15 Complaints, 12 Uninvestigated, One Expired Quality Agreement: Lessons from Ubi Pharma's FDA 483
Pharmacies reported empty vials. Patients received underfilled product. Over two years, 15 complaints accumulated — 12 were never investigated. The Quality Agreement with the US customer had expired 17 months earlier. No Field Alert Reports were filed. Investigations that did occur were conducted without complaint samples, then invalidated based on manufacturing records alone. The Ubi Pharma 483 reveals what happens when contract manufacturing quality oversight depends on agreements that no one tracks.
On October 14, 2025, FDA investigators issued a Form 483 to Ubi Pharma Inc. at their facility at 45 Guangfu N. Road, Hukou, Hsinchu, Taiwan. The findings exposed a contract manufacturing operation where the foundational quality agreement with the US customer had expired, market complaints for injectable products went uninvestigated, and the complaint investigations that did occur were conducted without ever collecting the complaint samples. The result: 15 market complaints for underfill vials, empty vials, and crack/leak defects accumulated since June 2023 — and 12 were never investigated.
Ubi Pharma manufactures injectable drugs for the US market under contract for a US customer. The Quality Agreement governing the relationship was signed on May 20, 2019, effective for five years, expiring on May 19, 2024. As of the October 2025 inspection, the firm had not reviewed or renewed the Quality Agreement with its customer. For 17 months, the contract manufacturer had been producing injectable drugs for the US market without a current, binding agreement defining quality responsibilities between the two parties.
The expired Quality Agreement compounded every downstream failure. The initial agreement did not include a provision for timely submission of Field Alert Reports. Under 21 CFR 314.81(b)(1), FARs must be submitted within three working days of information concerning significant changes or deterioration in a distributed drug product. Empty vials and underfilled injectable product clearly meet that threshold. The US customer decided not to investigate and not to file FARs. According to the expired agreement, Ubi Pharma was responsible for complaint investigation. Ubi Pharma did not investigate at least 12 complaints. No FARs were submitted by either party. The regulatory reporting obligation fell through the gap between two organizations — and no one picked it up.
When a Quality Agreement expires and neither party notices for 17 months, the agreement was never functioning as a quality system control — it was functioning as a filing cabinet document.
What the FDA Found
Three observations exposing systemic failures in complaint handling, quality oversight, and equipment qualification at a contract manufacturing facility
21 CFR 211.22 requires that the responsibilities and procedures applicable to the quality control unit be in writing and fully followed. At Ubi Pharma, the FDA found that the quality unit’s most fundamental external obligation — maintaining a current Quality Agreement with the customer whose product they manufacture — had lapsed. The initial QA did not address FAR submission timelines. When complaints arrived, the expired agreement’s allocation of investigation responsibility created confusion: the US customer decided not to investigate, and Ubi Pharma failed to investigate 12 of 15 complaints. The quality control unit’s procedures were neither in writing (the agreement had expired) nor fully followed (investigations were not conducted).
21 CFR 211.198 requires thorough review of any unexplained discrepancy and written records of investigations including follow-up. The FDA found that of 15 market complaints for partially filled vials, empty vials, and crack/leak vials, only 3 were investigated. Those three investigations were conducted without the complaint sample. The firm had no documentary evidence of attempting to acquire complaint samples. Based solely on manufacturing records, the firm invalidated the complaints. The FDA noted: “It is not clear how you determined that the complaint samples were not counterfeit without acquiring complaint samples or pictures.” When investigators reviewed the visual inspection reports for the three implicated batches, they confirmed the presence of underfill vials — directly contradicting the invalidation conclusions.
Separately, the FDA identified a change control failure under 211.22. When Ubi Pharma replaced equipment on the capping machine, the firm did not follow its own change control procedure. No product quality impact assessment was conducted for the equipment change. A CAPA had been issued for a related problem, but it was ineffective — a similar deviation was subsequently recorded. The capping machine change is directly relevant to the underfill and empty vial complaints: capping equipment that is replaced without impact assessment and generates repeated deviations is a plausible source of the very defects customers were reporting.
