The First Hour on Site: What It Reveals About the Inspection Ahead

‍An experienced investigator does not use the first hour to decide whether a facility will pass or fail. The first hour is used to decide where the truth is most likely to be found.

After more than three decades of inspecting life sciences operations across different countries, product types, and regulatory environments, I have learned that the trajectory of an inspection often becomes visible remarkably early.

Not the final outcome. No responsible investigator reaches a conclusion before examining the evidence. But within the first hour, experienced investigators begin forming hypotheses about the facility’s state of control, the authority of its quality unit, the reliability of its records, and the organization’s willingness to disclose problems.

FDA’s drug-manufacturing inspection program evaluates whether a facility operates in a state of control and consistently manufactures products of acceptable quality. It examines interconnected systems—not merely isolated documents—including quality, facilities and equipment, materials, production, packaging and labeling, and laboratory controls.  

The effectiveness of the pharmaceutical quality system is central to that assessment . Please refer to - FDA Compliance Program 7356.002 and PAI Compliance Program CP7346.832

The first hour provides the investigator with an initial view of whether those systems operate as an integrated whole—or exist primarily as a collection of procedures assembled for inspection.

The inspection begins before the opening meeting

The assessment starts at the gate.

I notice whether security personnel know how to handle the arrival of an investigator, whether senior management and the quality unit are promptly informed, and whether entry is facilitated professionally. A short, reasonable delay may be understandable. A prolonged delay accompanied by visible confusion, repeated telephone calls, or unexplained restrictions immediately raises questions.

What is happening during that delay?

Is the organization simply locating the appropriate personnel, or is it preparing the site for my arrival? Are employees being moved, records being gathered selectively, production areas being cleaned, or computer systems being accessed before I enter?

None of these possibilities is assumed without evidence. But the clock, the explanations provided, and the sequence of events become part of the inspection context.

The reception area also provides clues. Are visitor and gowning controls followed consistently, or are they improvised for the investigator? Am I required to follow the same safety and hygiene requirements as every other visitor? A facility that abandons its own procedures to accommodate an investigator inadvertently demonstrates that procedures are negotiable.

The opening meeting: alignment or theater?

The strongest opening meetings are brief, accurate, and operationally grounded.

The right people are present. Site leadership understands the facility’s products, processes, major risks, and recent performance. Quality leaders speak with authority rather than waiting for operations or corporate counsel to approve every sentence.

I listen closely to how the site describes itself.

Can management clearly explain what the facility manufactures, which markets it supplies, what has changed since the previous inspection, and where its most significant quality risks reside? Does the site presentation correspond with registration information, approved applications, organization charts, manufacturing licenses, and the actual activities observed later?

I may also ask a straightforward question: “Who here has read the FDA compliance programs applicable to this facility?” These programs provide a road map to the inspection objectives, the areas we examine, and the kinds of questions you should be prepared to answer. Have the relevant sections been incorporated into onboarding and ongoing training for quality, manufacturing, laboratory, and regulatory personnel? Can those individuals explain how the program relates to their responsibilities? I am interested in more than whether someone has read the document. I want to understand whether the firm has used it to evaluate its operations and identify gaps before I arrive.

An overproduced presentation is not reassuring if leaders cannot answer straightforward questions. Statements such as “we have had no significant quality issues” are particularly unhelpful when the site manufactures complex sterile, biological, or otherwise high-risk products. Mature organizations have problems. What distinguishes them is how quickly they detect, escalate, investigate, and correct those problems.  Be honest about issues because it does not look good if the investigator is told at the start of the inspection that there have been no issues and then they find that there have been.  You want to build trust with the investigator from the “get go”.  

I am more confident in a site leader who says, “We experienced an adverse environmental-monitoring trend in February, stopped the affected operation, identified weaknesses in disinfection practices, and are verifying CAPA effectiveness,” than one who insists everything is fine.  Be assured that the investigator is going to want to verify, through review of documents and interviews with personnel, that what was stated to them is what happened.  FDA investigators are trained to “trust but verify”.  

Who actually owns quality?

One of the most consequential early signals is the relationship between the quality unit and operational management.

When I direct a question to the Head of Quality, does that person answer? Or does the plant head intervene? Can Quality describe recent batch rejections, significant deviations, out-of-specification results, complaints, recalls, supplier issues, and overdue CAPAs without first seeking permission?

