prove the product went nowhere.

• Records that aren’t held for the full six years.

None of these are particularly unusual or new. They’re the predictable result of asking an operation built for throughput to behave, on short notice, like a quality unit.

What to consider putting in place

If you run distribution or dispensing, or you support a client who does, these are a few things to consider (not a full list of course and certainly subject to context):

The firms that come through the next round of inspections cleanly won’t be the ones with the most expensive serialization stack. They’ll be the ones who treated suspect-product handling as an investigation from day one and documented it like one.

If you’re not confident your operation could pass that test, that’s the kind of gap we help close. Talk to us.

Who is The FDA Group?

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on Quality Assurance, Regulatory Affairs, and Clinical Operations.

With thousands of resources worldwide, hundreds of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee. Learn more and schedule a call with us to see if we’re a fit to help you access specialized professionals and execute your projects on time and on budget.

Share The FDA Group's Insider Newsletter

We don’t often get to say this: our guest runs the organization that certifies much of the quality profession. Sid Bhatnagar is the CEO of ASQ, the American Society for Quality, a global body with members in more than 130 countries and 200-plus member-led communities. He sat down with Nick Capman for a conversation about how AI is changing what leadership actually requires, and we’re grateful he made the time.

Sid came to the CEO role by an unusual path for a quality organization. He’s an engineer by training who spent most of his career in technology strategy and implementation, much of it as a CIO in fast-moving financial services. He joined ASQ as Chief Information Officer in 2020, became COO and Chief of Staff, and has served as CEO for about three years. That mix, deep technology background running an organization built on process discipline, made him a sharp guest for this topic.

The conversation covers a lot: where AI pressure is actually coming from, how to balance human judgment against machine speed, what to do with the employee whose job AI will change, and how to think about return on investment when the technology won’t sit still. Here’s what stood out.

Apple Podcasts | Spotify | YouTube | Web + Others

Sid’s key insights and practical takeaways

If you’re short on time, here are the most important lessons from the discussion.

The pressure on executives now comes from the top down, which is new. For most of business history, culture around a tool came up from the technology group or the business unit. AI reversed that. Sid’s point is that the demand is coming straight from boards and investors, who want to see AI drive operational efficiency and keep products and services competitive. The executive’s job expanded as a result. You’re no longer just a decision-maker; you’re the steward of AI adoption and the driver of AI culture, and that’s a role the C-suite mostly hasn’t had to play before.

Insights arrive faster, so timelines compress. The old pattern was a leader posing a problem to their directors and VPs and waiting. Now an executive can get credible information quickly by working through a tool like ChatGPT or Claude over a cup of coffee, then handing it to the team differently: not “find this out for me” but “I’ve done this research, validate it.” Sid’s example is concrete. Work that used to carry a six-month expectation now gets expected in three.

AI may need to separate from IT the way IT once separated from everything else. Nick made the structural argument, and Sid extended it. Decades ago there were no IT departments. Now Nick sees AI and digital titles splitting off from IT, and he thinks executives shouldn’t be surprised to see a dedicated AI function emerge. The FDA Group is in the final steps of bringing on an AI consulting firm, and Nick frames it as an ongoing relationship like accounting or IT, not a one-time project, because annual planning can’t keep pace with the technology.

Who owns AI strategy is a genuinely open question. Sid asked a room of executives who owned their AI strategy and got different answers every time. Right now, ownership tends to land on whoever knows AI best in the organization. Nick ran the same experiment on a panel: he offered six departments and asked who owns it, and one panelist answered “yes.” Another attendee proposed organizing it like the branches of the U.S. government, executive, legislative, judicial, which Nick called the winner. The takeaway both men landed on: there’s no single right answer, because AI touches everything. What matters is what works for your organization, and the willingness to shift as you learn.

Keep humans on both ends of the process, not just in the middle. Most people talk about “human in the middle.” Sid reframes it. Ideation comes from the human side, what problem are we actually trying to solve. Then the tool does its work. Then validation comes back to the human. The tool is bookended by human judgment on both ends. That structure keeps an organization from getting caught in a loop of AI generating work for AI to check, and it keeps people focused on identifying real problems rather than letting the technology run loose.

Lean toward the critical parts of the business, not the tasks. Both converged here, and it’s the most useful career insight in the episode. Nick brought up a book he’s been listening to on the shift from tasks to systems. Task-oriented work will go away. What won’t is critical decision-making. His analogy: an anesthesiologist doesn’t get paid for “here comes the fun juice,” and a pilot doesn’t get paid for easy takeoffs and smooth landings. They get paid for what they know when things go wrong. His advice to professionals is to lean toward the critical decision-making parts of the business rather than sharpening skills AI will absorb. Sid tied it back to ASQ’s core: process thinking, system design, root cause analysis. Things go sideways regardless of which tool you pick, and the question is whether you have the mindset and skill set to design and think at the system level.

The mindset to hire for is innovation, change management, and system thinking. Sid named these three directly. The people worth bringing in aren’t the ones who claim they’ve done something 200 times, because nobody can say that about technology moving this fast. They’re the ones willing to say something new is happening and let’s explore it, not implement it blindly. He drew a line between the leading edge and the bleeding edge. As a nonprofit, ASQ can’t take R&D-style financial risks to sit on the bleeding edge, so the right mindset in the leadership team becomes essential.

Leadership through this is a healthy tension, a pull rather than a push. Sid described the executive’s job as pulling hard enough to bring people along but not so hard that something breaks. Historically leadership leaned more on push. The change management piece matters more now because the moment people fear for their jobs, innovation shuts down and appetite for change disappears. Building the ecosystem where that fear doesn’t take over is the executive’s responsibility, and it’s a space many leaders haven’t had to operate in before.

Executives owe their people upskilling, and employees owe the effort in return. Both framed this as a two-way responsibility. Sid sees it as critically important to upskill and reskill existing talent, not just hire for the future. ASQ has people with a year of tenure and people with 35 years, and he wants programs that serve across that whole range, which his people operations team is now putting in place. The flip side: if leadership is adopting AI constructively and has provided the skills, and an employee still refuses, that becomes a cultural-fit conversation. Nick added the employee’s half. It takes an employee of faith to raise their hand and say a tool could take over part of their job, trusting that a leadership team of good character and competence will reskill them or create new opportunities. That faith only works if leadership has earned it.

When there’s a right person but no seat, be honest and be generous. If the organization is shrinking and the revenue or investor pressure is real, hard decisions have to be made regardless of whether it’s the right person. You can’t create jobs to keep good people around. Reduction in headcount is never easy, and Sid said he takes it very seriously and feels it. The obligation is transparency: if there’s room and the person is willing to reskill, find the seat, but tell them plainly what’s happening. Nick went further into what he called the dark cave. When you do have to let someone go and it isn’t performance-based, make the severance generous, write the letter of recommendation proactively, connect them with recruiters, and do whatever you can to maintain the goodwill the employee gave first.

The executive skill set itself has to level up. System thinking at scale, governance, and accountability topped Sid’s list. AI isn’t a tool you can delegate the way you delegate a CRM or an ERP. Every executive has to have a role in the AI strategy, which wasn’t true of past technology. Nick noticed the irony: AI is supposed to make life easier and instead demands more of us. He compared it to education, learning things, then learning how to learn, then learning to think more analytically, and argued that what counted as an exceptional executive a few years ago no longer clears the bar. Sid added a specific new skill: prompt engineering. Not coding, but prompting. Nick joked it could replace college writing as a course, prompting for executives.

Operations and customer service give the fastest return; talent is the long play. Asked where AI delivers the most, Sid was direct. Operations is the bang for your buck, work that took two hours now taking twenty minutes. Customer service is close behind, chatbots built on real knowledge that cut call volume. The bigger, slower payoff is talent: invest in upskilling, reskilling, system and design thinking, and you build the foundation everything else sits on. For life sciences specifically, he flagged information gathering and product development, faster access to standards and regulatory information.

Advocacy for change has to be repeated, shared, and built on trust. For the people on the fence, neither the true believers nor the never-adopters, Sid’s advice is to build advocacy through transparency about what you’re trying to accomplish and why it matters, to the organization, the brand, and the individual. Most resistance to change is fear of losing a job, so leaders have to create a sense of security: we’re investing in this so you can work smarter, not harder, and we’re committed together. Nick added that the message has to be said repeatedly in different ways because a different version lands for different people, and advocacy can’t live in one person or one department. He returned to a formula he uses often: trust is character times competence. Employees will only extend the faith this transition requires if they believe leadership is both trustworthy in character and competent to deliver.

You can’t retrofit ethics. On keeping AI fair and aligned with organizational values, Sid was clear that ethics can’t be bolted on afterward. You start with what the organization stands for and work the AI through that lens. Feed it biases and you’ll get biases back. ASQ has an advantage here, since it lives in the world of standards where there’s little gray area, but for organizations with a potential bend, data quality is everything. Fair and ethical aren’t checkboxes. They’re continuous work.

Good data, not just data, is the real prerequisite. Nick shared an analogy from a recent conference: data isn’t power, good data is. Saltwater isn’t much use until you filter it. Many organizations discover that when they think they’re ready for AI, they’re not, because their data isn’t in order. Sid agreed so strongly that ASQ launched a specialized data quality credential, which has been getting good traction, built around the same three themes ASQ focuses on: innovation, risk, and data quality.

Think about return on time, not just return on investment. On ROI, Sid reframed the metric. Alongside return on investment, there’s return on time. A function that drops from five hours to three is real productivity in dollars and cents. If you’re looking to put in a dollar and get ten back, you’ll struggle to justify AI today. But a chatbot that cuts call volume 30 percent, built on the right system design and data quality, is real. Unless you’re launching a product directly tied to AI, the honest place to look for value is operational streamlining and faster, better-informed decisions. Nick was candid about his own real-world anxiety over the investment The FDA Group has made in its AI product, and Sid’s answer was to treat it as calculated risk with guardrails. ASQ launched an AI chatbot, Quincy, to give members access to its large repository of quality content in multiple languages, without burdening the bank, and it’s seeing good results.

One thing to bring back to your team

Ask who owns your AI strategy, and don’t accept a vague answer.

The most honest moment in this conversation was two experienced leaders admitting the ownership question has no clean answer, because AI touches every function. But “it touches everything” is not the same as “it belongs to no one.” Pick a structure, name the owner or owners, and stay willing to shift as you learn. Then ask the harder follow-up Sid and Nick kept circling: what do you owe the people whose work this will change, and are you actually delivering it, the upskilling, the transparency, the generosity when there’s no seat left?

The organizations that come through this well won’t be the ones with the flashiest tools. They’ll be the ones whose people trusted their leadership enough to lean into the change instead of bracing against it.

Sid Bhatnagar is the Chief Executive Officer of ASQ, the American Society for Quality, a global organization headquartered in Milwaukee with members in more than 130 countries and additional centers in Mexico and India. An engineer by training, he spent most of his career in technology strategy and implementation, serving as a CIO across manufacturing and financial services, including roles at Ziegler and Produce Alliance. He joined ASQ as Chief Information Officer in 2020, became Chief Operating Officer and Chief of Staff, and has served as CEO since 2023. He holds a bachelor’s degree from the University of Wisconsin-Whitewater and is completing an Executive MBA in International Business at Marquette University.

Connect with Sid on LinkedIn here.

Who is The FDA Group?

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on staff augmentation, auditing, remediation, QMS, and other specialized project work in Quality Assurance, Regulatory Affairs, and Clinical Operations.

With over 3,750 resources worldwide, over 325 of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee.

Here’s why 17 of the top 20 life science firms access their consulting and contractor talent through us:

• Resources in 75 countries and 48 states.

• 26 hours average time to present a consultant or candidate.

• Exclusive life science focus and expertise.

• Dedicated account management team.

• Right resource, first time (95% success).

• 97% client satisfaction rating.

Talk to us when you're ready for a better talent resourcing experience and the peace of mind that comes with a partner whose commitment to quality and integrity reflects your own.

Subscribe to The Life Science Rundown:

Apple | Spotify | YouTube | Web + Others

Check out our newly launched AI-powered QMS audit tool, AICA (the Audit Intelligence Compliance Assistant).

We’re resharing this conversation from 2023. The career questions in it haven’t aged, and it’s one we come back to often when people ask us how to think about their next move in regulatory affairs or quality.

Nobody grows up dreaming of a career in RA/QA. Most people, as Nick puts it, tripped and fell into it. And then they hit the industry’s opening paradox: you need pharma experience to get in, but you can’t get the experience until you’re already in.

Once you’re in, the questions don’t stop. Nick Capman sat down with Jessica Beaver, a regulatory and quality executive with over 20 years in the industry, to work through the ones people ask most in the first and second halves of their careers. Should I specialize or stay broad? Big company or startup? Stay a contributor or move into management? Go independent, or even start my own thing?

Jessica started as a pharmacist during graduate school, moved into large pharma as a contributing scientist on product development teams, and shifted into regulatory and then quality about halfway through her career. She’s worked across both pharma and medical device, from a seven-plus-year run at Targacept through executive roles at Kowa Pharmaceuticals and, most recently, as President and CEO of the medical device startup Verinetics.

Her throughline advice is simpler than any single answer: spend real time figuring out what you actually want, then make the move.

Apple Podcasts | Spotify | YouTube | Web + Others

Jessica’s key insights and practical takeaways

If you’re short on time, here are the most important lessons from the discussion.

Specialist or generalist comes down to what you actually like doing. Early in an RA career you often don’t get to choose. You start in a therapeutic area or molecule type and build depth there by default. The real question comes later: stay deep or broaden out. Jessica frames it as a genuine tradeoff, not a right answer. Specializing makes you knowledgeable and sought after. The cost is fragility, because work environments aren’t stable, and deep specialization makes a pivot harder. She leans generalist herself because she likes solving different kinds of problems and thinks of the work as puzzles. So ask yourself what you like: diving deep into one thing, or solving broadly across many. Nick’s shorthand for the choice is the laser versus the Swiss Army knife.

