Continuous manufacturing (CM) produces pharmaceutical product in an uninterrupted stream rather than in discrete batches. FDA has issued guidance supporting the technology, and ICH Q13 — finalized in 2023 — provides the international framework for CM implementation and quality system requirements. For quality teams, CM introduces fundamentally different process monitoring, batch definition, and control strategy requirements that a traditional batch-oriented QMS is not designed to handle. Understanding those differences is the first step to building a quality system that supports CM without creating compliance gaps.

What continuous manufacturing is and why it matters

In batch manufacturing, raw materials are loaded, processed, and released as a defined lot. The batch is a unit of production with a start, an end, and a defined set of records. In continuous manufacturing, materials flow through the process without discrete starts and stops. Product is manufactured in a continuous stream, and the “batch” is defined differently — by time, by mass of material processed, or by some other operational boundary that the manufacturer defines and justifies to the regulator.

FDA has been actively encouraging CM adoption in pharmaceutical manufacturing for over a decade. The agency’s interest is straightforward: CM processes, when properly monitored and controlled, can produce more consistent product than batch processes because they eliminate the scale-up discontinuities and between-batch variability that are inherent in batch production. CM also enables faster manufacturing cycles, smaller facility footprints, and more responsive supply chains.

Johnson & Johnson received the first FDA approval for a solid oral dosage form product manufactured by continuous manufacturing in 2016. Since then, the number of approved CM products has grown, and FDA has approved CM for an expanding range of product types and dosage forms. [Source: FDA Guidance for Industry — Advancement of Emerging Technology Applications, 2017]

ICH Q13: the international framework for continuous manufacturing

ICH Q13 — “Continuous Manufacturing of Drug Substances and Drug Products” — was finalized by the International Council for Harmonisation in November 2022 and adopted by FDA, EMA, and other major regulatory agencies in 2023. It is the primary international guidance document governing CM implementation and quality system requirements for pharmaceuticals.

ICH Q13 covers CM for both drug substance (API) and drug product manufacturing. It addresses the key quality science concepts that distinguish CM from batch manufacturing: the control strategy, real-time release testing, the integrated process monitoring approach, and the definition and justification of the CM batch. [Source: ICH Q13 Guideline]

The guidance makes clear that the fundamental quality principles — process understanding, risk-based control, data integrity — are the same for CM as for batch manufacturing. What changes is how those principles are implemented in a process that runs continuously rather than in discrete units.

How CM changes the quality system

Batch definition

In a batch process, the batch is defined by the manufacturing cycle: everything produced from a defined set of inputs in a single manufacturing run. In CM, the manufacturer must define what constitutes a batch, justify that definition in the regulatory submission, and design the quality system to manage it. A CM batch might be defined as a specific duration of continuous operation (e.g., 24 hours of production), a defined mass of output, or a segment of a continuous run bounded by specific operational events.

The batch definition has direct implications for every batch-associated quality record: batch records, release testing records, certificates of conformance, and the nonconformance records that reference specific batch numbers. The quality system must be designed to generate and manage these records consistently with the approved batch definition.

Control strategy and real-time monitoring

ICH Q13 emphasizes that CM requires a robust control strategy — a defined set of controls that assure product quality throughout the continuous process. The control strategy for a CM process typically relies much more heavily on real-time process analytical technology (PAT) measurements than a batch process does. Rather than testing a sample of finished product against specification, CM quality assurance often involves continuous measurement of critical quality attributes during production, using spectroscopic or other inline analytical methods.

Real-time release testing (RTRT), where finished product is released based on real-time process and analytical data rather than traditional end-product testing, is a common feature of CM quality systems. RTRT requires regulatory approval and a validated analytical strategy, but it offers significant efficiency advantages — particularly for products with long end-product testing cycles.

The quality system must support the control strategy: capturing, storing, and making available the continuous stream of process data that demonstrates the process was in a state of control throughout the production run. This is a data management challenge at a scale that traditional paper-based or batch-oriented QMS platforms are not designed for.

Diversion and rejection of nonconforming material

In batch manufacturing, a nonconforming lot is identified, segregated, and submitted to the Material Review Board for disposition. In CM, nonconforming material must be diverted in real time — before it has a chance to mix with conforming product or continue through the process. This requires automated diversion capabilities built into the manufacturing system, triggered by the real-time process monitoring.

The quality system must document the diversion event: what triggered it, when it occurred, what quantity of material was diverted, and the disposition of the diverted material. The deviation report for a CM diversion event looks different from a traditional nonconforming lot record, but the underlying documentation requirements are the same — the event must be recorded, reviewed, and connected to any required corrective action.