A third observation addressed equipment qualification under 21 CFR 211.63. The visual inspection equipment — the last line of defense against defective vials reaching the market — was not qualified for the defect categories it was supposed to detect. The qualification was based on a limited number of samples. The Production Manager could not produce a list of defect vials or documents describing defect types. When the equipment responsible for catching underfill and empty vials has not been qualified to detect those specific defect categories, the visual inspection step is not a control — it is a formality.
15
Market Complaints
For underfill, empty, and crack/leak vials since June 2023
12
Uninvestigated
Complaints that were never investigated by either party
17
Months Expired
Duration the Quality Agreement had lapsed without renewal
Why This Keeps Happening
The root cause is not one firm's negligence — it is the structural gap in how contract manufacturing quality relationships are documented, tracked, and enforced
The Ubi Pharma 483 is a case study in CDMO quality oversight failure. Every observation traces back to a single architectural problem: when quality obligations between a contract manufacturer and its customer depend on a static document that no one tracks, and when complaint workflows span organizational boundaries without system-level enforcement, the gaps the FDA found are not exceptions. They are the predictable output of a system that was never designed to prevent them.
Quality Agreements treated as one-time documents rather than active system controls
The Quality Agreement between Ubi Pharma and its US customer was signed in 2019 with a five-year term. When it expired in May 2024, neither party renewed it. This is not unusual — Quality Agreements in contract manufacturing are frequently treated as procurement documents rather than living quality system elements. When a QA sits in a file cabinet without automated expiration tracking, renewal workflows, or periodic review triggers, the document degrades from a binding quality control into a historical artifact. The 17-month gap was not discovered internally; it was discovered by the FDA.
Complaint investigation responsibility undefined across organizational boundaries
The expired Quality Agreement assigned complaint investigation responsibility to Ubi Pharma. The US customer decided not to investigate and not to file FARs. Ubi Pharma did not investigate 12 of 15 complaints. The result: no one investigated. This is the predictable failure mode when complaint workflows cross organizational boundaries without system-level enforcement. When investigation responsibility is defined in a document rather than in a shared workflow system, each party can assume the other is handling it — and neither does.
Investigations conducted without complaint samples, then used to invalidate complaints
The three complaints that were investigated were investigated without the actual complaint samples. The firm had no documentary evidence of even attempting to acquire samples. Manufacturing records alone were used to invalidate the complaints. But when FDA investigators reviewed the visual inspection reports for the implicated batches, they found underfill vials. The investigation conclusions contradicted the production data. Conducting complaint investigations without physical samples is not an investigation — it is a desk exercise that produces conclusions unsupported by evidence. Under 21 CFR 211.198, it does not meet the standard for thorough review.
Equipment changes made without impact assessment, generating repeat deviations
Ubi Pharma replaced equipment on the capping machine without following its own change control procedure and without conducting a product quality impact assessment. A CAPA was already open for a similar issue, but it was ineffective — the same type of deviation recurred. Meanwhile, 15 complaints for underfill, empty, and cracked vials accumulated. The connection between an uncontrolled capping machine equipment change and vial integrity complaints is not subtle. But without a system that links change control records, deviation trends, and incoming complaints, the correlation was never identified.
When a contract manufacturer investigates complaints without collecting samples, invalidates findings that contradict its own visual inspection data, and operates under an expired Quality Agreement for 17 months — the problem is not that people failed. The problem is that no system required them to succeed.
Paper-Based vs. System-Enforced CDMO Quality Oversight
How architectural differences determine whether contract manufacturing complaints are investigated or forgotten
Each comparison below addresses a specific gap documented in the Ubi Pharma 483. The architectural approach does not add layers of manual coordination between contract parties — it eliminates the conditions that allowed these failures to persist for over two years.