If you specialize, pick a niche with longevity. Jessica’s one caution for anyone going deep is to watch where the niche is headed. Nick’s example is radiopharmaceuticals, small but not necessarily shrinking, against his tongue-in-cheek image of specializing in buggy whips in the 1930s. Neither could name a current specialty they’d call a bad bet, which is the point. The move isn’t to avoid specializing. It’s to keep your head up, track where therapeutics and investment are flowing, and specialize in something durable rather than whatever’s hottest right now.

Stay current without drowning in information. The industry has more blogs, newsletters, and podcasts than anyone can follow, to the point of being overwhelming. Jessica’s advice is to pick a small number of sources, ask a mentor or supervisor what they actually read, and protect a little time each day for it, at lunch, before the day starts, whenever fits. Consistency beats volume.

The generalist path buys flexibility, with a real catch. A generalist is adaptable and can tackle many kinds of problems, which is valuable in consulting and in small organizations that change tactics fast. The downside mirrors the specialist’s advantage: some employers want a specialist with existing FDA contacts for a specific area, and a broad background can read as not deep enough for them.

Large companies teach you how the industry actually works. Jessica’s case for big pharma is education. You see specialists across CMC, clinical, and nonclinical, meet a lot of people, and get learning opportunities everywhere. For a regulatory professional who has to advise colleagues across all those functions, that exposure is hard to replicate. She spent the first part of her career there and values what it taught her, even though she hasn’t gone back. Nick adds that large companies tend to suit people focused on climbing the ladder, with the tradeoff being bureaucracy and personalities.

Big doesn’t mean stable. Both push back on the stability myth. Large companies feel more stable, and are, until a wave of layoffs hits, and Nick notes the last decade has been more volatile than the one before it. A large employer is not a job for life.

Startups give a regulatory professional room to shape the whole company. Jessica is openly a lover of startups at this stage of her career, and her metaphor is the best in the episode. In a large company, the people running alongside you toward success are close on either side. In a small company, the distance between you and the next person is much wider, which means there’s room to do more. A regulatory professional at a startup can influence the whole company’s success, not just a team’s, because so much of a company’s fate runs through regulatory, the stories you tell and how you assemble the information to get a candidate through the gauntlet. The risk is the obvious one: a startup can collapse when things don’t work out, with no broad portfolio to cushion the fall. Jessica’s answer is that the experience is still worth it, and you can carry what you learned into the next thing.

Nick’s pull toward startups comes down to three things. You wear a lot of hats, which broadens your experience the way generalist work does. There’s financial upside if you can negotiate equity, phantom stock, or options and the company succeeds. And you get to be part of a story, going from an idea to an established brand, which he finds genuinely exciting and useful to carry forward. He also makes a point about stress worth repeating: on a scale of one to ten, a two or three is healthy, three to five is manageable, above five is probably too much, and below two means you’re not growing. Startups can deliver the useful kind.

Small molecule or large, pharma or device, folds back into specialist versus generalist. Jessica notes that even the RA certification exams have shifted, from a regional, generalist approach spanning devices and pharma to a recognition that people now tend to work in one area rather than across. Her advice doesn’t change: go with what you feel called toward, whether that pull is scientific, problem-solving, or personal, like a therapeutic area that matters because of a family member. No area carries an inherent advantage. Principally the work is the same, building a story to tell toward a regulatory approval.

Moving from contributor to manager deserves real discernment, not reflex. Nick opens with his own story. His first job out of college, the hiring manager said he could become a manager within twelve months. He did it in eleven, because it looked like growth and he saw no reason not to. He was also removed from the role within three to six months, which he says was the right call. Jessica names the shift people underestimate: as a contributor you gauge productivity by what you produce, a report, a successful FDA meeting. As a people manager, what you’re producing is people, and you can’t point to that in a performance review the way you’d point to a finished deliverable. It takes time, the people depend on you, and it’s a real change in how you think and spend your time.

Being good at the work doesn’t make you good at managing. Nick’s addition is that a great baseball player isn’t automatically a great manager. Management is a separate skill set you have to learn. He also reframes the hierarchy: a contributor has one boss, a manager has many, and if you’re doing it well, you work for the people who report to you, not the other way around. Both reach for Kim Scott’s rock star versus superstar framing from Radical Candor. The rock star wants to stay an individual contributor in their space. The superstar wants to climb. Neither is better, and a good hiring manager figures out which someone is instead of assuming everyone wants to climb.

Listen to the people in your care instead of pushing your idea of advancement on them. Jessica closes the management section with something that’s stayed with her for twenty years. Early on, as a team lead, someone asked her why she thought everyone should have to advance or go back to school. She wasn’t saying that, but it was landing that way. The lesson she took, and still carries, is to actually listen to the people she leads rather than push them toward her own definition of advancement.

Going independent means giving up being part of something. Nick flags consulting as his world and teases a whole future episode on it. The core tradeoff Jessica names is belonging. As a full-time employee you’re on the team, invested in a shared goal. As a consultant you can still feel that sometimes, but often you’re viewed as someone who helps rather than part of the deeper, invested team. It also loops back to specialist versus generalist: you need to know what you bring. A specialist consultant needs clients who need that specialty. A generalist has more flexibility in who they can help, with the same catch, they may not be the first choice when a client wants deep experience in one thing. And there’s real financial risk, since consulting has an ebb and flow. Jessica points to the shift from the busy early-COVID stretch into a slower period as a reminder that independents have to watch the market and plan for their own stability.

Underneath all of it: spend time discerning where you want to go. Asked for closing lessons, Jessica returns to her theme. Her book recommendation is Marcus Buckingham’s Love and Work, which she keeps going back to. Its argument, in her telling, is that you don’t need to love what you do every second, but you do need to love it generally, because you spend so much of your life working. She reads by listening first, then returns to a hard copy with a notebook to pull out what matters, for professional and personal growth alike. Her parting advice is to stay flexible. She thinks about what she wanted to be as a child, which isn’t where she landed, and tells her teenager the same thing she’d tell an early-career professional: it’s the journey, not the destination. Set goals, because they matter for growth, but you may hit them faster than expected and then have to ask what’s next, so keep looking around at the opportunities that could take you somewhere you didn’t plan.

Guard your wellbeing, and build systems rather than chasing goals. Nick complements Jessica’s advice with a push on wellbeing. You spend more time with colleagues than family, so if you’re waking up with a case of the Mondays every day, you’re losing years and the ability to live well within them. Make a change. He also offers an unpopular opinion on goals: he focuses less on the goal and more on the systems, processes, and habits that put him in position to reach it. His analogy comes from Danny Ainge, that you can’t control the scoreboard, only your position to influence it. So he builds systems and holds himself accountable across the things that feed the work, when he wakes up, how he manages stress and mental health, how he keeps learning, and the relationships and sense of purpose outside the job that shape performance inside it.

One thing to bring back to your team

Block an hour this week to answer one question honestly: what do you actually want your next move to be, and why?

Nearly every fork in this episode, specialist or generalist, big or small, contributor or manager, employee or independent, comes down to the same thing. There’s no universally right answer, only the one that fits who you are and where you want to go. Most people skip the discernment and let momentum or a hiring manager’s challenge decide for them. The professionals who end up happy are usually the ones who spent the time.

Jessica Smith Beaver, PhD, is a pharmaceutical and medical device executive with over 20 years of experience across all stages of R&D. She began her career in clinical pharmacology focused on translational medicine, then moved into regulatory affairs and quality, and has since led regulatory strategy, quality systems implementation, and healthcare compliance across therapeutic areas including CNS, rheumatology, gastroenterology, women's health, oncology, and wound healing, with FDA interactions spanning CDER, CDRH, and CBER. She spent nearly eight years at Targacept in roles from clinical pharmacokinetics scientist to Senior Director of Regulatory Affairs and Quality Compliance, led regulatory and quality at KeraNetics, and served as Vice President of Regulatory and Business Operations at Kowa Pharmaceuticals America. Most recently she was Chief Regulatory Officer and then President and CEO of Verinetics, a medical device startup developing DispenSecur, an electronic lockbox to support take-home methadone access for people in treatment for opioid use disorder.

Connect with Jessica on LinkedIn here.

Who is The FDA Group?

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on staff augmentation, auditing, remediation, QMS, and other specialized project work in Quality Assurance, Regulatory Affairs, and Clinical Operations.

With over 3,750 resources worldwide, over 325 of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee.

Here’s why 17 of the top 20 life science firms access their consulting and contractor talent through us:

• Resources in 75 countries and 48 states.

• 26 hours average time to present a consultant or candidate.

• Exclusive life science focus and expertise.

• Dedicated account management team.

• Right resource, first time (95% success).

• 97% client satisfaction rating.

Talk to us when you're ready for a better talent resourcing experience and the peace of mind that comes with a partner whose commitment to quality and integrity reflects your own.

Subscribe to The Life Science Rundown:

Apple | Spotify | YouTube | Web + Others

Check out our newly launched AI-powered QMS audit tool, AICA (the Audit Intelligence Compliance Assistant).

We recently highlighted that the FDA has reported gaps in QMSR compliance after the first few months following the regulation's launch.

Today, we wanted to follow up on that with a companion piece from our own audit work. We’ve been running QMSR readiness assessments and gap analyses across clients' quality systems well before the effective date, and now that inspections are being conducted against the new framework, a few patterns keep recurring in our audits.

Most of them trace back to one assumption: that this was a relabeling exercise. It was not! The requirements are largely the same in substance, but several things changed that change how an inspection goes. These are the points we are flagging for clients, and the reason we want subscribers thinking about them now rather than after an investigator arrives.

If you still need to get compliant with the QMSR, talk to us ASAP. We’ll pair you with the ISO/FDA expertise that can be hard to find for projects like this.

Some records that used to be off-limits are now fair game

This is the single change most worth internalizing. Under the old QSR, 21 CFR 820.180(c) shielded three categories of records from FDA review during a routine inspection:

• Management review minutes

• Internal quality audit reports

• Supplier audit reports

Investigators couldn’t ask for them. In practice, that let companies document quality problems candidly in those records, knowing FDA would not see the underlying files.

That exemption is gone. The FDA did not carry 820.180(c) into the QMSR, and its own FAQ now states plainly that it has the authority to inspect management review, quality audits, and supplier audit reports. The new inspection program goes further and lists management reviews and internal audits among the elements investigators evaluate. So the records are not just theoretically accessible. The Compliance Program tells investigators to look at them.

Here’s what we’ve been seeing going wrong in practice:

Supplier audit reports deserve their own mention here. If your supplier “audit reports” are informal email summaries or notes from a phone call, they will not hold up now that FDA can request them.

Formal, structured supplier audit records, prioritized by the risk of the component or service, are worth standing up before an inspection forces the issue.

Risk now runs through the whole system, not just the product

We discussed this at length in our prior piece on the subject, but we definitely need to cover it again.

The old QSR mentioned risk exactly once, in the design validation section, and even there it said risk analysis "where appropriate." The QMSR, through ISO 13485, threads risk management through the entire quality system. There are two distinct things happening here, and conflating them is a common gap we flag.

• The first is product risk management, the hazard analysis and risk controls for the device itself. That lives in the world of ISO 14971 and it’s familiar territory.

• The second is newer as an inspectable expectation: the risk-based approach to the quality management system itself, in ISO 13485 clause 4.1.2. That clause requires you to apply risk-based thinking to how you control your QMS processes and outsourced processes, and clause 4.1 requires you to document how those processes interact.

The exposure this creates is that FDA can now cite you not only for a weak product risk file, but for QMS decisions that are not risk-based and for QMS processes that do not connect to each other.

If your complaint handling, CAPA, design, and postmarket surveillance operate in silos and do not feed one another, that is a defensible observation under clause 4.1. When we assess a client system, the interfaces between processes are often where the real gaps are, more so than any single procedure actually!

This is also where quality culture and top management enter the picture. ISO 13485 puts management responsibility up front in clause 5, and the new inspection posture expects to see leadership actively owning quality and risk across the product lifecycle. If the connections between your QMS processes are broken, an investigator can reasonably read that as a management and culture problem, and it lands on top management, not on the individual process owners.

“ISO 14971 isn’t a rule” is the argument that gets companies in trouble

The FDA did not codify ISO 14971 in the QMSR. Strictly speaking, it’s 'voluntary. We hear this used as a reason to push back on risk management expectations, and it’s a losing argument in an inspection. The FDA has recognized 14971 as a consensus standard for years; it trains its investigators to it, and it inspects to its expectations. The regulation points you to ISO 13485’s risk requirements, and the practical content of meeting those requirements looks like 14971 whether or not you cite it.

So your risk management file, functionally speaking, needs to do what 14971 does:

Read more

Sid and Nick get into why AI pressure now comes from boards and investors rather than the ground up, how to keep human judgment on both ends of the process, what executives owe their people through the change, and how to think about ROI when the technology won't sit still.

A few of the takeaways:

• AI made the executive a steward of adoption and culture, not just a decision-maker, and the pressure is coming top down

• Keep humans on both ends: ideation from people, work from the tool, validation back to people

• Lean toward critical decision-making, not tasks. Pilots and anesthesiologists get paid for what they know when things go wrong

• Hire for innovation, change management, and system thinking. Lead by pulling people along, not pushing

• Executives owe their teams upskilling; employees owe the effort. When there's no seat left, be transparent and generous

• You can't retrofit ethics, and data isn't power, good data is

• Think about return on time, not just return on investment. Operations and customer service pay off fastest; talent is the long play

About Sid Bhatnagar
Sid Bhatnagar is CEO of ASQ, the American Society for Quality, a global organization with members in more than 130 countries. An engineer by training and a longtime technology executive and CIO across manufacturing and financial services, he joined ASQ in 2020 as CIO, became COO and Chief of Staff, and has led as CEO since 2023.

About The FDA Group
The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on staff augmentation, auditing, remediation, QMS, and other specialized project work in Quality Assurance, Regulatory Affairs, and Clinical Operations. Learn more: https://www.thefdagroup.com/

These RA/QA news rolls are available in full only to paid subscribers. Upgrade here. If you haven’t already taken our reader survey, it should only take 1 minute here.

This is our biweekly recap of news and industry happenings curated for our paid subscribers. Our goal is to bring you the headline news and the stories, research, and insights that might be harder to find.