The Quality unit need not dominate the meeting. It must, however, demonstrate independence, technical competence, and access to senior leadership.

Warning signs include:

1. Quality personnel repeatedly deferring to manufacturing on quality decisions.  
2. Senior leaders minimizing known problems before Quality can explain them.  
3. Disagreement about whether an event was significant.  
4. Quality representatives unfamiliar with major deviations or product complaints.  
5. Commercial or supply considerations being offered as justification for incomplete investigations.  
6. Responsibilities that appear clear on the organization chart but become ambiguous in conversation.

A pharmaceutical quality system is not measured by the number of SOPs under quality ownership. It is measured by whether Quality can stop an operation, reject a batch, require a scientifically adequate investigation, and maintain that position when schedule or revenue is at risk.

Document requests reveal how your systems work

An experienced investigator’s initial request is rarely random. It is designed to test the organization’s command of its information.

Depending on the inspection scope, I may request the site master file, product list, organization chart, recent quality metrics, deviation and CAPA registers, complaint data, recalls, out-of-specification results, stability failures, rejected batches, change controls, data-integrity assessments, or the status of commitments from a previous inspection.

I am observing more than the contents.

How long does retrieval take? Is the document complete? Does it cover the correct time period in question and how was the time period covered determined? Are records presented in their original or appropriately controlled form? Can the site explain how the report was generated? Does the electronic system produce the same information as the spreadsheet prepared for the inspection?  Be assured an investigator will cross- compare documents and information for consistency and to alert any anomalies.

A polished spreadsheet that cannot be reconciled with the source system will raise a “red flag” in the eyes of an investigator.

I also note whether records are supplied consistently. If every request triggers a long private discussion, multiple revisions, or negotiation over wording, the inspection will become more forensic. If information arrives promptly, with known limitations openly identified, confidence develops more quickly.

Transparency does not shorten an inspection by concealing complexity. It prevents ordinary complexity from being mistaken for concealment.

The front room–back-room interface

The inspection-management process itself often mirrors the company’s quality culture.

A well-managed front room has a knowledgeable coordinator, an accurate request log, controlled document transfer, appropriate subject-matter experts, and a process for recording commitments. The back room verifies that documents are responsive, complete, and protected from inadvertent disclosure of unrelated confidential information.

That is legitimate inspection management.

The concern begins when the back room becomes an answer-engineering function. Long delays, rehearsed responses, documents altered for presentation, or subject-matter experts instructed to avoid direct answers suggest that the site is trying to manage the investigator rather than demonstrating the system.

Experts should be prepared, but they should sound like experts—not witnesses reciting approved language. The strongest subject-matter experts distinguish clearly among fact, technical judgment, and information that still needs verification.

“I do not know, but I will confirm it from the record” is a credible answer.

"I believe we always do that" is not.

The first walk-through: where procedures meet reality

If the opening meeting tells me what the organization believes about itself, the walk-through tells me how work is actually performed.

I look at personnel and material flows, equipment status, segregation, housekeeping, maintenance condition, line clearance, label control, cleaning practices, environmental controls, and the accessibility of current procedures. I examine status labels, logbooks, notebooks, loose paper, whiteboards, temporary repairs, uncontrolled forms, shared passwords, and local workarounds. Sometimes, I simply stand and observe operations for an hour or more to understand how work is actually performed. While firm management may want to move the inspection along, I determine how long I need to observe. I may also return to a location later or observe the same operation after a shift change to assess consistency across personnel and shifts.

The objective is not to find cosmetic imperfections. A scratch on a wall is not equivalent to loss of process control. The question is whether the physical environment supports the process and whether observed conditions agree with the records.

Small inconsistencies can be highly informative:

1. An equipment log shows a unit as clean while visible residue remains.  
2. A room is labelled ready for use, but an alarm is active.  
3. Operators describe a sequence differently from the approved batch record.
4. Materials are physically segregated, but their electronic status is unclear.
5. A temporary repair has no associated work order or quality assessment.
6. An analyst uses a private calculation sheet outside the validated system.  
7. Personnel perform an undocumented step because “that is how we make the process work.”

I follow discrepancies, not appearances. A beautiful facility can have a weak quality system, while an older facility can remain in control through sound maintenance, disciplined operations, and scientifically justified controls. To understand how those controls work in practice, I may choose to begin an inspection during the night shift, when fewer managers are present, and contract vendors may be cleaning critical areas. Observing these activities can be particularly important if the firm has never directly observed or assessed its night-shift cleaning practices. ‍

The operators are often the clearest signal

Senior management explains the system. Operators reveal whether the system is usable.