Quality Agreement Lifecycle Management
Quality Agreement signed at contract initiation, filed, and forgotten. No automated expiration alerts. No periodic review triggers. No system-level link between the QA and the operational workflows it governs. When the agreement expires, production continues under the same terms — or no terms at all. At Ubi Pharma, 17 months passed before anyone noticed the QA had lapsed.
Quality Agreements are registered as active system documents with defined expiration dates, review periods, and renewal workflows. The system generates alerts at 90, 60, and 30 days before expiration. If a QA expires without renewal, the system flags all associated product lines and escalates to quality leadership. Production planning receives notification that the QA status is non-current. The agreement's terms — including complaint handling responsibilities and FAR obligations — are encoded as workflow rules, not just document clauses.
Cross-Organization Complaint Investigation
The Quality Agreement states which party is responsible for complaint investigation. When a complaint arrives, the receiving party must manually determine responsibility, route the complaint, and track whether the responsible party actually investigates. At Ubi Pharma, the US customer decided not to investigate. Ubi Pharma did not investigate 12 of 15 complaints. Neither party submitted FARs. The document said someone was responsible; the system did not enforce it.
Complaints for contract-manufactured products are routed into a shared workflow that assigns investigation tasks based on the Quality Agreement terms. Investigation timelines are enforced with escalation rules — if the responsible party does not initiate investigation within the defined period, the complaint escalates automatically to both quality units. FAR evaluation is triggered at complaint classification regardless of which party investigates. Neither party can close the complaint until investigation, sample collection, and regulatory reporting evaluation steps are documented.
Complaint Sample Collection and Counterfeit Assessment
Investigator reviews manufacturing records, concludes the complaint is not supported by production data, and invalidates the complaint without ever requesting or receiving the physical sample. No documentary evidence of sample acquisition attempts exists. The firm cannot determine whether the complaint sample is counterfeit because it never examined the sample. At Ubi Pharma, visual inspection data for the implicated batches actually confirmed underfill — contradicting the investigation conclusions.
Complaint intake triggers an automated sample request to the reporting entity. The investigation cannot advance to root cause analysis until sample receipt is confirmed or a documented, QA-approved justification for proceeding without a sample is recorded. Investigation conclusions must reference physical evidence — laboratory test results, visual inspection comparison against qualified defect categories, or documented sample photographs. Manufacturing record review alone cannot be the sole basis for complaint invalidation when physical samples are available.
Change Control and Complaint Correlation
Equipment replacement on the capping machine is processed through change control. Underfill and empty vial complaints are processed through the complaint system. Deviations are processed through the deviation system. No system links the three data streams. The correlation between an uncontrolled equipment change and a spike in vial integrity complaints is never identified because the data lives in disconnected silos.
When a complaint is classified by defect type (underfill, empty, crack/leak), the system automatically cross-references recent change control records and deviation logs for the implicated equipment and process steps. If a capping machine equipment change coincides with the onset of underfill complaints, the system flags the correlation for investigation. CAPA effectiveness is evaluated against both deviation recurrence and incoming complaint trends — not in isolation.
What a Modern CDMO Quality System Must Do
Four architectural capabilities that prevent contract manufacturing quality gaps from becoming regulatory findings
Preventing the Ubi Pharma pattern requires capabilities that span organizational boundaries. The failures documented in this 483 were not confined to one firm’s quality system — they occurred in the space between two organizations, where neither party’s system enforced the obligations that the Quality Agreement was supposed to define. Closing that gap requires architecture, not agreements.
Quality Agreement Lifecycle Management with Operational Linkage
A CDMO quality system must treat Quality Agreements as active, system-managed documents — not static files. Expiration dates must trigger automated renewal workflows. The terms of the agreement — complaint handling responsibilities, FAR submission obligations, change notification requirements — must be encoded as operational workflow rules that the system enforces. When a QA expires, the system must escalate, not silently allow production to continue under expired terms. The 17-month gap at Ubi Pharma is architecturally preventable.