If you haven’t upgraded to a paid subscription yet, you can do so here. Also, make sure to follow us on LinkedIn and catch the latest episodes of our podcast, The Life Science Rundown.

Here are some key items from the past couple of weeks.

💊 = Pharma, biologic, and biotech.
💉 = Medical device, medtech, diagnostics, and digital health.

FDA Leadership and Administration

• Commissioner search reaches its final stages: The White House is reviewing a handful of potential nominees for FDA commissioner, and a person familiar with the search told POLITICO the goal was to make a pick this week. Several names surfaced over the two weeks: Bloomberg reported that White House policy aide and physician Heidi Overton is among the finalists, and later added New York Cancer & Blood Specialists CEO Jeff Vacirca. Axios also named Vacirca, Overton, and Principal Deputy Assistant Secretary of Defense for Health Affairs Stephen Ferrara. Norman “Ned” Sharpless and Alkermes CEO Richard Pops remain under consideration, and RBC analysts floated BIO CEO John Crowley and Ferrara. Overton previously coauthored academic papers with former Commissioner Martin Makary at Johns Hopkins.

• Administration vetting a full senior leadership team: Alongside commissioner candidates, the White House is reportedly vetting candidates to lead the drug and biologics centers, currently run on an interim basis by Mike Davis and Karim Mikhail. People familiar with the search told Bloomberg the administration aims to announce a senior leadership team alongside the commissioner nominee, and that former Oncology Center of Excellence Director Richard Pazdur is under consideration to return as a senior agency leader.

• Office of Therapeutic Products acting director to depart: 💊 Vijay Kumar will leave his post as acting director of the Office of Therapeutic Products, the super office responsible for cellular therapies, gene therapies, therapeutic vaccines, and plasma-derived products. Kumar wrote that he and agency leadership mutually decided not to renew his detail. He worked closely with former CBER Director Vinay Prasad and former Commissioner Makary and cosigned new policies on CAR T-cell therapies and manufacturing flexibility. OTP was led by Nicole Verdun until Prasad placed her on administrative leave about a year ago.

• Trump nominates Klomp as HHS deputy secretary: President Trump is nominating HHS Chief Counselor Chris Klomp, who also serves as director of Medicare and CMS deputy administrator, to be HHS deputy secretary. Klomp has already exercised authority over the FDA as chief counselor, and POLITICO reported he was involved in Makary’s ouster and the elevation of former Principal Deputy Commissioner Sara Brenner to Kennedy’s senior counselor for public health. Confirmation would give him greater oversight of the FDA. Sen. Bill Cassidy (R-La.), whose committee would vet him, gave the pick a positive review.

• FDA and MHRA to embed liaison officers in each other’s headquarters: At the Drug Information Association meeting, U.K. MHRA CEO Lawrence Tallon and FDA deputy commissioner for policy Grace Graham introduced an initiative to establish reciprocal liaison roles. An FDA representative will be embedded at MHRA headquarters in London, and an MHRA representative at the FDA’s White Oak campus. This follows an April announcement of closer cooperation on medical devices and the MHRA’s recent hiring of several former FDA digital health leaders.

• FDA and African Medicines Agency sign a cooperation statement: 💊 Also at DIA, Associate Commissioner for Global Policy and Strategy Mark Abdoo announced that the FDA and the African Medicines Agency signed a Statement of Cooperation to strengthen engagement on regulatory and scientific matters related to the products both regulate.

• Acting leaders emphasize continuity and staffing at DIA: Acting CDER Director Mike Davis said at a DIA town hall we attended that his top priority is delivering on the agency’s mission and that trust and integrity matter at a time of headwinds. He focused on staff retention and hiring and encouraged former employees to consider returning. Oncology Center of Excellence Director Angelo De Claro said advisory committees remain an important part of the review process and that the commissioner’s office supports convening ODAC meetings when the center wants to. Associate Commissioner for Legislative Affairs Brian Fahey described how the past year leaned on Complete Response Letters as a de facto advisory committee mechanism.

FDA General

• FDA launches Operation TrialBlazer to keep clinical trials in the U.S.: 💊 On June 22, the FDA rolled out its part of a broader HHS initiative meant to reverse the shift of early clinical research to countries such as China and Australia. The FDA’s contribution includes guidance revisions, new sponsor resources, and a planned pilot program aimed at speeding first-in-human trials and streamlining later-stage testing. The effort also involves the NIH, ARPA-H, and the Office of the National Coordinator for Health IT. Kennedy introduced the initiative in a Fox News op-ed. The medtech side is not yet included: the roadmap does not mention device issues or the Investigational Device Exemption system, so for now the effort is focused on drug policy.

• Phase 1 IND pilot in the works: 💊 Central to the effort is a proposed pilot to expedite Investigational New Drug applications. The FDA wrote in a request for information that accelerating early development in the U.S. requires reimagining the IND process. The pilot would recruit external research institutions to provide advice and preliminary review of sponsors’ IND submissions, and would explore rolling IND submissions that let the agency review completed components before the full application arrives.

• Research and evidence guidance overhaul accompanies the launch: 💊 As part of Operation TrialBlazer, the FDA issued three draft guidances and an RFI. The revised master protocols guidance adds about 10 pages clarifying the agency’s thinking on basket trials, informed consent, and how trials with sub-studies should handle IND applications, and is open for comment through August 24. A revised draft guidance on first-in-human dose selection using quantitative systems pharmacology relies on the ICH M15 framework with QSP-specific considerations.

• White House clears the substantial evidence guidance update: 💊 The Office of Information and Regulatory Affairs completed its review of the FDA’s revised draft guidance on demonstrating substantial evidence of effectiveness, marking it consistent with change, which clears the way for publication. The update to the December 2019 draft endorses one adequate and well-controlled trial plus confirmatory evidence as a viable norm, following the policy Makary and Prasad outlined in a February NEJM piece. Last year, 59% of new medicines approved by the agency’s main review office listed just one pivotal study in their original labeling. The guidance is open for comment through September 22.

• PreCheck pilot participants selected: 💊 The FDA chose seven drug companies building U.S. manufacturing plants for its PreCheck pilot, which promotes domestic production: Eli Lilly, Regeneron, Kyowa Kirin, Amneal Pharmaceutical, Cellares, Fujifilm Biotechnologies, and Kriya Therapeutics. The seven were selected from more than 80 requests. Participants will receive technical guidance as they build facilities and enhanced engagement with earlier inspections when they submit applications referencing those sites. Three facilities are in North Carolina, two in New York, and one each in New Jersey and Indiana. Pre-operational engagements begin July 1.

• White House clears the expedited IND RFI and a QSP dosing guidance: 💊 OIRA cleared the RFI for the expedited IND pilot, submitted June 3, on June 17. It also cleared a draft guidance on using quantitative systems pharmacology for dose selection in first-in-human trials, an unusually fast turnaround given it was received June 15. The QSP guidance was not on the drug centers’ guidance agendas for the year.

• Advisory committee returns for two previously rejected drugs: 💊 The FDA plans to convene the Cellular, Tissue, and Gene Therapies Advisory Committee on July 29 to discuss Capricor’s Duchenne muscular dystrophy therapy deramiocel, rejected last July, and separately accepted Replimune’s resubmission of the melanoma immunotherapy RP1, rejected twice before, with an advisory meeting planned for late July. The agency expects to decide on RP1 by August 2 and deramiocel by August 22. The plans suggest advisory committees may play a larger role than they did under Makary.

• FDA schedules a peptide compounding advisory meeting, and staff weigh against the peptides: 💊 Documents posted June 29 show FDA scientists concluded that publicly available data weigh against adding seven peptides, including BPC-157, TB-500, and MOTs-C, to the 503A bulks list that would let compounding pharmacies make them from bulk substance. None are components of approved drugs. All seven were previously flagged for potential safety risks before the agency removed them from that category in mid-April at Kennedy’s encouragement. The Pharmacy Compounding Advisory Committee meets July 23-24, with a roster that includes physicians and pharmacists who have promoted peptides, among them Robert Harshbarger, son of Rep. Diana Harshbarger (R-Tenn.), who urged Kennedy to hold the meeting.

• New technical guidance on immunogenicity datasets: 💊 The FDA released a final Level 2 technical specification on submitting clinical trial datasets evaluating the impact of immunogenicity on drug pharmacokinetics, describing recommendations for three datasets in CDISC’s SDTM and ADaM formats.

Read more

Welcome to Insider issue #50.

Also, hello to the 610 life science professionals who joined us last month. If you haven’t already upgraded to a paid subscription to unlock our in-depth analysis issues like this one, you can do so here. Read our past deep dives here.

This month, Jonathan Wacks covers two new pieces of FDA guidance. The first addresses CMC flexibility for cell and gene therapy products pursuing a BLA, and the language in it is genuinely remarkable if you’ve spent a career inside traditional pharma manufacturing. No process validation for investigational product. No cGMP controls for Phase 1 safety studies. Permissive release criteria. Jonathan explains why the agency is doing this and what it actually requires of a CGT firm on the ground. The second is the finalized human factors guidance for device marketing submissions, which introduces a yes/no flowchart that finally right-sizes the HF data package to the actual risk of the device.

Leon Chagal closes the issue with a warning letter to GC America, Inc. It’s a textbook multi-system breakdown across component testing, stability, and data integrity, with a quality unit that had essentially gone dark. Leon walks through the citations and lays out what remediation actually looks like.

Download our slide decks to use any way you’d like:

Read more

This is a guest post from Nick Capman, President and CEO of The FDA Group.

This year, I went to DIA to present on a panel, but the honest truth about any conference is that the useful part is the rooms you sit in, not the one you talk in. So here’s what I’ve been chewing on since.

Two themes ran through almost every session. The first was AI, which surprises no one, though what struck me was how uniform the qualifiers were. Speakers kept saying the same thing in different words: AI can cut risk, add speed, and show you things you’d otherwise miss, but only alongside real data governance, validation, transparency, and someone accountable at the end.

The second theme was quieter and, I think, more important. The science is now moving faster than the systems built to handle it. That showed up in drug repurposing, GLP-1s, rare disease work, trial modernization, regulatory technology, and precision medicine. The opportunity is real here, but so are the constraints around access, affordability, trust, data quality, and regulatory alignment.

DIA framed itself this year as neutral ground where regulators, industry, academics, patients, payers, investors, and technologists can actually meet. The theme was finding hidden potential in today’s science and turning it into tomorrow’s therapies through collaboration that’s organized rather than accidental.

They also announced a couple of global efforts worth watching, a DIA and MHRA Global Summit and an Africa Access Summit.

The session I keep telling people about

David Fajgenbaum’s keynote on drug repurposing was the one that stayed with me the most. His group, Every Cure, is using AI and knowledge graphs to scan something like 75 million drug-disease pairs, built out of roughly 4,000 approved drugs and about 18,000 diseases. They’ve cut the time to run that kind of scan from around 100 days to about 17 hours.

The sort of uncomfortable part is why nobody had done it. A lot of the promising matches involve generic drugs, which often carry less commercial incentive to develop. So the treatments may already exist, sitting in plain sight, and the reason they haven’t reached patients isn’t scientific. It’s that the money doesn’t line up.

His ask to the room was practical: share your repurposing signals, contribute data, bring forward abandoned assets, and work with regulators on non-commercial pathways. I’d like to see our corner of the industry take that seriously.

GLP-1s and the mess a breakthrough exposes

The obesity plenary, on the past, present, and future of care in the age of GLP-1 therapies, did two things at once. Robert Califf called GLP-1s a genuine scientific breakthrough, with benefits reaching well beyond weight into cardiovascular outcomes and other organ systems. Then he used the same breakthrough to point at everything it exposed: high cost, supply shortages, unregulated compounding, misinformation spreading online, and access that isn’t remotely equal.

Scott Butsch of Novo Nordisk made the case that obesity is a physiological disease, not a lifestyle choice or a failure of character, and walked through how the newer therapies change the body’s biology, including its weight set point, with some benefits that seem to run past weight loss itself.

Brad Jordan of Eli Lilly stayed on access and stigma and the real danger of unapproved compounded products, and argued that industry and regulators have to deal with quality, safety, misinformation, pricing, comorbidities, and prevention together rather than one at a time.

Where AI is earning its place in development

Audrey Greenberg of Mayo Clinic Ventures moderated a panel on where AI is actually creating value in drug development and where people need to calm down. The framing I liked was AI as a biological risk reduction engine, useful for target selection, biomarker discovery, patient stratification, toxicology, drug-drug interaction analysis, and enrollment.

The nearest-term wins looked like discovery, reducing risk before you ever get to the bench, toxicology, and trial optimization.

They also got into money (which conferences usually avoid). One idea was clinical trial insurance, where investor capital gets reimbursed if a trial fails, which could free up financing for mid-stage assets that are hard to fund under normal risk math. The regulatory side covered sandboxes, data infrastructure, rare disease pathways, and AI-assisted submission review.

The panel also kept circling back to the same non-negotiables: transparency, reproducibility, third-party vetting, published evidence, and a human being who is accountable. You’ll notice that list shows up everywhere at this conference.

Quality is moving from sampling to surveillance

For years, “AI in quality” was mostly slideware and hype. This year it looked like something people are actually running now.

One of the core arguments was that traditional QA, leaning on manual sampling and point-in-time site audits, can’t keep up with the complexity and data volume of modern trials, and that E6(R3) is pushing everyone toward continuous, proactive, risk-based oversight built on analytics. The examples were specific enough to believe!

There was a business-case point aimed squarely at quality leaders, and it’s one I’d repeat. Stop justifying analytics on financial ROI alone. The stronger frame speakers used was value of investment, which counts time saved, cost reduced, inspection readiness, data maturity, workforce development, and regulatory risk taken off the table. Those are harder to drop into a spreadsheet, and they’re the real reasons to do this.

The agencies seem to be behind the direction.

• Staff from the FDA described moving toward directed inspections built around critical-to-quality factors and leaning on Remote Regulatory Assessments.

• MHRA was encouraging about innovation but firm that anything new has to be fit for purpose, independent, documented, piloted, and validated before you rely on it.