Process validation

FDA’s 2011 process validation guidance describes three stages: process design, process qualification, and continued process verification. For CM, the continued process verification stage is especially important — and especially data-intensive. Because CM runs are long and the process is expected to operate continuously, the quality system must monitor process performance data continuously and flag deviations from the established control limits.

ICH Q13 provides specific guidance on process qualification for CM, including considerations for how to demonstrate process capability at the scale of continuous production rather than across a small number of discrete qualification batches. The statistical framework for CM process qualification is more sophisticated than for batch processes, and the quality team needs the analytical infrastructure to support it.

Data management requirements for a CM quality system

The data volume generated by a continuous manufacturing process is orders of magnitude larger than a batch process generating comparable product quantities. PAT instruments, process sensors, and environmental monitoring systems generate continuous data streams throughout the production run. All of that data is quality-relevant — it is the evidence that the control strategy was operating correctly throughout the run.

Managing this data requires a quality system with genuine data management capabilities: the ability to ingest large data streams from connected instruments, store them with appropriate metadata and timestamps, make them searchable and accessible for batch review and regulatory inspection, and connect them to the batch records and deviation records that reference specific time periods or process events.

Data integrity requirements apply to CM data with the same force as to any other quality-critical record. FDA expects that CM process data is attributable, legible, contemporaneous, original, and accurate (ALCOA). For continuous data streams from automated instruments, this means the system capturing the data must have validated data integrity controls, including protection against alteration and a complete audit trail.

Common quality system gaps for companies adopting CM

Organizations transitioning from batch to continuous manufacturing often find that their existing quality system was designed around batch concepts and does not map cleanly to CM requirements. The most common gaps include the following.

Batch record templates designed for discrete lots. A CM batch record needs to capture continuous process data over a defined time window, not a set of discrete manufacturing steps with individual sign-offs. Batch record templates designed for batch manufacturing require significant redesign for CM.

Nonconformance processes that assume discrete lots. Traditional nonconformance workflows are designed around identifiable lots that can be segregated and held. CM nonconformances require a different workflow that accounts for real-time diversion, continuous process recovery, and the documentation of which output was affected by a process excursion during a continuous run.

Change control processes that do not account for continuous operation. In batch manufacturing, a process change is implemented between batches. In CM, implementing a change requires careful consideration of how to manage the transition in a continuously operating system. The change control process and its documentation must reflect this complexity.

Insufficient data management infrastructure. The volume and velocity of CM process data typically exceeds what paper-based systems or basic electronic QMS platforms were designed to handle. Organizations adopting CM typically need to invest in data infrastructure that can manage continuous data streams alongside their traditional quality records.

How a cloud-based eQMS supports continuous manufacturing quality

A modern cloud-based eQMS provides several capabilities that are particularly valuable for CM quality management. Real-time data integration from connected manufacturing systems enables the quality team to monitor process performance as the run progresses, not just after the fact. Configurable batch record templates can be designed to match the specific batch definition used in the CM process. Deviation workflows can be configured to handle the diversion-and-recovery pattern characteristic of CM excursions.

Cloudtheapp’s platform includes 60+ applications covering quality, compliance, and operations management — all on a single validated system. The platform’s integration tools connect with MES, LIMS, and PAT data systems to consolidate the data streams that CM quality management requires. The no-code configuration capability allows quality teams to adapt batch record formats, deviation workflows, and control chart parameters to the specific requirements of their CM process without requiring IT development resources.

For regulated pharmaceutical manufacturers moving into continuous manufacturing, the quality system is not a secondary consideration — it is central to the regulatory approval pathway. A platform that can handle the data volume, the workflow complexity, and the documentation requirements of CM, while maintaining 21 CFR Part 11 compliance and ISO 13485 alignment, removes one significant barrier to CM adoption.

To see how Cloudtheapp supports continuous manufacturing quality programs, schedule a demo with the team.

Regulatory submissions for continuous manufacturing

For new product submissions that involve CM, FDA expects the submission to address the specific quality considerations described in ICH Q13: the batch definition and justification, the control strategy description, the process monitoring approach, the real-time release testing strategy (if applicable), and the approach to process validation including continued process verification.

For existing products that a company wants to transition from batch to CM, a prior approval supplement (PAS) or comparable regulatory mechanism is required. The supplement must demonstrate that the CM process produces product of equivalent or superior quality to the batch process and that the quality system is appropriately adapted to manage CM-specific quality requirements.