I may ask an operator to describe what happens when an alarm occurs, when a result looks unusual, when a step is missed, or when the procedure does not match the situation. I am not testing memorization. I am evaluating whether the employee understands the process, its risks, and the escalation pathway.

A healthy quality culture is visible when employees can stop, question, document, and escalate without fear.

A weak culture becomes visible when employees look toward a supervisor before answering, use compliance slogans without explaining the underlying risk, or indicate that deviations are discouraged because they affect performance metrics.

Training records cannot compensate for a process that employees do not understand. A signature establishes that training was documented; it does not establish competence.

Data integrity: the question beneath every question

Data integrity is not confined to the laboratory, and it is not simply an information-technology issue. It is the basis on which the investigator decides whether the evidence can be trusted.

FDA emphasizes that senior management is responsible for establishing a quality culture in which data integrity is a core value and employees are encouraged to identify and report problems. FDA has also warned that failures of management support can allow quality systems to break down. (Source: FDA Data Integrity and Compliance With Drug CGMP Guidance)

During the first hour, I look for early indications of data vulnerability:

1. Uncontrolled user access or shared credentials.
2. Paper transcriptions where direct electronic capture is available.
3. Unexplained duplicate forms or unofficial notebooks.  
4. Reports that exclude aborted, invalidated, or repeated activity.  
5. Audit trails that are unavailable, disabled, or rarely reviewed.  
6. Metadata that the responsible manager cannot retrieve or interpret.  
7. A culture in which undesirable data are treated as inconvenient rather than informative.  

One unexplained inconsistency does not prove systemic data-integrity failure. But it changes the inspection strategy. If the reliability of one record is uncertain, I must determine whether the issue is isolated, procedural, technical, behavioral, or systemic.

How the site discusses bad news

Perhaps the most revealing first-hour question is some version of: “What is the most serious quality problem this site has faced?”

The substance of the problem matters, but so does the narrative.

Does management describe the event consistently? Was the patient or product risk assessed promptly? Did the investigation go beyond “operator error”? Were other batches, products, systems, and sites considered? Was the root cause supported by evidence? Did CAPA change the conditions that allowed the failure? Has effectiveness been demonstrated?

Weak organizations tell reassuring stories. Compliant organizations show evidence.

Attributing recurring failures to individuals is particularly concerning when the process design, workload, supervision, procedure usability, training model, or facility configuration has not been examined. Recent FDA enforcement correspondence continues to emphasize deeper analysis of conditions that create “human error,” rather than using the phrase as the endpoint of an investigation. You can see an example here -  FDA warning letter example

What predicts the rest of the inspection?

By the end of the first day, I am usually assessing five propositions:

1. Does leadership know the true condition of the operation?  
2. Can the quality unit act independently and effectively?  
3. Do records provide a reliable reconstruction of what occurred?
4. Do procedures, employee practices, physical conditions, and electronic data agree?
5. Does the organization investigate bad news—or manage its visibility?

When these elements align, the inspection is likely to remain risk-based and confirmatory. The investigator can test representative systems, follow meaningful signals, and evaluate the facility’s state of control efficiently.

When they do not align, the inspection expands. More source records are requested. Audit trails receive closer attention. Individual events are traced across products and systems. Interviews broaden. Questions shift from “Was this deviation handled appropriately?” to “Can this organization reliably detect and disclose deviations?”

That shift is decisive.

The lesson for industry is not to perfect the opening meeting. It is to build an operation that does not require a performance when an investigator arrives.

The first hour goes well when leadership routinely knows its risks, quality has real authority, employees can speak honestly, systems retain complete and accurate evidence, and problems are treated as opportunities to protect the patient—not threats to the site’s metrics.

An investigator does not expect perfection. We expect control, scientific integrity, reliable records, and honesty.

‍About Author

‍‍Elvia J. Cervantes spent 32+ years with the U.S. FDA as a Branch Director, Compliance Officer, and Investigator, specializing in GCP, GMP, BIMO, and Trial Master File inspections. She is a recipient of the FDA Investigator of the Year and Commissioner's Awards.

At AuraGxP she advises on inspection readiness and provides quality control review of high-risk audit reports, bringing the regulator's perspective to every engagement.