How to read a 483

A separate quality panel took on Form 483s, and I wish more people had been in the room for it. The core message was that a 483 is not automatically a critical finding or a serious breach. It’s a formal communication tool, and it calls for a proportional, evidence-based response with strong root cause analysis and interpretation that accounts for context, not a company-wide panic every time one lands.

Read more

AICA, the Audit Intelligence Compliance Assistance, is a patent-pending AI-powered compliance auditing tool we launched this year that analyzes all of your QMS documentation at machine speed.

It eliminates the sampling and blind spots common with fully manual document reviews. Upload your documents and get results the same day. If you’re new to AICA, learn more about it at aica.thefdagroup.com. Get a personalized demo here, or watch our intro below.

A lot of teams getting in touch with us about AICA are asking about how to actually integrate it into everyday operations. We run through it in every demo, but we wanted to take a minute to explain it in detail.

Let’s back up and start with some context: Just about every Quality department runs on a quiet, largely unsolved problem:

Very few teams have the headcount and hours to read every SOP against every applicable requirement, so review gets sampled. People check what they can, trust that the rest holds, and find out otherwise when an auditor, a customer, or an inspector reads the one procedure nobody got to.

That gap between what a Quality team is responsible for and what it can actually review is the space AICA was built to fill. To understand how, we talked with two consultants who helped develop it: Neal Siegel, a senior auditor with decades of QMS experience, and Mohammed Yousffi, a former FDA investigator who spent most of his time on the project training the system.

They explained what the tool does, what it does not, and how a Quality team should actually use it.

One job, done well

Start with what AICA does, because the honest version is narrower and more useful than the marketing version. We covered this in more depth here:

In short though, AICA reads a procedure against the regulation and judges whether the procedure contains what the regulation requires. That is the whole job. Neal puts it plainly:

“AICA looks at a regulation and says, ‘here are the things you should be doing.’ Then it takes your procedure and answers a single question: are the steps to meet those requirements actually written down? That’s the work. It’s not a checklist. It’s a contextual comparison between what the regulation requires and what your document says.”

The gaps it finds are often the quiet ones. Neal described a site where he asked the Quality team about their record retention period. Everyone knew it was five years. The managers, the QA staff, the CEO when he walked past. The number just was not in the procedure anymore. It had been dropped in a revision, and nobody noticed, because everyone already knew the practice. Harmless that time, but the same mechanism, a required element quietly missing from the document, is exactly what an inspector finds and what AICA is built to catch first.

Why it’s not just ChatGPT pointed at an SOP

The obvious question from anyone who has used a general AI tool is what makes this different. The answer is the part Mohammed spent his time on, and it is the reason the tool is worth trusting.

A general model, handed a regulation and a procedure, flags everything that looks like a gap. Many of those flags are wrong! Mohammed, Neal, and several other auditors fed AICA real procedures from companies they had audited and corrected its instincts, case by case, until it stopped raising things that did not matter.

His example is the one to hold onto here:

"There might be a requirement to have a QC lab, but you don't need to write that into a procedure. You can walk over and see that it's there. We taught the system: as long as the physical utility is there and you can verify it, that does not have to be in the document."

A raw model does not know this! An auditor does! That difference, multiplied across hundreds of real documents, is what keeps AICA from burying a reviewer in false positives. As Mohammed put it, "that sharpening is the uniqueness of it. A general model relies on hearsay. This was verified by people who actually use the regulation and write observations." The judgment of an experienced auditor is part of what got built into the tool, not left entirely to the person reading the output.

How to use it day to day

This is where the idea comes down out of the sky and into the department. Mohammed laid out four uses, and they move from a one-time check to something woven through the week, month, etc.

1. Start with a baseline

The first pass with AICA is diagnostic. Run your current procedures through the tool to see where the gaps are, then decide what to fix and in what order.

“The first thing I’d do is use it to see where my deficiencies are, so I know what to realign to. It is the use most people picture, and it is actually the smallest of the four.”

2. Then make it continuous

This is where it stops being a project and becomes part of the work. Procedures never hold still. Mohammed worked at a company with a vault of roughly 3,000 SOPs that revised close to 300 of them a month. At that pace, something slips past human review, not through carelessness but through volume.

"From a human perspective, we are always going to miss something. You're updating that many procedures a month. This is the double check."

Used this way, AICA runs against the regulation continuously and flags what has been addressed, what has not, and what looks risky enough to revisit. It becomes a standing check on a system that is always moving. Neal frames the same habit from the quality-culture side: a manager who runs it regularly reads the report not as fires to put out but as a running list of improvements, which is the evidence of an active, functioning quality program that an inspector wants to see.

3. Build your risk register before an inspection

The third use is preparation. The tool surfaces gaps, and the team turns them into a self-identified risk register with a plan attached. Mohammed is careful about what that is:

"You're not handing this to the regulator. But you have an internal plan you can take to management. We self-identified these. Here is the three-to-five-year plan, here is the stopgap until the funding comes."

Neal draws the same line. The report is an internal tool, not a document you give an investigator. Its value in preparation is that it lets a company show, with a documented plan, that it found its own problems and is working them.

If a gap surfaces during an inspection, the answer is already in hand: we know, here is the plan, management is aware, here is the interim control. That posture is worth far more than a clean-looking binder.

4. Sharpen your audits (internal or supplier!)

The fourth use is the one Mohammed knows best, because it’s his job. When he audits a supplier or conducts an internal audit, he is on-site for a day and has to cover six or seven systems. There is never enough time to read everything.

If he can get the procedures in advance, he runs them through the tool first. AICA sorts what it finds into confirmed, unconfirmed, and potential observations. He filters to the confirmed list, validates each with his own judgment, and walks in with a focused line of questioning.

"Say they give me the calibration procedure. Some of these run 10 to 50 pages, with an umbrella procedure and one for every piece of equipment. I run it through, it tells me where the gaps are, and that becomes my line of questioning on site."

The heat map does the triage here (example below). Instead of reading fifty pages at the same depth, he spends his limited hours where the report shows the most red.

This is the same move Neal described from the audit-prep side: walking in able to say, “here are the six procedures I want to look at out of everything you gave me.” The slow part happens before anyone is in the room, so the time on site goes to the questions only a person can ask.

Where the human stays essential

A tool that found everything and judged it correctly would not need a reviewer. AICA is not that tool (nor do we want it to be), and both consultants are direct about it.

One reason why we’re adamant that humans be kept in the loop is that AICA only sees what you give it. If you don’t upload a procedure, AICA will report it as missing, which is true of what it was handed, but possibly false about your quality system. As Neal puts it, absence in the report does not mean absence from the QMS. The tool is built to ask the reviewer to confirm rather than to declare a non-conformance outright.

Also, it reads documents, not records. It’s not opening batch records, deviation files, or CAPA histories, so it cannot tell you whether a procedure is actually being followed. That question stays with a person (where it should at least for now).

AICA obviously can’t walk the floor, either. An equipment or facility issue that an investigator would see on site is outside what a document review can reach.

And it’s not perfect, which Mohammed says plainly. The output is a strong, fast first pass that a qualified person then screens, validates, and prioritizes. The tool screens and challenges. The expert decides. That division is the design, not a shortcoming, and it is the reason the results hold up.

Where it earns its keep

AICA pays off most where the work is heaviest. Neal’s point is that the value compounds in large, multi-site operations, where the same procedures repeat across departments and a traditional review pulls senior people off their jobs for a week. Cut that to a few days and surface the worst procedures first, and you give back real time on every audit cycle.

Mohammed’s point is that the reach is broad: because the regulation covers the whole operation, every department governed by a procedure can use it, from facilities and equipment to production to the lab.

The piece that was missing, which AICA now fills

Step back, and the shape of the value here is clear: For years, a Quality team had two ways to check procedures against the regulation. People could read them, slowly, a sample at a time, with results that varied by who did the reading.

Or a general AI tool could read them quickly and return a pile of flags too unreliable to act on. Neither covered the actual need, which was a fast, consistent first pass across everything, grounded in how an auditor actually reads.

• That first pass is what AICA adds. It doesn’t replace the Quality team, it does not review records, and it does not solve compliance.

• What it does is take the one layer that was always too large to do by hand, the line-by-line comparison of every procedure against the regulation, and do it at machine speed with auditor judgment built in, so the experts can spend their time where judgment is the work.

For a Quality department that has spent years sampling what it could and trusting the rest to hold, that is the missing piece.

This article draws on conversations with Neal Siegel, a senior quality consultant, and Mohammed Yousffi, a former FDA investigator and GxP compliance consultant, both of whom contributed to AICA's development. Quotes are lightly edited for readability.

Read our other AICA articles:

• Eric Boyd Introduces the Latest Version of AICA in 3 Minutes

• What AICA Actually Sees When It Looks at Your Procedures

• AICA's Security Architecture: What Your IT Team Needs to Know

• A Few Use Cases for AICA

• Your AICA Questions, Answered

• AICA is Officially Live — Here's Everything You Need to Know

Have questions or want to get a demo?

Want to see whether AICA fits your team? Visit aica.thefdagroup.com to learn more and get in touch. We’re onboarding teams to the tool right now.

Share The FDA Group's Insider Newsletter

This guidance breakdown is available in full to paid subscribers. Only paid subscribers get regular full access to our guidance breakdowns and other analyses. If you’re not already a paid subscriber, you can upgrade here.

We’re catching up with the FDA’s recent guidances, starting with a new final one from CDER’s Office of Generic Drugs: Certain Postapproval Requirements and Resources for ANDAs.

The story behind it is short: The FDA has been trimming the standardized postapproval reminders it used to print in ANDA approval letters and consolidating them into this reference document instead.

If your team has treated the approval letter as its postapproval checklist, that habit is now on borrowed time! The obligations themselves have not changed. What’s changing is that the FDA is no longer going to spell them all out for you at the moment of approval.

One caveat the guidance makes about itself is worth stating up front: this document does not cover every postapproval requirement that may apply to an ANDA, nor does it list every relevant CDER guidance. It covers “certain” requirements and points to “related” resources. Treat it as a starting map, not the whole territory. Everything below applies to ANDAs that have received final approval.

Here’s what’s on the map.

Changes and supplements to an approved ANDA

Some manufacturing changes to an approved ANDA require an approved supplement before you can make the change and distribute product made with it. That comes from section 506A of the FD&C Act, with the implementing regulations at 21 CFR 314.70 and 314.97.

How much reporting a change triggers depends on its potential to affect the product, and the long-standing reference is Changes to an Approved NDA or ANDA (April 2004, Rev. 1). The point worth holding onto here is that if a change is the kind that needs prior approval, you cannot make it first and report it later.

The tentative-approval trap after a lost patent suit

Read this section closely if you have launched, or plan to launch, on a paragraph IV certification while patent litigation is still live.

Let’s run through it:

The exposure is real here. A product you are already selling can have its approval pulled back to tentative, and the 14-day window to get documents in is short.

The reference is MAPP 5220.2.

Postmarketing reporting: safety, combination products, and shortages

Postmarketing reporting applies whether or not you are actively marketing the product. The core regulations sit at 21 CFR 314.80, 314.81, and 314.98, and they include keeping the FDA informed of the product’s marketing status. FDARA added section 506I in 2017, which layers on additional marketing-status reporting in certain circumstances, with the companion guidance Marketing Status Notifications Under Section 506I (August 2020).

On safety surveillance, every approved ANDA holder has to maintain written procedures for collecting, evaluating, and reporting postmarketing adverse drug experiences, and has to review adverse experience information from any source, foreign or domestic, including commercial use, postmarketing studies, and the published and unpublished literature.

Periodic safety reports are due quarterly for the first three years after the U.S. approval date and annually after that. The annual report has to summarize significant new information from the prior year that could affect the product’s safety, effectiveness, or labeling, plus what you have done or plan to do about it. The current standards reference is the final E2D(R1) guidance (March 2026, Revision 1).

If your product meets the definition of a combination product under 21 CFR 3.2(e), the combination-product postmarketing safety reporting rule applies on top of the standard requirements. See Postmarketing Safety Reporting for Combination Products (July 2019).

Shortage notifications are their own obligation, and timing matters to the FDA here.

• Under section 506C, you have to notify the FDA of a permanent discontinuance or an interruption in manufacturing that is likely to cause a meaningful supply disruption, and that duty covers both finished products and the API for those products.

• Four triggers in total:

  • Discontinuance of a finished product

  • An interruption in a finished product likely to meaningfully disrupt supply,

  • Discontinuance of API

  • An interruption in API likely to meaningfully disrupt API supply

The notification has to state the reasons. The reference is the draft guidance Notifying FDA of a Discontinuance or Interruption in Manufacturing of Finished Products or Active Pharmaceutical Ingredients Under Section 506C (February 2024, Rev. 1).

Promotional materials

When you put out promotional labeling or advertising for a prescription generic, you have to submit a specimen to the FDA at the time of first dissemination or first publication, on a Form FDA 2253, with the product’s current professional labeling attached.

Separately, you can ask the Office of Prescription Drug Promotion for advisory comments on proposed launch materials before they run. That review is voluntary, submitted in eCTD format with the draft or mock-up materials, annotated references, and the applicable labeling. It is a useful check when a campaign is high-stakes, but using it is your call.

Annual facility fees and self-identification

If you own a facility that makes generic drug products or APIs, or a site that supports their manufacture or approval, GDUFA requires you to self-identify with the FDA electronically and update that information every year.

Self-identification feeds two things: the list of facilities that owe user fees, and the FDA’s visibility into the generic supply chain. Most facilities that self-identify owe an annual facility fee. GDUFA III runs from October 1, 2022, through September 30, 2027. For the mechanics, see Self-Identification of Generic Drug Facilities, Sites, and Organizations (September 2016) and the FDA’s GDUFA pages, which carry the current-year fee amounts.

Keeping labeling current

Content of labeling goes to the FDA electronically, currently in structured product labeling (SPL) format, under 21 CFR 314.94(d).

The part that gets under-resourced is the ongoing duty.

