Crumbling Tablets, Closed Investigations: Lessons from Cohance Lifesciences' FDA 483
A pharmacy reported tablets crumbling in the bottle. The investigation compared photos and concluded the complaint sample resembled the control. No physical sample was collected. No laboratory testing was performed. No Field Alert Report was filed. The Cohance 483 reveals a complaint system designed to close investigations, not to find answers.
On August 12, 2025, FDA investigators issued a 483 observation to Cohance Lifesciences Limited (formerly Ra Chem Pharma) at their Secunderabad, Telangana facility. The finding was direct: “Written records of investigation of a drug complaint do not include the follow-up.” What the observation revealed was not a minor documentation gap. It was a complaint handling system that treated investigation closure as the objective rather than root cause identification and corrective action.
The specific case centered on Complaint #PC/NCR/FDF/23-020, filed on October 5, 2023, after a pharmacy customer reported that tablets USP were disintegrating in the bottle — described as crumbly, pitted, with dust from the pills themselves, and lacking crisp markings. The investigation’s root cause analysis concluded that “based on the investigation, it can be drawn that the tablets photographs of the control/reference sample resemble the tablet photographs of the complaint sample. The probability of generation of powder could be due to nature of the tablets.” In other words, the firm compared photos, decided the complaint sample looked like what they expected, attributed the defect to the inherent nature of the product, and closed the case.
The FDA found that complaint investigations were closed without adequate follow-up, without collecting actual complaint samples, without performing laboratory testing on complaint samples, and without assessing the impact on distributed batches. To compound the failure, the firm did not submit timely Field Alert Reports (FARs) for defects that could affect product quality — a separate regulatory obligation under 21 CFR 314.81(b)(1) that exists precisely because complaint data can signal distributed product risk.
When a complaint investigation’s root cause analysis compares photographs and concludes the defect is “due to the nature of the tablets,” the system is not investigating — it is rationalizing.
What the FDA Found
A complaint handling process that failed at every required step under 21 CFR 211.198 and 21 CFR 314.81
21 CFR 211.198 sets out the requirements for complaint handling in pharmaceutical manufacturing. The regulation requires that written records of each complaint include, among other things, the findings of the investigation and the follow-up. The word “follow-up” is not decorative. It mandates that when an investigation identifies a potential quality defect, the firm must take concrete subsequent action — collecting and testing the complaint sample, evaluating whether other distributed batches are affected, and determining whether corrective actions are needed.
At Cohance Lifesciences, the investigation for Complaint #PC/NCR/FDF/23-020 failed to perform any of these steps. The root cause analysis relied entirely on visual comparison of photographs. No physical complaint sample was collected from the pharmacy or the customer. No laboratory testing — dissolution, hardness, friability, or any other relevant assay — was performed on the complaint samples. No assessment was conducted to determine whether other batches of the same product might exhibit similar degradation. The investigation was closed as though the photographic resemblance between complaint and control samples constituted a finding.
Separately, 21 CFR 314.81(b)(1) requires approved NDA/ANDA holders to submit Field Alert Reports within three working days of receiving information concerning any significant chemical, physical, or other change in, or deterioration of, a distributed drug product. Tablets that are disintegrating, crumbling, and shedding dust inside their container represent a clear physical change that could affect safety or efficacy. The failure to file timely FARs means that the FDA was not notified of a potential distributed product defect — leaving both the agency and the public without information needed to assess patient risk.
The combined failure creates a particularly concerning pattern: the firm’s complaint system did not escalate the finding internally (no follow-up investigation, no batch impact assessment), and it did not escalate externally (no Field Alert Report to the FDA). The complaint effectively disappeared into the system.
0
Samples Collected
Complaint samples collected for laboratory testing
0
FARs Filed
Field Alert Reports filed for product quality defects
0
Impact Assessments
Distributed batch impact assessments conducted
Why This Keeps Happening
The root cause is not negligence — it is a complaint system architecture that allows investigations to close without completing required steps
The pattern is structural. When a complaint handling system permits closure without enforcing sample collection, laboratory testing, batch impact assessment, and regulatory reporting evaluation, the gaps documented at Cohance are not exceptions — they are the expected output of the system’s design.