FDA has an Emerging Technology Program specifically designed to support early engagement with companies developing novel manufacturing approaches, including CM. Early engagement through this program can significantly reduce the regulatory uncertainty associated with first-of-kind CM implementations. [Source: FDA Emerging Technology Program]

Conclusion

Continuous manufacturing changes the practical reality of pharmaceutical quality management in ways that go beyond updating a few procedures. The batch definition, the control strategy, the real-time monitoring requirements, the data management challenge, and the deviation workflows all require substantive redesign for CM. Organizations that understand these requirements — and invest in the quality system infrastructure to support them — will be better positioned to realize the manufacturing flexibility and product quality benefits that continuous manufacturing offers. A modern cloud-based eQMS that can handle the data complexity and workflow requirements of CM is a necessary part of that infrastructure.

This post created by and appeared first on Cloudtheapp

An out-of-specification (OOS) result is any test result that falls outside the acceptance criteria established in a drug application, compendial standard, or manufacturer specification. FDA’s 2022 revised guidance requires a structured two-phase investigation: Phase I covers the laboratory, and Phase II covers the manufacturing process. OOS results that are not properly investigated, documented, and resolved are among the most frequently cited cGMP failures in FDA inspections.

Every regulated laboratory that tests pharmaceutical products, medical device components, or raw materials will eventually produce a result that falls outside an established limit. What happens in the next several hours determines whether that result becomes a documented, defensible investigation or a regulatory liability.

An out-of-specification result is not a quality failure by itself. It is a signal. The failure happens when the investigation is incomplete, the documentation is vague, or the result is invalidated without scientific justification. FDA investigators know this, and OOS-related citations appear consistently across drug and device inspection reports year after year.

This guide covers the regulatory definition, FDA’s current two-phase investigation framework, documentation requirements, common mistakes, and how a validated quality management system structures OOS workflows from initiation through closure.

What Is an Out-of-Specification (OOS) Result?

An out-of-specification result is any test result that falls outside the specifications or acceptance criteria established in a drug application, drug master file, official compendium, or by the manufacturer. FDA’s definition also applies to in-process laboratory tests that fall outside established specifications.

The term covers a broad range of situations: a finished product that fails potency testing, a raw material that falls outside purity limits, a stability sample that exceeds degradation thresholds, and a manufacturing in-process test result outside validated control limits. In each case, the same fundamental requirement applies: the result must be investigated.

FDA’s regulatory authority for OOS investigations comes from 21 CFR 211.192, which requires that all discrepancies or failures of a batch to meet any of its specifications be investigated. That investigation must be completed and documented before the batch is approved or rejected. The regulation makes no distinction between failures attributable to laboratory error and failures attributable to manufacturing problems — both require investigation.

OOS vs OOT vs OOE: Key Differences

Quality teams working in GMP environments encounter three related but distinct categories of anomalous results. Understanding the difference matters for triaging and investigation scope.

Out-of-Specification (OOS): A result that falls outside established acceptance criteria as defined in the specification, pharmacopeial standard, or regulatory filing. OOS results always trigger a formal investigation.

Out-of-Trend (OOT): A result that is within specification but shows a statistically significant deviation from historical data or the expected trend for that product or batch type. OOT results require review and documentation but follow a different and typically less intensive investigation path. Stability studies are the most common context for OOT assessments.

Out-of-Expectation (OOE): A result that is within specification and within historical trend, but differs from the expected outcome in a specific experimental context. OOE designation is used when a result is unexpected based on prior knowledge about the process or product, even though it technically passes the specification.

The distinction between these three categories shapes both the urgency of the response and the depth of investigation required. OOS results carry the highest regulatory risk and demand the most structured, documented response.

The Regulatory Basis: FDA’s 2022 OOS Guidance

FDA first issued guidance on OOS investigation in October 2006, formalizing an investigation framework that had developed through enforcement actions, warning letters, and court decisions dating back to the 1990s. In May 2022, FDA published a revised version that updated terminology for consistency with current guidance and clarified concepts related to outlier results and the practice of averaging OOS results. (FDA.gov)

The 2022 guidance applies to finished pharmaceutical products regulated under 21 CFR Parts 210 and 211. For medical device manufacturers operating under 21 CFR Part 820 and ISO 13485, the underlying principles of the two-phase investigation framework and documentation expectations apply equivalently through those regulations, even though FDA has not issued a parallel guidance document specific to devices.