Your obligation to keep labeling accurate does not end at approval. When new information makes your labeling inaccurate, false, or misleading, you have to act to correct it, and you are expected to watch the RLD’s labeling for changes and update yours to match. In practice that means routinely checking Drugs@FDA and the Orange Book rather than waiting to be told. The how-to is Revising ANDA Labeling Following Revision of the RLD Labeling (January 2024).

REMS obligations for generics

If the FDA requires a REMS for the listed drug your ANDA references, your ANDA needs a REMS too!

A few specifics particular to generics are worth knowing:

• Your ANDA’s REMS has to carry the same Medication Guide or patient package insert, and the same packaging or disposal requirement, as the listed drug.

• If the listed drug’s REMS includes Elements to Assure Safe Use (ETASU), your ANDA can either join a single, shared-system REMS with the listed drug or use a different but comparable aspect of those elements, though the FDA can require the shared system if it concludes no comparable alternative would satisfy the statute.

• An ANDA REMS does not include a communication plan, and ANDAs are not subject to the standard statutory timetable requirement for REMS assessments, although the FDA can still require REMS assessments when needed and include the schedule in the ANDA approval letter.

Two exclusivity clocks, two different notification rules

Sections H and I of the guidance both deal with telling the FDA when you start selling, and the fastest way to get burned is to assume the two work the same way/ They do not!

The short version: CGT is a 75-day deadline to start selling, with notice that is technically optional but practically required. Paragraph IV is a 30-day deadline to report selling once you have started, with notice that is mandatory and a costly default if you miss it.

Compendial standards

If your drug’s name is recognized in the USP-NF, it generally has to meet the compendial standard for strength, quality, and purity unless any difference is stated plainly on the label.

One practical wrinkle the guidance flags here is that the FDA usually cannot share the application-specific information in your filings with third parties, and that includes USP. So if a monograph needs to change for your product to keep complying, you may need to work with USP directly to revise it.

A few checklist items to consider

If you hold approved ANDAs, here is where this guidance should prompt some internal work. A few of our suggestions:

• Build or refresh an internal postapproval obligations register, owned by a named person. With the FDA pulling reminders out of approval letters, the tracking burden shifts entirely to you. Map each obligation in this guidance to a responsible function, a trigger, and a deadline, and stop relying on the letter to surface them.

• For any at-risk or paragraph IV launch, pre-stage the 314.107(e) document package and assign the 14-day clock to a specific owner before you go to market. If you lose the patent suit, you do not want to be assembling court documents from scratch against a two-week deadline while also managing the commercial fallout of a possible conversion back to tentative approval.

Read more

Today’s piece is for the device and medtech teams among us. As you know by now, the FDA’s Quality Management System Regulation took effect earlier this year, folding ISO 13485:2016 into 21 CFR Part 820.

The agency also retired the Quality System Inspection Technique that investigators had used for decades and switched to a new inspection process under Compliance Program 7382.850.

A few months in, the early data is starting to arrive. FDA officials have now described findings from the first wave of QMSR inspections at several industry conferences, and the picture lines up closely with what our own auditors and quality consultants are seeing in readiness work.

If you still need to get compliant with the QMSR, talk to us ASAP. We’ll pair you with the ISO/FDA expertise that can be hard to find for projects like this.

The early numbers, with a caveat

At the FDA Law Institute’s annual conference in May, Keisha Thomas, associate director in CDRH’s Office of Product Evaluation and Quality, said the agency had completed just over 100 inspections under QMSR.

For Form 483 observations issued between February and mid-April, she ranked the top areas as follows:

• Risk management

• Outsourcing and purchasing

• Complaint handling and feedback

• UDI

• Corrective action.

Her framing of the central change was pretty blunt:

“Risk, risk, risk, risk. That is the fundamental change to QMSR.”

The ordering of these areas is not fixed yet. At MedCon in late April, officials put risk management first, followed by outsourcing and purchasing, for the February-through-March window.

Speaking at the RAPS Quality Conference about four months in, Thomas listed risk management, then corrective action, then risk-based approach, complaint handling, and purchasing. She added the caveat worth keeping in mind: it’s early, and the agency is largely seeing the same citations it saw before QMSR, only in a different order.

That last point is the one to sit with. The categories are old. What changed is how an inspection arrives at them.

Risk is now the roadmap (and the audit trail is open)

Under the old QSIT, investigators worked through a handful of subsystems, each more or less on its own. Under the new CP 7382.850, the inspection is organized around six QMS areas and built to follow product risk across the lifecycle. Investigators identify the risks a device could pose, then use the company’s own risk documentation to navigate the rest of the system. Risk management is no longer one file among many. It’s the route the inspection takes.

One change deserves more attention than it has gotten. QMSR removed the long-standing protection that kept internal audit reports, supplier audit results, and management review records out of the FDA’s reach during an inspection. Those records are now fair game.

There’s a related and easy-to-miss trap here: some companies read a clean internal audit history as proof of strong compliance, when an absence of findings can just as easily mean the audits are not looking hard enough. The agency is encouraging firms to run internal audits more like an FDA inspection, aimed at finding real problems before an investigator does.

With that context, here is where the early problems are clustering, and what our auditors tend to find underneath each one.

Risk management

The clearest signal is that risk management is being treated as a living system rather than a design deliverable. Thomas said firms are being cited for a lack of process and evidence showing that risk is actually rooted in their decisions, and she pressed the point that risk management has to keep evolving after launch.

Many companies still manage the risk file as something built during design and revisited only when a formal change or audit forces it.

On paper, the file looks complete. What it often can’t show is the connective work: how complaints, nonconformances, service data, field failures, and supplier issues feed back into the analysis, who owns those updates, when they trigger a change, and how control effectiveness gets verified over time. Producing an FMEA is no longer the test. You have to show that the FMEA reflects the current product performance.

Supplier and outsourcing controls

Supplier control has always drawn inspection attention, but QMSR sharpens the question from “was this supplier approved” to “is the level of control proportionate to the risk this supplier carries.”

The weakness our auditors find is rarely an empty supplier file. It’s much more often a generic one. We’ve seen quite a few low-risk packaging vendors and critical contract sterilizers governed by the same template, the same review cadence, and the same thin quality-agreement language.

Under an inspection that starts from risk, that flatness stands out pretty quickly.

Complaints and feedback

Complaint handling is surfacing as an early observation area, and it usually fails at the handoffs rather than the procedure.

Read more

The easy part of leadership is the part with a “right” answer. But most of the job isn’t that! It’s the decisions where the facts, the opinions, and the emotions all pull in different directions, where you’re making the best call you can on limited information, and people are watching to see what you do next.

Nick Capman sat down with Sterling Chung, Chief Regulatory and Quality Officer at Aurion Biotech, to talk about exactly that: how leaders handle tough conversations and stay grounded when there’s no clean answer to reach for.

Sterling has spent his career coming up through regulatory and quality from the ground level, across companies large and small, through major submissions and the kind of pivots that change a company’s direction. His take on where leadership actually gets tested is simple. The good times are easy. The tough times are where you find out what leadership is really about.

Key insights from the discussion

The higher you go, the fewer clear answers there are. Coming up through an organization, you think the answers are obvious. Then you reach positions where a single decision you once saw as a straightforward regulatory or quality call turns out to carry a dozen competing factors. The skill that replaces certainty is judgment: taking in the information, weighing what’s best for the organization, and making the best educated guess you can, often on limited facts. You won’t be right every time! Sterling borrows from baseball here. A player who bats .300 is admired because hitting a third of the time safely is enough to be great. Leadership works the same way. You average out good decisions over time, and you keep stepping up to make the next one.

Leadership humbles you, and servant leadership is the model worth chasing. Sterling describes being humbled over the years by what leadership actually demands. The leaders he admires are the ones who think first about who they’re serving, whether that’s an individual, the organization, patients, or physicians. One realization stands out: you have far more customers than you think. It’s easy when you’re a junior person answering only to the person above you. It gets harder, and more meaningful, when you’re answering to the whole organization and to the patients at the end of the chain.

Trust is character times competence. Nick offered a formula he’s borrowed for years: trust equals character multiplied by competence. You need both. Good character without competence won’t earn the trust of the people who report to you, and competence without character won’t either. Get both right and you’re pointed in the direction of actually leading people somewhere.

Prepare for hard conversations by bringing facts and then listening. When Sterling can prepare for an emotionally charged conversation, his approach is to weigh what’s best for the company against what makes sense for the individual, and look for the overlap, even though the two interests aren’t often aligned. The preparation itself comes down to arriving with the facts of why the conversation is happening and being ready to talk it through rationally, paired with genuine empathy for the other person’s situation. Then comes the part people underrate: listening. You’re usually dealing with smart people who can see the situation clearly. Often they don’t need an answer. They need to be heard. And when their feedback surfaces something worth adjusting, make the micro-adjustment and acknowledge the part they got right.

When a hard conversation arrives faster than you can prepare, anchor to two things. Roughly a quarter to half the time, these conversations erupt before you’re ready. In that case Sterling narrows to two questions. What is the goal of this conversation, and what needs to come out of it? And how did this situation actually come about? Since you won’t have all the facts in the moment, much of the work is simply listening for why something grew bigger than it should have.

Accountability is what people remember, and it’s where growth comes from. Both Nick and Sterling kept returning to this. Owning a mistake, saying “I got this wrong, let’s fix it together,” is what lets an organization and its people relate to you. The humbleness underneath it matters more than the apology itself, because we’re fallible by design, and only by truly owning a mistake can anyone improve on it. The alternative, leading from “I’m the boss and I’ll tell you how it is,” produces no growth in any direction.

Transparency protects you from the stories people invent. Nick’s argument for transparency is blunt: if you don’t tell people what’s going on, they’ll write their own version in their heads, and nine times out of ten the version they invent is worse than the reality. And accountability alone isn’t enough. As he put it, the truth only sets you free when it comes with hope. “I made a mistake” needs to be followed by “and here’s how we’re going to fix it.” Without the second half, you’ve given people a problem and no path out of it.

If people can’t see progress, they feel trapped. Sterling’s phrase for a stalled situation is purgatory. Employees feel stuck in it when they sense their leader won’t grow alongside them. His own animating mindset, held since he was young, is to never stop growing. He may not start a race fastest, but he intends to go the longest distance having learned, because the lessons compound. The truest leaders he meets are often the humblest people in the room, the ones still saying there’s so much more to learn.

Vulnerability does the listener a service, not just you. Nick connected this to Brené Brown’s work on vulnerability, including her widely viewed TED talk and her book Dare to Lead. When a leader is open about their own struggles, it tells the people listening two things: that you’re human like them, and that it’s normal to be working through hard things. The opposite, projecting a life where everything is perfect and nothing ever goes wrong, leaves people thinking they’re the only ones falling short. Withholding your humanity doesn’t just shortchange you. It shortchanges them.

Distill facts down to tables to take the emotion out. This was one of the most practical ideas in the conversation. Sterling credits a colleague, Andrew Torres, with the principle that facts can be distilled into tables. When a decision is crowded with inputs and feelings, agree first on the structure, the rows and columns, the X and the Y, the goal everyone is working toward. Once everyone accepts that the data set is the data set, you can read what it actually says without the emotion. Much of what makes a decision feel charged is that the right and wrong of it isn’t clear to anyone. A shared table makes it clear. Sterling applies this in data-heavy CMC and quality work, and also in softer, more variable areas like clinical, where the primary endpoint is the conclusion of the story and the secondary endpoints are the supporting cast. Laying the studies out in a table lets a team ask, honestly, whether the secondary endpoint someone feels strongly about actually supports the arc, or just stands alone.

Communicate to be understood. In Nick’s household, the family runs on three values: no lying, no disrespect, and communicating to be understood. The third is the leadership lesson. It’s often not what you say but how you say it, and the ability to communicate so that you’re actually understood is half the battle. Sterling closed the episode on the same note: even within a shared language, people misunderstand each other constantly, and part of a leader’s job is overcoming those barriers by having the frank conversations rather than avoiding them.

Nobody is purely good or purely evil, so do the work to understand people. Both men landed on the same idea from different directions. Real people aren’t heroes or villains. Calling someone crazy, an idiot, or evil is usually a way of declining to understand them. The harder, more useful move is to remember that the difficult person is also somebody’s parent, child, sibling, or friend, and to spend the extra energy figuring out where they’re coming from. It produces understanding and outcomes you’d otherwise never reach.

The decisions that affect people’s livelihoods are the heaviest, and Sterling is honest that he still wrestles with them. An across-the-board cut by percentage is, in a strange way, the easy version, because it’s arbitrary and applied evenly. The genuinely hard call is the individual one, where something isn’t working for the person or the organization. Sterling’s hard-won view is that the most honest thing is sometimes to part ways frankly and transparently, without assigning blame, because everyone owns it: you hired them for a reason and saw potential, and in this situation it simply isn’t working. Often the person isn’t happy either, and parting can turn out to be what’s best for them.

Extreme ownership is a prison that turns out to be freedom. Nick referenced Jocko Willink’s Extreme Ownership and the idea that everything is your fault. It sounds confining and ends up being liberating. Applied to people who don’t work out, the logic is that you either hired them that way or trained them that way, so the responsibility lands with you more often than not.

Half of how someone remembers you is how you handled the final moments. This was Nick’s most striking lesson, and a recent one. He’d been handling departures badly, he says, because lawyers tell you to keep quiet to protect against litigation, which made separations cold, as if the years and the relationship had never happened. He decided to stop following that advice when it comes to letting people go, to stay open and supportive even at some risk, because closing people off carries its own risk to relationships and reputation. He tied it to Maya Angelou’s line that people remember how you made them feel, with a twist he found surprising: how people remember you comes down to the best thing you ever did for them and the last thing you did for them. In a separation, that last thing is half the memory. So Nick’s organization implemented a program to help departing employees find their next position. Don’t just tell someone you feel bad. Prove, to whatever extent you can, that you care what happens to them after they leave.