Cross-Organization Complaint Workflows with Enforced Timelines
When complaint investigation responsibility spans two organizations, the workflow must span them too. A complaint for a contract-manufactured product must be visible to both quality units, with enforced investigation timelines and automatic escalation when deadlines are missed. FAR evaluation must be triggered at complaint classification — independent of which party is responsible for investigation. The Ubi Pharma case shows what happens when one party decides not to investigate and the other party does not fill the gap: the complaint disappears. System-enforced cross-organization workflows prevent that disappearance.
Enforced Sample Collection with Evidence-Linked Investigation Closure
Complaint investigations must not be closeable without documented evidence that physical samples were requested, received, and tested — or that a QA-approved justification exists for proceeding without them. Investigation conclusions that rely solely on manufacturing record review, while physical samples were never collected, do not meet the standard of thorough review under 21 CFR 211.198. The system must enforce the evidentiary standard, not leave it to investigator discretion.
Integrated Change Control, Deviation, and Complaint Correlation
Equipment changes, process deviations, and market complaints must be linked in a single data architecture that enables automated correlation analysis. When 15 complaints for underfill and empty vials accumulate while the capping machine undergoes equipment replacement without impact assessment, the system should surface that correlation — not require a human to manually cross-reference three separate databases. CAPA effectiveness must be evaluated against complaint trends, not just deviation closure.
30
Facilities
Running on a single integrated quality platform (Cipla deployment)
100%
Part 11 Compliance
21 CFR Part 11 compliance across 10+ sites (Piramal deployment)
2,700
Hours Saved Annually
Through digitized batch records and automated workflows (Valent BioSciences)
From Gap to Prevention
A three-phase approach to building CDMO quality oversight that cannot produce the failures observed at Ubi Pharma
The objective is not to add more review layers to existing contract manufacturing workflows. It is to build a system where the failures documented at Ubi Pharma — expired Quality Agreements, uninvestigated complaints, investigations without samples, uncontrolled equipment changes, unqualified inspection equipment — are architecturally impossible.
Phase 1: Assess — Audit every Quality Agreement, complaint backlog, and equipment qualification gap
Start with a complete inventory of active Quality Agreements across all contract manufacturing relationships. For each: Is the QA current? Does it define complaint handling responsibilities with specificity? Does it address FAR submission timelines? Does it require change notification between parties? Then audit the complaint backlog: How many complaints are open? How many were closed without sample collection? How many were invalidated based on manufacturing records alone? Finally, review equipment qualification records for inspection equipment: Is each defect category the equipment is expected to detect included in the qualification protocol? The Ubi Pharma 483 tells you exactly what the FDA will look for. Map your current state against those expectations.
Phase 2: Implement — Deploy system-enforced workflows that span organizational boundaries
Replace static Quality Agreements with system-managed documents linked to operational workflows. Encode complaint handling responsibilities, investigation timelines, FAR obligations, and change notification requirements as workflow rules — not document clauses. Deploy shared complaint investigation workflows that are visible to both the contract manufacturer and the customer's quality unit, with enforced timelines and automatic escalation. Implement mandatory sample collection gates in complaint investigation workflows. Link change control records, deviation logs, and complaint data into a single architecture that enables automated correlation analysis. The goal is a system where the Ubi Pharma failure pattern is architecturally impossible.
Phase 3: Validate — Prove the system works under audit conditions across organizational boundaries
Validate the CDMO quality oversight system under 21 CFR Part 11 requirements, ensuring complete audit trails, electronic signatures at each workflow gate, and data integrity across the complaint-to-CAPA-to-batch lifecycle. Run challenge scenarios that mirror the Ubi Pharma findings: let a Quality Agreement reach its expiration date and verify the system escalates. Process a complaint for a contract-manufactured product and verify that investigation cannot close without sample collection evidence and FAR evaluation. Introduce an equipment change and verify that the system correlates it with subsequent complaint trends. Document the validation evidence. When the FDA arrives, the system's architecture should demonstrate compliance across organizational boundaries — not just within a single quality unit.