The guidance defines OOS results broadly to include all in-process tests outside established specifications, not just finished product release tests. This scope is important: in-process failures that are not properly investigated are as problematic during an inspection as release failures.

Phase I: The Laboratory Investigation

Phase I is the laboratory-focused portion of the OOS investigation. Its purpose is to determine whether the OOS result was caused by an identifiable laboratory error. FDA’s guidance sets a clear expectation: the laboratory investigation should be completed within 20 business days of identifying the OOS result, although this is a target, not an absolute regulatory deadline.

The Phase I investigation should be conducted and documented by the laboratory analyst and reviewed by the laboratory supervisor or quality unit. Key elements include:

Review of analyst technique and instruments. The investigation begins with an assessment of whether the analyst followed the approved procedure exactly as written. Were the correct standards used? Were solutions prepared correctly? Was the instrument calibrated and operating within qualified parameters? Were integration parameters and calculations applied correctly? This review covers the raw data, including chromatograms, balance printouts, and instrument logs.

Assessment of sample preparation and storage. Sample preparation errors, including incorrect dilution, improper extraction, or sample degradation from improper storage, are among the most common identifiable causes of laboratory error. The Phase I investigation should document the condition of the sample, preparation records, and the handling history of the retained sample.

Analyst qualification records. The investigation should confirm that the analyst who performed the testing was qualified to perform that method. If qualification is not current, that finding must be documented and addressed.

Re-injection of retained solutions. If the existing sample solution is still valid, re-injection of that solution is permitted in Phase I to check for instrument or preparation error. A re-injection is not a retest. It tests the same prepared solution under the same conditions and is only permissible if the solution’s stability supports it.

Documentation of findings. Every action taken during Phase I must be documented in real time. Notes, calculations, instrument printouts, and the investigator’s conclusions must be preserved in the investigation record. If Phase I identifies a confirmed laboratory error with a specific, documented root cause, the investigation may be closed at Phase I. The original OOS result must remain in the batch record. The confirmed error must be documented, and corrective action must be assigned.

If Phase I does not identify a confirmed laboratory error, the investigation must proceed to Phase II. The guidance is explicit: Phase I cannot be used to simply reassign the result. A Phase I invalidation requires a specific, documented, scientifically justifiable cause.

Phase II: The Full-Scale Production Investigation

Phase II expands the investigation scope beyond the laboratory to include the manufacturing process, raw materials, equipment, and environmental conditions that could have caused the OOS result. The Phase II investigation is typically led by the quality unit with involvement from manufacturing, engineering, and where applicable, contract manufacturing or contract laboratory partners.

Phase II elements include:

Manufacturing process review. A thorough review of the batch production record, including all in-process checks, equipment logs, environmental monitoring results, and any documented deviations. Any deviation or anomaly observed during manufacturing that was not investigated at the time must be assessed for a causal relationship to the OOS result.

Root cause investigation. The Phase II investigation must include a documented root cause analysis. Methods such as fishbone diagrams, 5 Whys, or fault tree analysis are used to move beyond symptom description to the underlying cause of the failure. If no root cause can be confirmed, that conclusion must itself be documented with a clear explanation of what was investigated and why no cause was identified.

Retesting with additional samples. Retesting under Phase II requires the quality unit’s involvement and must follow a pre-defined retesting protocol that documents the justification for retesting, the number of samples, and the criteria for interpretation. Retesting is not an acceptable substitute for investigation. An OOS result cannot be discarded based solely on passing retest results. The original result stands and must be explained, not overridden.

Lot disposition decision. Phase II concludes with a documented batch disposition decision. If the investigation identifies a confirmed manufacturing cause, the batch must be rejected unless retesting under the approved protocol demonstrates that the product meets specification. If no cause is confirmed and retesting passes, the quality unit must document the rationale for disposition and accept responsibility for the decision.

CAPA initiation. Any confirmed OOS finding with a root cause must result in a formal corrective and preventive action to address both the immediate failure and the systemic conditions that allowed it to occur.

When Can an OOS Result Be Invalidated?

Invalidation of an OOS result without a confirmed, specific, documented laboratory error is one of the most serious findings an FDA investigator can make. The guidance is clear: averaging of OOS results with passing results to generate an acceptable composite result is not acceptable practice. A passing average does not resolve an OOS result. Each individual result must be evaluated.

Legitimate bases for invalidation include: a documented instrument malfunction confirmed by calibration or maintenance records, a documented sample preparation error with an identifiable cause, and a confirmed analyst technique error that is directly traceable to the specific sample and test. Even with a confirmed error, the investigation record must document the error’s nature, the evidence supporting the conclusion, and the corrective action assigned.