Silence is a leadership tool. Sterling’s practice is running, something he never liked and took up for health and stress reasons coming out of college. He runs without headphones, without a watch, without a phone, on purpose, because the mesh of body and mind in that quiet is where he works through the hard problems. Nick has been reaching for silence too, recently, driving without the radio, sitting in front of a fish tank for twenty minutes doing nothing, as an antidote to constant overstimulation. Both framed it as intentional. You can’t escape to a wellness retreat every week, but you can take twenty or thirty minutes to be quiet with yourself, empty the mental trash, and reset. Sterling added a nice observation: the way you find a lost object once you stop looking for it, answers tend to surface from the edges of your mind once you stop straining for them and give yourself the quiet to let them come.

One thing to bring back to your team

Pick the conversation you’ve been avoiding, and have it this week.

Sterling’s closing point is that the longer you put off the tough conversation, the faster things get misunderstood. Most of what makes those conversations feel impossible is the absence of two things: facts about why the conversation is happening, and the willingness to actually listen once it does. Bring the first. Practice the second.

And ask yourself the harder question underneath the whole episode. When someone leaves your team, whether by their choice or yours, what’s the last thing you do for them? Because that, as much as anything, is what they’ll carry.

For leaders looking to take this further, Nick Capman’s book The Passionate Workforce: How to Create and Maintain Maximum Employee Engagement expands on creating high-trust cultures where ideas and people can thrive.

Drawing from practical strategies, real-world examples, and proven methods, Nick provides actionable insights for leaders at every level. Whether you’re refining your leadership style, implementing engagement initiatives, or preparing your team for the future, this Amazon #1 bestseller equips you with the tools to transform your organization from the inside out.

If you haven’t picked up a copy yet, it’s an invaluable resource for any leader serious about maximizing employee engagement and creating a workplace where passion and performance are inextricably linked. Grab your copy on Amazon.

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates.

Need expert help planning and executing compliance projects or connecting with a quality and compliance professional? Get in touch with us. If we haven’t worked together yet, watch our explainer video below or head to our company introduction page to see if we could help you execute your projects now or in the future.

These RA/QA news rolls are available in full only to paid subscribers. Upgrade here. If you haven’t already taken our reader survey, it should only take 1 minute here.

This is our biweekly recap of news and industry happenings curated for our paid subscribers. Our goal is to bring you the headline news and the stories, research, and insights that might be harder to find.

If you haven’t upgraded to a paid subscription yet, you can do so here. Also, make sure to follow us on LinkedIn and catch the latest episodes of our podcast, The Life Science Rundown.

Here are some key items from the past couple of weeks.

💊 = Pharma, biologic, and biotech.
💉 = Medical device, medtech, diagnostics, and digital health.

FDA Leadership and Administration

• Diamantas is not interested in keeping the top job long term: Acting Commissioner Kyle Diamantas does not want to hold the role permanently, according to reporting from Endpoints News. Because the FDA commissioner is a Senate-confirmed position, his acting tenure is governed by the Vacancies Act of 1998, which generally limits an acting official to 210 days from the date the vacancy occurred, putting his window at around early December. That clock pauses once a formal nomination is submitted and resets if a nomination is rejected, withdrawn, or returned. Diamantas previously served as deputy commissioner for food, a role now held by Donald Prater in an acting capacity.

• Sharpless and Pops reported on the commissioner short list: The administration appears to be working from a short list for the permanent commissioner job. CNN reported that former acting commissioner and National Cancer Institute director Norman “Ned” Sharpless and Alkermes chairman and CEO Richard Pops are both under consideration, with some support also voiced for keeping Diamantas. An announcement on a formal nominee is reportedly still weeks away. Whoever takes the role will inherit an agency where many senior positions, including the directors of CDER and CBER, are filled on an acting basis.

• Diamantas keeps a low public profile but an active stakeholder schedule: Since taking over, Diamantas has stuck to official channels rather than the frequent podcast appearances his predecessor favored. Public calendars show him meeting with BIO, the Medical Device Manufacturers Association, Friends of Cancer Research, former Commissioner Mark McClellan, AdvaMed, PhRMA, and legislators who oversee the FDA. He also met with PhRMA head Stephen Ubl and Right to Try bill sponsor Rep. Diana Harshbarger (R-Tenn.), joined by policy deputy Grace Graham.

• Diamantas pledges support for staff and more advisory committee meetings: At a June 3 meeting with rare disease advocacy groups, Diamantas voiced support for FDA reviewers and pledged to prevent interference with their decision-making, along with a renewed push to hire reviewers, including former agency employees. According to reporting from Endpoints News, he also indicated that more advisory committee meetings should be expected, reversing a slowdown during Makary’s tenure. The hiring pledge comes as the administration pursues a workforce policy that would make it easier to remove staff seen as out of step with its priorities.

• Lisa Romano named inspections and investigations lead: The FDA promoted Lisa Romano to principal deputy associate commissioner for inspections and investigations. She previously served as acting in that role and before that as deputy director for the Medical Products Inspectorate.

• White House counsel joined an FDA meeting with Replimune: Public calendar logs show that on May 27, two days before Replimune said it would resubmit its rejected melanoma drug RP1, the company’s CEO and head of regulatory affairs met with Deputy Commissioner for Strategic Initiatives Lowell Zeta and Chief Counsel Sean Keveney. White House associate counsel Joshua Hoyt also attended. The FDA’s second rejection of RP1 in mid-April drew significant pushback and contributed to the tensions that preceded Makary’s resignation. Replimune had separately met with the White House Domestic Policy Council in May to present its case.

• Kennedy reportedly out of the loop on Høeg firing: A New York Times story describing HHS Secretary Robert F. Kennedy Jr.’s limited involvement in day-to-day management of his agencies cited the recent FDA leadership shake-up as an example. According to three people who spoke with the paper, Kennedy did not learn of Tracy Beth Høeg’s firing as acting CDER director until afterward.

• Kennedy uses his podcast to press on peptides: A June 10 episode of Kennedy’s podcast featured a biohacking influencer who described personally using several peptides that the FDA recently removed from the 503A bulks nomination list. Those peptides are on the agenda for upcoming Pharmacy Compounding Advisory Committee meetings, one in July and another by the end of February. Their removal was unusual given that the agency had previously cited safety concerns. Kennedy has continued to push the FDA to loosen restrictions on compounded peptides.

• New executive order reshapes the federal workforce: President Trump signed an executive order converting some policymaking roles into “Schedule Policy/Career” positions, removing certain protections such as adverse action procedures and appeals and allowing for faster termination. The policy was known as “Schedule F” during the first Trump administration and could affect FDA staff.

FDA General

• FDA finally publishes delayed gene therapy guidance: 💊 The FDA released “Leveraging Prior Knowledge in the Development of Human Gene Therapy Products Incorporating Genome Editing,” a guidance the White House had cleared back on March 31. The unusual delay had stood out for an otherwise uncontroversial document. The guidance explains how companies can use existing knowledge of a manufacturing platform or other public knowledge to advance development and improve review efficiency across multiple programs.

• Revised guidance on company communications with payors: 💊💉 The FDA issued a revised version of its 2018 draft guidance on communicating health care economic information to payors and formulary committees. The update reconfirms that companies can share such information with these sophisticated audiences in ways that would be problematic if directed at doctors or patients, including for products not yet approved or cleared.

• FDA adopts ICH M15 on model-informed drug development: 💊 The FDA formally adopted the International Council for Harmonisation’s M15 guideline on model-informed drug development, which the ICH adopted February 9.

• New draft guidance on Forms FDA 3542 and 3542a: 💊 The FDA published a draft guidance answering 29 common questions about the forms used to submit patent information with a New Drug Application or after approval. The information feeds the FDA’s Orange Book listings, which carries significant implications for sponsors and generic competitors. The agency framed the release as part of its Drug Competition Action Plan.

• ICH biannual assembly readout: 💊 The ICH Assembly held its biannual meeting and reported additions to its observer roster, approval of an ICH Strategic Vision document outlining priorities through 2030, and adoption of the ICH E6(R3) Annex 2. The readout noted progress on high-priority quality, efficacy, safety, and multidisciplinary guidelines, and said the MedDRA Version 29.0 Companion Document is expected later this year.

• Two new CBER SOPP documents: 💊 CBER released Version 7 of SOPP 8401.1 on issuing and reviewing responses to information request communications for pending applications, and Version 11 of SOPP 8406 on processing PDUFA application payments.

• FDA reorganizes and expands its real-world evidence pages: 💊💉 The FDA revamped its webpage on the use of real-world evidence in decision-making, the first update since the pages launched in September. The CDER page now lists 13 approvals using RWE and nearly 50 postmarket studies, and the device center’s newly incorporated page lists 138 device-specific regulatory uses of RWE. Formats and definitions were standardized across centers to allow some cross-comparison.

• First in silico drug development tool accepted into ISTAND: 💊 The FDA accepted a letter of intent from Absentia Labs for a digital liver model that uses artificial intelligence to assess the liver injury risk of small-molecule candidates, the first in silico tool advanced into the agency’s ISTAND program. Acceptance is the first of three steps toward qualification, and the FDA raised 16 considerations for the developer to address. Read the FDA’s statement on it here.

• FDA considering an expedited IND pilot: 💊 The White House is reviewing an FDA request for information on an “Expedited Investigational New Drug Pilot Program,” according to a notice from the Office of Information and Regulatory Affairs. The effort responds to concerns that early-stage trial activity is shifting to countries such as China and Australia. Lawmakers and the agency have proposed testing whether certain lower-risk early trials could launch with less oversight, with Australia’s clinical trial notification system cited as one possible model.
  

• FDA reassesses the CNPV program after a public hearing: 💊 The FDA held a June 4 public hearing on the Commissioner’s National Priority Voucher program, with 17 drug company representatives, physicians, researchers, and patient advocates offering feedback. Deputy Chief Medical Officer Mallika Mundkur and acting CDER Director Mike Davis opened the session. Several speakers, including PhRMA’s Kelly Falconer Goldberg and former FDA official Lowell Schiller, urged the agency to add transparency and structure through guidance or notice-and-comment rulemaking, warning that informal selection criteria could invite legal challenges. The agency intends to continue the program but may change its processes, and is collecting comments through June 29.

• FDA adds bemotrizinol to the sunscreen monograph: 💊 The FDA finalized an order adding bemotrizinol to its sunscreen monograph, the first new sunscreen active ingredient cleared for the U.S. market in roughly two decades. The ingredient, submitted by DSM-Firmenich under the CARES Act pathway, will be considered generally recognized as safe and effective at concentrations up to 6%. The ingredient has been authorized in Europe since 2000.

• OTP town hall covers cell and gene therapy submissions: 💊 At a June 4 town hall, the FDA’s Office of Therapeutic Products fielded questions on cell and gene therapy BLA submissions, emphasizing the pre-BLA meeting. Staff described it as not required but highly recommended, advised requesting it at least four months before the anticipated BLA submission, and recommended a maximum of 10 questions for a 60-minute meeting.

Read more

Last week, we shared our model for interrogating “human error” in CAPAs. Today, we have a conversation about why, despite meeting metrics, hitting closure dates, and getting clean reports, problems addressed through CAPAs often come back.

That gap between a system that’s technically compliant and one that actually drives improvement is where many QA teams quietly lose ground. A closed CAPA that fails again for the same reason was never a successful CAPA. The numbers say you’re fine while the repeat offenders say otherwise.

Our Nick Capman recently sat down with Dan Eagles, Vice President of Quality Assurance at NeuraSignal, to talk about why CAPA systems become a check-the-box exercise and what it takes to rebuild them into something that strengthens compliance, efficiency, and team ownership.

Dan has spent over 20 years in the medical device and pharmaceutical industries. He got into medical device work before he even graduated college, held roles in manufacturing and R&D, and has spent the past 23 years in quality assurance at increasing levels of responsibility. Before NeuraSignal, he held quality leadership roles at NuVasive, Dendreon, Endologix, and Edwards Lifesciences. He’s a CAPA and NCR mentor, an auditor, and he’s listed on 21 ophthalmic device patents. At NeuraSignal, he leads the quality program for a company developing robotic transcranial Doppler ultrasound technology for stroke detection.

Apple Podcasts | Spotify | YouTube | Web + Others

Dan's key insights and practical takeaways

If you're short on time, here are the most important lessons from the discussion.

CAPA failure comes down to two things: process and stewardship. And they’re entwined. The process can be sound, but if nobody owns it well, it breaks. The stewardship can be strong, but if the process is clunky, people work around it. Dan’s framing applies beyond CAPA to the rest of the quality management system, but it shows up in CAPA more visibly than anywhere else.

The clearest warning sign of a broken-but-compliant system is repeat offenders. Your metrics can be timely. People can be hitting their marks. But if you’re seeing the same failure recur after an executed CAPA, the CAPA didn’t work, no matter what the closure data says. Metrics matter, and the one that matters most isn’t whether you closed on time. It’s whether the problem stayed solved.

The process is simple but not easy, and most of the trouble starts at the very beginning. A CAPA is straightforward: you have an issue, you write a problem statement, you investigate, you build an implementation plan, and you build a verification of effectiveness plan. The failure point is usually the problem statement. A poorly worded problem description creates ambiguity that follows you through the entire investigation. Dan’s recommendation: build training into your CAPA program specifically to teach owners how to write a good problem statement. And apply it to NCRs too, since NCRs often feed into CAPA. Train your owners to write strong problem statements and you get a lot of value for the effort.

The most common root cause mistake is back-engineering the answer you already wanted. Humans jump to conclusions. We see an obvious gap, decide that’s the cause, and then quietly reverse-engineer the root cause analysis to justify the corrective action we’d already picked. Dan illustrated this with a Six Sigma training example from his time at J&J:

Stop defaulting to “retrain the operator.” It’s the favorite CAPA for almost any finding, and it usually isn’t the answer. Dan’s approach is to handle training inside the NCR first and track it. If the same issue keeps recurring (he gave the example of cable crimping), then you’ve built up compelling data, and you can escalate to a CAR, which serves as a precursor to a CAPA. By that point you have evidence that something deeper is going on, rather than reflexively writing “retrained the operator” and moving on.