No enforced workflow gates between investigation and closure
When a complaint handling system allows an investigator to close a case without documenting that physical samples were requested, received, and tested, the system is structurally permissive. The absence of mandatory workflow checkpoints means that the path of least resistance — closing without follow-up — is also the easiest path through the system. At Cohance, the investigator could conclude the case based on photograph comparison alone because nothing in the system required anything more.
Disconnected complaint and regulatory reporting systems
Field Alert Report obligations under 21 CFR 314.81 are triggered by complaint data, but in many firms the complaint management system and the regulatory reporting function operate as separate processes with separate ownership. When the complaint system does not automatically flag potential FAR-triggering events based on complaint classification — physical defect, chemical change, contamination — the regulatory reporting step depends entirely on manual recognition and escalation. That manual handoff failed here.
No batch-level traceability linking complaints to distribution records
Assessing impact on distributed batches requires connecting complaint data to batch genealogy and distribution records. When these data sets live in separate systems — complaints in one database, batch records in another, distribution records in a third — the investigation team faces a manual correlation exercise that is time-consuming and easy to skip. The absence of integrated traceability makes batch impact assessment an optional effort rather than a system-generated output.
Root cause analysis templates that accept narrative conclusions
A root cause analysis that concludes 'the probability of generation of powder could be due to nature of the tablets' is not a root cause — it is a hypothesis, and a weak one. When investigation templates accept free-text narrative conclusions without requiring structured causal analysis (fishbone diagrams, 5-why chains, statistical comparison to specification limits), the system permits investigators to close with speculation rather than evidence.
The question is not whether investigators should have done more. The question is why the system allowed them to do less.
Paper-Based vs. System-Enforced Complaint Handling
How architectural differences determine whether complaint investigations produce answers or rationalizations
Each comparison below addresses a specific gap documented in the Cohance 483. The architectural approach does not add layers of manual oversight — it eliminates the conditions that allowed these failures to exist.
Sample Collection and Laboratory Testing
Investigator visually compares complaint sample photographs to control sample photographs, concludes they resemble each other, attributes the defect to 'nature of the tablets,' and closes the complaint. No physical sample collected. No laboratory testing performed. No specification limits referenced.
Complaint intake triggers an automated sample request workflow. The investigation cannot advance to root cause analysis until sample receipt is confirmed or a documented justification for proceeding without a sample is reviewed and approved by QA. Laboratory testing results are linked directly to the investigation record, and root cause conclusions must reference quantitative data against specification limits.
Distributed Batch Impact Assessment
Investigation closes without evaluating whether other distributed batches of the same product exhibit similar defects. Batch records, stability data, and distribution records are stored in separate systems with no automated correlation. The investigator would need to manually search multiple databases to perform the assessment — so it does not happen.
When a complaint is classified as a potential product quality defect, the system automatically identifies all batches manufactured with the same process parameters, raw material lots, and equipment train. Distribution records are cross-referenced to determine which batches are in the field. A batch impact assessment is generated as a system output, not an investigator task.
Field Alert Report Filing
FAR obligations depend on an individual recognizing that a complaint meets the threshold for reporting under 21 CFR 314.81(b)(1). The complaint system has no automated link to regulatory reporting. When the investigation itself is inadequate, the data needed to trigger FAR awareness never surfaces.
At the point of complaint classification, the system evaluates the defect type against FAR-triggering criteria (significant physical, chemical, or biological change; deterioration; failure to meet specifications). When criteria are met, a FAR preparation workflow is initiated automatically with a three-working-day deadline, and the regulatory affairs team is notified. The complaint cannot be closed until the FAR decision is documented.
What a Modern Complaint System Must Do
Three architectural capabilities that prevent investigation gaps from becoming regulatory findings
Preventing the Cohance pattern requires architectural enforcement — not better SOPs, not more training, not periodic audits of a system that was never designed to enforce the steps the FDA expects. The three capabilities below directly address the root causes behind each finding.
Enforced Investigation Workflow with Mandatory Gates
A complaint management system must enforce sequential workflow stages — intake, sample collection, laboratory testing, root cause analysis, batch impact assessment, regulatory reporting evaluation, CAPA linkage — with mandatory completion criteria at each gate. Investigators should not be able to advance to closure without documented evidence that each required step was completed or a justified, QA-approved deviation was recorded. This is not about adding bureaucracy; it is about making the system incapable of producing the gap that the FDA found at Cohance.