Ubi Pharma manufactured injectable drugs for the US market for 17 months under an expired Quality Agreement. Fifteen complaints accumulated. Twelve were never investigated. The three that were investigated contradicted the firm’s own visual inspection data. The question every CDMO quality leader should ask: do our systems track Quality Agreement expiration, enforce cross-organization complaint workflows, and link equipment changes to complaint trends — or do we depend on someone remembering to check?
The Ubi Pharma 483 is not primarily a complaint handling failure — it is a contract manufacturing quality oversight failure. The distinction matters. Every observation in this 483 traces back to the space between two organizations: a contract manufacturer in Taiwan and its US customer. The Quality Agreement that was supposed to govern that space had expired. The complaint investigation responsibilities it defined were not enforced by either party. The regulatory reporting obligations it should have addressed were never included. The equipment changes that may have caused the defects were not communicated across the relationship.
This is the structural vulnerability of contract manufacturing. When quality obligations are defined in documents rather than enforced in systems, and when those documents span organizational boundaries where neither party’s quality system has full visibility, the gap is not a risk — it is a certainty. The only question is how long it takes for the gap to produce a finding.
Under 21 CFR 211.22, the quality control unit’s responsibilities must be in writing and fully followed. At Ubi Pharma, the written document had expired. Under 21 CFR 211.198, unexplained discrepancies must be thoroughly reviewed. At Ubi Pharma, 12 of 15 discrepancies were not reviewed at all, and the 3 that were reviewed produced conclusions contradicted by the firm’s own data. Under 21 CFR 314.81, Field Alert Reports must be filed within three working days of information concerning distributed product defects. No FARs were filed by either party. Under 21 CFR 211.63, equipment must be of appropriate design for its intended use. The visual inspection equipment was not qualified for the defect categories it was supposed to detect.
For quality leaders at CDMOs and their customers, the Ubi Pharma finding is a prompt for a specific audit. Not a general review — a specific, relationship-by-relationship test: Is every Quality Agreement current? Does each QA define complaint handling, FAR submission, and change notification with operational specificity? Are complaint investigation workflows enforced across organizational boundaries, or do they depend on manual coordination? Can our system close a complaint investigation without physical sample evidence? Do our change control records link to complaint trend data? If the answer to any of these questions reveals a gap, the system has the same structural vulnerability that the FDA identified in Hsinchu. The time to address it is before the inspection, not after.
Related Articles
Three Facilities, Three FDA Actions, Five Architectural Gaps: How AI Agents Address Cipla's Regulatory Exposure
Between 2023 and 2026, three Cipla facilities — Pithampur, Raigad, and Pharmathen Greece — received FDA enforcement actions documenting the same five systemic failures: complaint investigation, CAPA effectiveness, electronic data review, contamination control, and QC oversight. LeucineOS AI agents map directly to each gap.
35% OOS Invalidations, Zero Scientific Justification: Lessons from Aurobindo Pharma's FDA 483
A February 2026 FDA 483 at Aurobindo Pharma's Unit-III found 35% of OOS invalidations in the QC Chemistry lab — with 57% blamed on analyst error and 18% on equipment, none supported by adequate scientific justification. Batches shipped to the US after unresolved Grade A maintenance interventions.
Equipment Swapped, Cleaning Not Revalidated, OOS Dissolved Away: Lessons from Dr. Reddy's FDA 483
A December 2025 FDA 483 at Dr. Reddy's FTO-SEZ facility in Srikakulam found cleaning validation not performed after equipment replacement, OOS dissolution results invalidated despite contradictory evidence, and process qualification gaps — all traceable to a single uncontrolled equipment change eighteen months earlier.
Newsletter
Stay ahead in the Industry
Regulatory updates, pharma quality insights, and AI in manufacturing — written for quality leaders, not marketers.
Please use your official work email. Personal email addresses (Gmail, Yahoo, etc.) will not receive the newsletter. No spam. Unsubscribe anytime.
Ready to see what an AI-native quality platform looks like? Leucine unifies quality management, regulatory compliance, and production operations into one intelligent system.