Trade retrospective waiting for proactive validation where you can. A lot of organizations have drifted toward a “sit on it for three months and watch for recurrence” model for verifying effectiveness. That’s retrospective by design. In many cases, you could instead run a process validation that proves capability directly. Demonstrate that the operators understand the training, that they can do the task correctly, and that the process holds. When you wrap up that validation, you’ve wrapped up your VOE with objective evidence that the CAPA actually worked, rather than just waiting to see if the problem reappears.

Know the difference between a correction and a corrective action. Human nature pushes us to fix the thing in front of us, and in medical device work the instinct to protect patient safety is strong. But a correction just fixes the immediate problem, like filling in the missing fields on a record. A corrective action prevents recurrence. When you assign a corrective action, make sure it’s actually going to stop the problem from happening again. And keep risk management in mind throughout, because CAPA and risk management work hand in hand. The goal is to keep your risks as low as possible.

Good CAPA programs have leadership buy-in across every functional group. Dan reframed the leadership question around what good looks like. Strong programs have leaders who understand that CAPA is built into medical device DNA, that it isn’t just a compliance exercise but a project that drives continuous improvement. Shift the mindset from “I need to check this box for an auditor” to “I can finally fix the problems that have been aggravating us,” and CAPA becomes better operationally, better for compliance, and better for keeping manufacturing lines running.

Without that buy-in, CAPA becomes a hot potato. You get people who position themselves as CAPA experts mainly to argue that something isn’t a CAPA, though if you press them to define a CAPA, they often can’t give you a solid answer. The underlying attitude is “my department is already stretched, we don’t need the extra work.” But if you’ve ever been through a recall, you know that takes far more time and money. Dan describes CAPA as a bridge between your product before it leaves the dock and your product out in the field. Kept under your umbrella at headquarters, it’s a powerful tool for stopping discrepant product from ever reaching the market.

Rebuilding a CAPA system is a project, so treat it like one. Put a CAPA leader in charge, someone who genuinely understands CAPA, and pull in team members from every cross-functional group. Start with CAPA training: what it is, what the benefits are, what the downsides are, what it means to FDA and ISO. Look at warning letters and you’ll see the same heavy hitters every time, CAPA, training, and complaint management, so you know where regulators will focus. Then keep the process itself simple.

Standardize your forms so nobody is inventing anything on the fly. Whoever owns the CAPA should be working from the same templates. Build a problem statement template that captures who, what, why, and how. Provide a template for the five whys. Provide one for the fishbone diagram, since most people can’t recall all the categories from memory anyway (man, materials, process, environment, and the rest). If you can pull a released document from your QMS with the fishbone already laid out, you don’t have to remember it. That consistency drives quality across the whole process.

Build mentorship into every phase. Train experienced people to mentor CAPA owners at each step: reviewing the investigation, checking the implementation plan, pressure-testing the VOE plan, and tightening up anything that looks loose. Dan recalled a colleague who had a genuine knack for problem statements, whose CAPAs were brief, whose investigations weren’t long-winded, and whose implementation plans were succinct. As your CAPA owners grow, develop them into leaders, into technical experts, and into mentors who can pass that knowledge to their teams and delegate. That’s how you spread CAPA literacy across the entire company culture.

Track effectiveness, not just closure. Closure and on-time rates are useful high-level signals that tell you whether you’re resource-constrained or task-saturated. But the most important metric is whether your VOEs are actually effective. After a CAPA closes, keep tracking those specific errors to see whether they resurface, and stay objective about whether a new nonconformance is genuinely the same issue. Dan pointed to a medical device white paper involving a torque problem: the engineer assumed the fix was a torque standard on the floor and verified torque wrenches. Standards went in, wrenches were checked, parts were torqued correctly, and the problem persisted. The real answer was upstream in the manufacturing process, in finding a different way to capture the fastener rather than relying on torque alone.

Use AI as a content tool, not an oracle. Most teams are running CAPA in an electronic system, and Dan’s first piece of advice is to make that system easy to work in. If it’s clunky, change it, because that’s what your change management program is for. A clean electronic system helps the owner, helps the approver, and earns its keep at audit time, since you can print a CAPA that an auditor can follow easily. When an auditor pulls a few CAPAs and they all share the same look and feel, that signals your company takes CAPA seriously. Dan uses AI personally to help draft CAPAs and work through investigations, often feeding in his quality procedures to check what he might be missing. He reads the output and doesn’t take it as gospel. As he put it, if the output tells him to install a muffler bearing, he’s not going to abide by that. The value is in generating content quickly so he can pull the pieces he needs, since the typing itself eats up time even when you already know what you want to say. Down the line, AI could also analyze company data to assess whether a CAPA system is effective, which would add real value for continuous improvement.

Three moves to rebuild a CAPA system from scratch.

1. First, understand what your current system is actually doing. Gather your team and plot out the workflow the old-school way, noting where things get stuck. If your performance metrics show a pile-up in investigation or implementation, the workflow map will usually show you why.

2. Second, design your blue-sky system and keep it simple. There’s more material available on CAPA than ever, and FDA and ISO lay out exactly what you need, so use the regulations as your baseline and scale the system to your company, whether that’s 30 people or 3,000.

3. Third, put in an electronic system, sandbox it, and validate it. The activity is the same as it was on paper, so make sure the system genuinely works before you rely on it.

One thing to bring back to your team

Pull your last several closed CAPAs and ask one question: have any of these problems come back?

If they have, your closure metrics are telling you a story that isn’t true. Then look at where those CAPAs started:

• Was the problem statement clear enough that someone outside the team could understand exactly what went wrong, or was it vague from the first sentence?

• Did the corrective action actually prevent recurrence, or was it a correction dressed up as one, plus a “retrain the operator” line?

• Did anyone back-engineer the root cause to match a fix they’d already decided on?

• Did you verify effectiveness by waiting three months and hoping, or did you prove capability with a validation?

Dan Eagles is Vice President of Quality Assurance at NeuraSignal, where he leads the quality program for a company developing robotic transcranial Doppler ultrasound technology for stroke detection. He has over 20 years of experience in the medical device and pharmaceutical industries, with roles spanning manufacturing, R&D, and the past 23 years in quality assurance. Before NeuraSignal, he held quality leadership positions at NuVasive (as Site Quality Lead and Management Representative), Dendreon, Endologix (as Head of Quality Control), and Edwards Lifesciences. His expertise covers FDA Quality Management System regulations, ISO 13485, MDR implementation, quality system remediation, audit management, risk management, and CAPA/NCR processes. He holds an MBA from UC Irvine’s Paul Merage School of Business and a BS in Mechanical Engineering from Cal Poly Pomona, and is listed on 21 ophthalmic device patents.

Connect with Dan on LinkedIn here.

Who is The FDA Group?

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on staff augmentation, auditing, remediation, QMS, and other specialized project work in Quality Assurance, Regulatory Affairs, and Clinical Operations.

With over 3,750 resources worldwide, over 325 of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee.

Here’s why 17 of the top 20 life science firms access their consulting and contractor talent through us:

• Resources in 75 countries and 48 states.

• 26 hours average time to present a consultant or candidate.

• Exclusive life science focus and expertise.

• Dedicated account management team.

• Right resource, first time (95% success).

• 97% client satisfaction rating.

Talk to us when you're ready for a better talent resourcing experience and the peace of mind that comes with a partner whose commitment to quality and integrity reflects your own.

Subscribe to The Life Science Rundown:

Apple | Spotify | YouTube | Web + Others

Check out our newly launched AI-powered QMS audit tool, AICA (the Audit Intelligence Compliance Assistant).

Dan discusses why the problem statement is where most CAPAs go wrong, how teams back-engineer root cause analysis to match a fix they've already chosen, and why "retrain the operator" is usually the wrong answer.

A few of Dan's key takeaways:

• A closed CAPA that fails again for the same reason was never successful, repeat offenders are the clearest sign of a broken-but-compliant system

• Most CAPA trouble starts with a poorly worded problem statement, train your owners to write good ones and you get value across NCRs too

• Humans back-engineer root cause analysis to justify the fix they already wanted, the J&J forklift story shows how that goes wrong

• Trade retrospective "wait three months and see" verification for process validation that proves capability with objective evidence

• Know the difference between a correction (fixes the immediate problem) and a corrective action (prevents recurrence)

• Without leadership buy-in across functional groups, CAPA becomes a hot potato, but a recall costs far more time and money

• Track whether your VOEs are effective, not just whether CAPAs closed on time

About Dan Eagles
Dan Eagles is Vice President of Quality Assurance at NeuraSignal, a company developing robotic transcranial Doppler ultrasound technology for stroke detection. He has over 20 years in the medical device and pharmaceutical industries, with 23 years in quality assurance and prior roles at NuVasive, Dendreon, Endologix, and Edwards Lifesciences. He is a CAPA and NCR mentor and is listed on 21 ophthalmic device patents.

About The FDA Group
The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle — from clinical development to commercialization — with a focus on staff augmentation, auditing, remediation, QMS, and other specialized project work in Quality Assurance, Regulatory Affairs, and Clinical Operations. Learn more: https://www.thefdagroup.com/

Few topics in quality draw as many strong opinions as how to run a corrective and preventive action investigation.

People’s perspectives often diverge on just about every dimension of it: tools, about how deep is deep enough, and about when a root cause is really the root cause. We think that range of views is actually healthy as it challenges models and techniques in an area where no single method genuinely fits every situation.

We’re okay wading into that debate space from time to time, given just how many CAPA programs we encounter (and build or optimize) in the course of our fieldwork.

Here, we want to cover one part of the work that trips up a lot of investigations: deciding whether a problem that looks like human error actually is one. The approach below is not the only way to do this, but it has worked well for the Quality teams we work alongside and for us, and it holds up under inspection. So let’s break it down in case it’s a model you want to adopt, in full or in part.

First, a little context.

Human error is cited far more often than it happens

Genuine human errors are real, and in FDA-regulated manufacturing and adjacent functions, they do occur. But we see them get written into investigations far more often than they should, and that gap causes problems when it (intentionally or not) papers over what’s actually a systems problem that looked like a human error.

Most issues that first look like human error, especially ones that recur, trace back to a process or system that will keep producing the same problem until someone changes it. Labeling the cause “human error” puts a band-aid over that system and leaves it intact. The next occurrence is already on its way.

There’s an obvious inspection risk here, too. When an investigator sees human error cited more often than it could plausibly happen, it reads as a sign that problems aren’t being investigated thoroughly. That can move the investigator into problem-hunting mode and open the wider quality system to more scrutiny. We see this quite often in our audit and mock inspection work.

The exception is the one-time error made by a well-trained person following a sound process, where a passing distraction led to a mistake. That classification can be justified, but only after a thorough investigation has looked everywhere else and found nothing. It’s a conclusion you earn by exhausting the alternatives, not a place to start.

A model for telling genuine human error apart

It helps to have a structured way to test whether an apparent human error was deliberate or inadvertent, and if inadvertent, what kind of error it was.

The Skills, Rules, Knowledge framework, paired with the Generic Error Modeling System, does this well. It’s simple to follow, and it pushes you past the easy label toward the place a real problem is likely to sit. Just about every team can use this.

Let’s break it down:

The first question is whether the action was deliberate or inadvertent. An inadvertent action is the only kind that counts as a genuine human error in this model. From there, inadvertent errors fall into three types: skill-based, rule-based, and knowledge-based.

Skill-based errors are further split into slips and lapses. Both point to the same root cause: a lack of attention, showing up either as momentary memory loss or as a routine action that wasn’t performed. These are the closest things to a true human error, and they should be pretty rare.

Rule-based errors are more technical, such as applying the wrong rounding rule, resulting in an out-of-specification result. Before faulting someone’s judgment, ask whether they were set up to perform the task correctly:

• Did training give them enough practice?

• Did that practice match the operation they actually perform on the floor?

• Does the procedure spell out the details needed to do the task correctly?

Knowledge-based errors tend to show up when someone is multitasking beyond their capacity or can’t give a task the focus it needs. The useful question is why they were in that position at all:

• Were they asked to do too many things at once?

• Is the department under-resourced?

• Did the supervisor know the person was stretched past their limit?

In the rule-based and knowledge-based cases, notice that the questions point away from the individual and toward training, procedures, workload, and oversight. That’s the point of the exercise!

Two places systemic “human” error usually hides

When an error turns out to be less human than it first appeared, the next question is about how the work actually gets done. Two areas are worth a hard look.

Read more

We’d like to quickly introduce Eric Boyd, who’s leading our AICA client team, and share his brand new demo of the latest version of the tool.

Since launching AICA earlier this year, we’ve refined its user interface and continued to sharpen its analysis capabilities. If you’re a newer subscriber and this is the first time hearing about the tool, we built it to help Quality teams move beyond limited document sampling and toward a more complete, system-wide view of compliance by marrying the latest in machine learning with a ruleset developed by 15+ regulatory compliance experts.

Instead of manually reviewing one procedure at a time, you can upload your documentation and run an assessment against applicable FDA requirements. AICA then reviews the written system, identifies potential gaps, ties those gaps back to specific regulatory expectations and source documents, and helps generate a structured report that the Quality team can use for follow-up.

As Eric explains in the demo, the tool is designed to support teams by:

• Auditing QMS documentation against FDA regulations

• Highlighting gaps in procedures and controls

• Suggesting steps for remediation

• Supporting targeted or broader CFR-based assessments

• Maintaining a human subject matter expert in the review loop

• Producing transparent, traceable outputs that can be used as a working gap assessment tool

Head over to aica.thefdagroup.com and fill out our demo form to let us know you’re interested in seeing it in person. Or email Eric directly and let him know you’d like to set something up: eboyd@thefdagroup.com.

A simple workflow anyone can run

We’ve simplified AICA’s workflow into just a few steps that even the less tech-minded among us can manage very easily. In the demo, Eric shows AICA’s user interface and how a user can define the audit scope by answering a few short questions.

For example, the user can identify which procedures are performed in-house, which are outsourced, and which are not applicable. From there, the assessment can be configured either broadly across an entire CFR part or more narrowly within a specific subpart.

Once the scope is defined, the user uploads the relevant documentation. AICA then begins its initial review and notifies the user when observations are ready.