Automated Regulatory Reporting Triggers
Field Alert Report obligations, MedWatch submissions, and other regulatory reporting requirements should be evaluated automatically based on complaint classification data — not left to manual recognition. When a complaint is classified as involving a physical defect in a distributed product, the system should evaluate FAR criteria, flag the complaint for regulatory review, and initiate a time-bound reporting workflow. The three-working-day FAR deadline under 21 CFR 314.81 does not allow for the delays inherent in manual escalation chains.
Integrated Batch Genealogy and Distribution Traceability
Assessing the impact of a complaint on distributed batches requires real-time access to batch genealogy (raw materials, process parameters, equipment) and distribution records (which lots went where, and when). When these data sets are integrated into a single platform, the system can automatically generate a batch impact assessment the moment a product quality complaint is confirmed — identifying potentially affected lots, their distribution status, and whether retention samples are available for testing.
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 a complaint handling system that cannot produce the failures observed at Cohance
The objective is not to add more review steps to a broken process. It is to build a system where the failures documented at Cohance — closure without samples, closure without testing, closure without impact assessment, closure without regulatory reporting — are architecturally impossible.
Phase 1: Assess — Map every gap between current complaint workflows and 21 CFR 211.198 requirements
Conduct a structured gap assessment of the existing complaint handling process against the specific requirements of 21 CFR 211.198 and 21 CFR 314.81. For each complaint closed in the past 24 months, verify: Was a physical sample collected? Was laboratory testing performed? Was a batch impact assessment documented? Was regulatory reporting (FAR, MedWatch) evaluated? The Cohance 483 tells you exactly what the FDA will look for. Map your current state against those expectations before the FDA does it for you.
Phase 2: Implement — Deploy enforced workflows with mandatory gates, automated triggers, and integrated traceability
Replace free-form complaint investigation processes with structured, system-enforced workflows. Each investigation stage must have defined completion criteria and mandatory documentation requirements. Regulatory reporting triggers must be automated based on complaint classification. Batch impact assessment must pull from integrated batch genealogy and distribution data — not from manual cross-referencing of separate systems. The goal is a system where the Cohance failure pattern is architecturally impossible.
Phase 3: Validate — Prove the system works under audit conditions with complete traceability
Validate the complaint handling 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 Cohance finding: process a complaint for a physical product defect and verify that the system enforces sample collection, laboratory testing, batch impact assessment, and FAR evaluation before allowing closure. Document the validation evidence. When the FDA arrives, the system's architecture should demonstrate compliance — not just the content of individual investigation records.
Cohance Lifesciences received a complaint about tablets disintegrating in the bottle. Their system compared photographs, attributed the problem to the nature of the product, and closed the investigation. The question every quality leader should ask: could our system do the same thing?
The Cohance Lifesciences 483 is not an isolated incident. Complaint handling deficiencies under 21 CFR 211.198 are among the most frequently cited observations in FDA inspections of pharmaceutical manufacturing facilities. The pattern repeats across the industry: complaints are received, investigations are opened, and investigations are closed — but the steps between opening and closing fail to produce the evidence that the regulation requires. What makes the Cohance case instructive is the specificity of the failure. A pharmacy reported a tangible, observable product defect. The investigation did not collect the defective product. It did not test the defective product. It did not assess whether other products in the market might have the same defect. It did not report the defect to the FDA.
Each of these failures traces back to a system that did not require the step to be completed before the investigation could close. The root cause is not that individual investigators made poor decisions — it is that the system architecture permitted those decisions to stand. When a complaint handling system allows closure without follow-up, closure without follow-up is what you will get.
For quality leaders at multi-site pharmaceutical manufacturers, the Cohance finding is a prompt for a specific audit of their own complaint handling workflows. Not a general review — a specific, observation-by-observation test: can our system close a product quality complaint without collecting a physical sample? Can it close without laboratory testing? Can it close without a batch impact assessment? Can it close without evaluating FAR obligations? If the answer to any of these questions is yes, the system has the same structural vulnerability that the FDA identified in Secunderabad. 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.