Documentation and Audit Trail Requirements

OOS investigations that cannot be reconstructed from the documentation record are treated as investigations that did not occur. FDA investigators examine not only whether an investigation was completed but whether the documentation demonstrates that it was completed contemporaneously, by qualified personnel, and with sufficient detail to support the conclusion.

The investigation record must include: the date the OOS was identified, the identity of the analyst and the method used, all raw data generated during Phase I, all decisions about Phase I scope and conclusions, the Phase II investigation scope and findings if initiated, the root cause conclusion, the batch disposition decision and the rationale, and the CAPA record if initiated.

An audit trail that captures who took each action, when, and with what data is a non-negotiable component of any electronic OOS record. 21 CFR Part 11 requirements for electronic records apply to any OOS investigation conducted or stored in a computer system.

Common OOS Investigation Failures FDA Investigators Find

A review of FDA warning letters and 483 observations related to OOS investigations reveals patterns that appear year after year:

Phase I closure without a confirmed laboratory error. Teams that close investigations at Phase I because retesting passed, without identifying a specific laboratory error, are among the most commonly cited in warning letters. “No cause identified” is not an acceptable conclusion for Phase I closure.

Inadequate documentation of the investigation timeline. Records that cannot demonstrate a contemporaneous, real-time documentation sequence raise data integrity concerns. Backdated investigation notes, records reconstructed after the fact, and investigation documents with implausible completion timelines have triggered enforcement actions.

Retesting without quality unit oversight. Retesting conducted without a documented protocol approved by the quality unit, or retesting results used to override the original OOS without explanation, are consistently cited as cGMP violations.

Lack of connection between OOS results and CAPA. Investigations that identify a root cause but do not generate a CAPA leave the systemic condition unaddressed. FDA investigators look for evidence that recurring OOS results in the same category have triggered a systemic corrective action, not just individual batch investigations.

OOS results not shared with contract partners. When a CMO or contract laboratory produces an OOS result and does not promptly notify the sponsor company, or when the sponsor company’s quality agreement does not define notification requirements, the investigation record at the sponsor is often incomplete. The 2022 guidance addresses this expectation explicitly.

How a Modern eQMS Manages OOS Investigations

The OOS investigation process involves multiple parallel workflows that are difficult to manage reliably without a system that enforces structure: a laboratory investigation record, a production investigation record, a retesting protocol, a CAPA, a batch disposition decision, and a final closure review. Managing these across paper forms, email chains, or disconnected spreadsheets creates the exact documentation gaps that generate inspection findings.

Cloudtheapp’s Out of Specification application provides a structured, validated workflow for the complete OOS investigation lifecycle. When a result is flagged, the system opens an investigation record with a defined scope checklist. Phase I is completed within the record, with required fields for analyst identification, instrument records, and preliminary conclusions. If Phase I does not identify a confirmed error, the system automatically opens Phase II and routes it to the quality unit for expanded investigation. The investigation record captures all actions with timestamped, audit-trail-controlled documentation throughout.

Retesting, if required, is initiated directly from the OOS record and linked to the test results. The batch disposition decision is recorded within the same record with a required rationale field. If a CAPA is opened, it links directly to the OOS investigation record so the connection between the event and the corrective action is permanently documented.

When FDA investigators request OOS investigation records, Cloudtheapp customers can pull complete, current, and auditable investigation packages within minutes. That capability changes the inspection experience fundamentally.

Build OOS Readiness Into Your Quality System

The companies that manage OOS results most effectively are not the ones that rarely produce OOS findings. Anomalous results are inherent to laboratory testing at the volumes regulated companies operate. The differentiating factor is whether the system surrounding those results is structured enough to investigate, document, and resolve them consistently, every time, without relying on individual knowledge or manual coordination.

If your current quality system manages OOS investigations through spreadsheets, email approvals, or disconnected document templates, the investigation record that results is difficult to reconstruct and harder to defend. The question is not whether an OOS result will occur. The question is whether your system is ready to handle it when it does.

Cloudtheapp is an AI-powered, no-code eQMS platform built for regulated industries. The Out of Specification application is part of a fully validated platform that connects OOS investigations directly to lab testing, CAPA, and batch records. Request a demo at cloudtheapp.com to see how Cloudtheapp manages OOS workflows from initial detection through final closure and CAPA completion.

This post created by and appeared first on Cloudtheapp