This is where the system is intended to save you a ton of time. You don’t need to remain in the platform while the initial assessment runs. AICA performs the first-pass analysis and alerts the team when it is time to review the observations.

Keeping the human expert in the loop

One of the most important points in the demo is that AICA is not presented as a replacement for judgment.

Eric shows where a human subject matter expert reviews and validates the observations produced by the tool. Each observation includes an explanation of the gap identified, a suggested remediation step, and references to the relevant documents and CFR sections. The reviewer can approve an observation, reject it, or add additional context. Only after the observations have been reviewed does the user move on to the final report.

That review step is critical. It allows teams to combine AI-enabled speed with human quality oversight, helping make sure that the outputs are not only fast but also meaningful, defensible, and grounded in the realities of the organization’s quality system.

From observations to an actionable report

After the review process, AICA generates a downloadable report.

Eric walks through several key components of the report, including a heat map that shows where the highest concentration of gaps exists within the written system. In the demo, the assessment is being run against 21 CFR Part 211, allowing the team to see where gaps cluster across the relevant subparts.

Users can drill into individual heat map cells to view observation-by-observation detail, including the documents reviewed, the gaps identified, and how those gaps relate to the broader health of the QMS.

This is especially useful for determining whether an issue appears isolated within a single procedure or systemic across the quality system.

The audit observations tab aggregates the findings by CFR subpart and includes the reviewed documents, observations, remediation suggestions, and reference materials. The result is an output designed around transparency and traceability.

A living gap assessment tool

Because the final output is delivered in Excel, teams can continue working with the report after the initial assessment is complete.

Users can add columns for owners, due dates, status, or other internal tracking needs. That means the report can actually function as a living gap assessment tool that supports remediation planning, accountability, and follow-through.

For Quality teams managing multiple priorities, this is an important distinction and something our early beta users asked for. The value of an audit is not only in identifying gaps, but in turning those findings into practical, trackable action.

Where teams are using AICA now

There are several use cases where organizations are seeing success with AICA, including:

• Internal audits

• Supplier audits

• Pre-inspection preparation

• Year-round compliance monitoring

Each of these use cases depends on the same underlying need: a faster, more complete, and more traceable way to evaluate the written quality system against regulatory expectations.

For teams preparing for inspection, strengthening supplier oversight, or trying to maintain a more proactive compliance posture throughout the year, AICA offers a new way to approach document-heavy quality review.

Have questions or want to get a demo?

Want to see whether AICA fits your team? Visit aica.thefdagroup.com to learn more and get in touch. We’re onboarding teams to the tool right now.

A few weeks back, we shared the insights from our Q1 audit reports across sterile pharma, biologics, contract labs, GLP oversight, medical devices, and computer systems assurance. Fifty findings, zero critical, and most of them were the routine stuff that fills any audit report.

In case you missed it and you’re a paid subscriber:

We wanted to return to that dataset to extract some higher-level lessons applicable to every team out there. A few patterns we saw in that dataset stuck with us after the reports were signed because they pointed to something deeper than the findings themselves.

Talk to us if you need auditing, mock inspection, remediation, or other RA/QA/Clinical support.

1. The system of record can be perfect, and you can still fail the audit

This was sort of the lesson of the quarter, and it showed up just about everywhere.

At a well-run biotech that brought us in, the validation lifecycle management platform was mature, with risk-based thinking very well baked into the computer systems assurance program. But the PDF reports it generated for system periodic reviews came out with the data tables cut off. The reports were built in portrait orientation, and the tables were in landscape orientation. Several reports in a row were all truncated. The data was complete in the system. But the documentation an investigator would actually read was not.

We saw versions of this at other sites, too.

• Weighing slips without the balance ID, so the data couldn’t be traced to a calibrated instrument (even though the instrument was calibrated).

• Raw spreadsheets were handed over as audit evidence with no review or approval.

• An SOP was pointing analysts to a training system that had been replaced a year earlier.

None of these were data problems. It’s the gap between “our data is correct” and “our data is inspection-ready,” and that gap is silently wider than most facilities think. Internal users know the workarounds. They know which report to ignore, which export is the real one, which field is always blank for a reason. An external reviewer, of course, knows none of that, and the gap only becomes visible when someone outside the building asks for evidence. By then, it’s a finding.

2. A finding you didn’t fix is often worse than the finding you first had

At a clinical packaging facility, complaint handling cycle time had crept past the 30-day procedural limit. On its own, that’s a minor finding. But the same issue had been flagged at the previous audit two years earlier, and it hadn’t been corrected. Cycle time had actually gotten slightly worse, from about 38 days to 40. So it got written up as a major.

Instead of being about cycle time, the escalation was about what an unresolved repeat finding signals: that the CAPA from the last audit either wasn’t implemented or wasn’t checked for effectiveness. A repeat finding tells our auditors that root cause analysis stopped at the symptom, or that someone closed the CAPA on paper without verifying it worked, or that management review isn’t allocating resources when corrections slip.

This is worth thinking through because it changes how you should treat prior findings. A closed CAPA that didn’t actually fix the problem isn’t neutral. It’s a compounding liability, and the next auditor who finds the same issue shouldn’t grade it the way the first did.

3. Sometimes, nobody owns the supply chain gap until someone asks who does

The most serious single finding from last quarter came from a question that took about thirty seconds to ask.

At a sterile drug manufacturer, the primary packaging containers were sterilized by a contract sterilizer. We asked to see the sterilization assurance. The component supplier’s Certificate of Analysis didn’t mention sterility. The contract sterilizer’s documentation didn’t either, and the drug manufacturer had never audited the sterilizer.

There are three parties in the chain there, each one assuming, apparently, that sterility assurance was somebody else’s job. Under 21 CFR 211, the finished product manufacturer holds ultimate responsibility for the quality of its inputs, and under the FDA’s contract sterilization guidance, the manufacturer and sterilizer share responsibility for the sterilization itself. So the answer to “who owns this?” was “the manufacturer, and they hadn’t been doing it.”

Read more

Some important congressional news for the drug developers among us. As reported in Endpoints News, the House Appropriations Committee has directed the FDA to rework how it clears new clinical trials. It’s part of a wider effort in Washington to keep early drug development from moving offshore.

The instruction appears in the committee’s report accompanying the fiscal year 2027 Agriculture and FDA funding bill, which the House just passed by a vote of 213 to 210.

In the report, the committee directs the agency to revise its investigational new drug (IND) processes and data requirements for initial human trials to:

• Streamline administrative requirements.

• Reduce filing burdens.

• Make them phase- and risk-appropriate.

The committee said it recognizes that many Phase 1 trials are better understood as a continuation of early proof-of-concept research than as part of a traditional commercial drug development program, language that points toward lighter requirements at the very start of testing.

The committee also encouraged the FDA to build a pilot modeled on Australia’s clinical trial notification system. Under that approach, lower-risk INDs would be reviewed by academic medical center institutional review boards, with agency or third-party support to supplement that expertise rather than a full FDA sign-off before a trial begins.

This lines up with the FDA’s own request

The direction from House appropriators tracks closely with what the agency asked for in its fiscal year 2027 budget. That request proposed an optional, risk-based “expedited IND” pathway for experimental drugs whose existing preclinical data can be confirmed without animal testing.

The proposal is now under review at the White House Office of Information and Regulatory Affairs.

Much of the early momentum is reported to have come from former FDA Commissioner Marty Makary, who pushed for institutional review board reform before stepping down. Speaking at the 2026 CMS Quality Conference in Baltimore alongside CMS Administrator Mehmet Oz and NIH Director Jay Bhattacharya, Makary said China starts about four times as many Phase 1 trials as the United States.

He put the average pre-IND timeline at roughly 380 days in the US against about 220 in China, and said China had announced plans to compress that to 60. He argued that a research system depending on a monthly volunteer committee of staff, the way IRB review often works, can’t keep that pace.

(Makary resigned on May 12 amid White House pressure following a dispute over flavored e-cigarette approvals. Kyle Diamantas, the agency’s top food official, is serving as acting commissioner while the administration looks for a permanent replacement, which leaves the future of these reforms partly in the hands of leadership that has not yet been named.)

Reform interest looks bipartisan

The appropriations directive is not the only push on early-phase trials.

• Rep. Jake Auchincloss (D-MA) has advanced a bipartisan proposal that would expand the FDA’s role in clinical trials embedded in routine clinical care.

• Sen. Bill Cassidy (R-LA) earlier released a report outlining IND-related problems at the agency as part of a broader argument for FDA modernization.

The money and what’s probably coming next

The underlying bill would give the FDA about $7.1 billion for fiscal 2027, roughly $200 million more than the current year and somewhat below the $7.2 billion the administration requested. It cleared the House 213 to 210, with five Republicans voting against and four Democrats voting in favor.

The Senate has not taken up its version. The Senate Appropriations Committee postponed its markup to a date it has not announced, so the IND language is not final and could change before any bill reaches the president.

For sponsors planning first-in-human work, we see three questions that are still open:

• Whether the expedited IND pilot clears White House review

• Whether the Senate keeps the House’s IND language

• How a permanent commissioner approaches early-phase reform once one is in place.

None of that is settled yet, but the direction of travel in both the budget request and the appropriations report is hard to miss. We’ll keep you posted when there’s more to report.

Talk to us if you need IND submission or related regulatory support. We have a deep bench of experienced IND experts and routinely help with all dimensions of IND strategy, preparation, and submission.

Who is The FDA Group?

The FDA Group helps life science organizations rapidly access the industry's best consultants, contractors, and candidates. Our resources assist in every stage of the product lifecycle, from clinical development to commercialization, with a focus on Quality Assurance, Regulatory Affairs, and Clinical Operations.

With thousands of resources worldwide, hundreds of whom are former FDA, we meet your precise resourcing needs through a fast, convenient talent selection process supported by a Total Quality Guarantee. Learn more and schedule a call with us to see if we’re a fit to help you access specialized professionals and execute your projects on time and on budget.

Share The FDA Group's Insider Newsletter

Inspection readiness has always been a moving target. But for many drug and device manufacturers, the target feels a little less predictable right now.

We’re seeing (and hearing from teams we help with auditing and readiness) that inspection timing has sometimes become less routine than it was a few years ago. There also seems to be a more “varied” group of investigators running inspections. On top of that, we’re seeing relatively minor 483 observations that might previously have been handled as discussion items addressed while an investigator was on-site.

More uncertainty and inconsistency here mean that inspection readiness has to be built into your quality system, which operates every day, rather than being a discrete exercise it sometimes becomes (whether intentional or not).

In a recent episode of The Life Science Rundown, we spoke with Jeff Brenneman, Vice President of Global Operations Quality at Alora Pharmaceuticals, about how manufacturers can sustain high-performing teams and robust quality systems in this kind of environment. We wanted to return to it because Jeff packs quite a bit of great advice into it.

Much of that conversation came back to a simple point: when inspections are harder to predict, quality systems need to be easier to understand, defend, and execute. And that posture has to exist at all times.

Here are six practical ways to do that that Jeff touched on. We fold these into our readiness philosophy all the time. Talk to us if you need audit or inspection readiness support.

1. Write investigations for someone outside your organization

One of the most useful inspection-readiness tests is also one of the simplest: Could someone who does not work at your company understand this investigation?

That matters because companies often write documents from inside their own bubble, so to speak. They use internal acronyms and assume the reader understands their product, equipment, process, deviation history, and organizational shorthand. That may work for people who live inside the system every day, but it often does not work for an investigator coming in cold. They shouldn’t have to ask for constant definitions and clarity.

This is especially important for deviations, investigations, risk assessments, and other documents that are likely to become focal points during inspections.

A strong investigation should orient the reader before it gets technical.

• What product was involved?

• What happened?

• When did it happen?

• Where in the process did it occur?

• What equipment, batch, lot, or system was implicated?

• Why did the event matter?

The technical details still need to be there, of course. But they should come after the basic story is clear. If an investigator has to work too hard to understand what happened, they are naturally more likely to keep digging. If the document lays out the event plainly and logically, the discussion starts from a better place, and often ends faster.

2. Make risk-based decisions easier to defend

Risk assessment is often where quality systems move from black-and-white compliance into gray space. That’s not a problem by itself, but the rationale behind a risk-based decision has to be clear enough that someone outside the company can follow it. In that way, it relates to our first tip.

This becomes especially important if the investigator doesn’t have the same background or product-specific experience your team is used to seeing (which is happening more these days). A decision that seems obvious internally may look unsupported externally if the record does not show how you actually got there.

The practical fix is to treat every risk-based decision as something you may need to explain later. That means documenting:

• The question being evaluated

• The applicable requirement or expectation

• The data used to support the decision

• The assumptions made

• The rationale for the path chosen

• Any controls or follow-up actions used to manage the remaining risk

The idea here is to make your reasoning super visible wherever you might be asked about it. If the decision is defensible, the documentation should support it rather than relying on memory or other less tangible sources.

3. Simplify the quality system before complexity becomes the finding

“Robust” is one of those empty adjectives we often hear used to describe a quality system. But in practice, it’s very often a nicer way to say “big and complicated.”

In reality, most genuinely “robust: quality systems we see are simple, efficient, and compliant. Problems actually tend to emerge when the system becomes too complex for people to follow consistently. We see that complexity show up as a few common symptoms:

• Procedures that contradict themselves

• Procedures that conflict with other procedures

• Unclear roles and responsibilities

• Process handoffs that are not defined

• Training assigned too broadly

• Electronic systems governed by procedures that do not match how the system is actually used

These issues create risk because they make failure easier. If following one SOP causes someone to deviate from another, the problem is way worse than user error. It’s a real system design problem.

The same is true when responsibilities are vague. If a procedure says QA is responsible for a step, but doesn’t specify which QA role, when the step occurs, what evidence is required, or who owns the final decision, the organization has left too much to interpretation and is quietly asking for trouble.

When we come into a firm for inspection-readiness support, one thing we do is look for friction points in the system. Where do people hesitate? Where do handoffs break down? Where does the procedure require more effort than the risk justifies? Where are teams creating workarounds because the official process is too hard to use? Those are all places where complexity is already weakening compliance.

4. Be more precise about training

Read more