Pharmaceutical Water Results: TOC and Conductivity Do Not Replace Microbial Evidence

Executive Summary: Why Chemistry Cannot Answer the Microbiology Question

When a quality control manager is asked by a customer or regulatory auditor why routine microbial sampling is still needed on a purified water (PW) or water for injection (WFI) loop that has shown flat, in-spec online total organic carbon (TOC) and conductivity for months, the short answer is: in-spec chemistry does not close the microbiological question. TOC and conductivity answer a chemical cleanliness question, and they answer it well: conductivity tracks dissolved ionic content, and TOC tracks oxidizable organic carbon against the compendial 500 µg C/L target limit response. Neither test is designed to count viable organisms, detect biofilm at an outlet or measure endotoxin, and none of the regulatory or compendial sources reviewed here treats them as microbial evidence.

FDA's 1993 inspection guide to high-purity water systems—historical, non-binding reference material, but still the plainest regulator statement of the mechanism—puts it directly: “Because of the similar ionic quality of distilled and deionized water, conductivity meters cannot be used to monitor microbiological quality.” The same guide explains that biofilm continuously sloughs off organisms, so contamination is not uniformly distributed and a single sample may not be representative. Separately, USP's water FAQ explains that the Purified Water and Water for Injection monographs contain no microbial requirements because these waters serve many different uses; microbial guidance sits in informational chapter <1231>, and each user sets fitness-for-use levels for its own products. Chemistry and microbiology therefore run on different tests, different limit logic and, in the EU sterile framework, different approval gates.

What TOC and Conductivity Actually Measure (and What the Monographs Say)

To understand why chemical parameters cannot substitute for microbiological surveillance, one must examine what USP <643> (Total Organic Carbon) and USP <645> (Water Conductivity) physically measure. These procedures represent standardized attribute assays designed to demonstrate broad chemical control over water purification and distribution systems.

Under USP <643>, total organic carbon is determined by oxidizing dissolved organic molecules to carbon dioxide and measuring the carbon dioxide produced. The compendial limit is expressed as a target limit response equivalent to 500 µg of carbon per liter (0.50 mg/L), and system suitability compares the instrument's response to an easily oxidized reference (sucrose) with its response to a hard-to-oxidize one (1,4-benzoquinone). TOC is a sensitive indicator of organic contamination in the bulk water. It is a bulk chemical measurement, though, not a microbial assay: a result well under the target response says that total oxidizable carbon is low, and nothing about whether viable organisms are present at a particular outlet, hose or dead leg.

Similarly, USP <645> provides a staged measure of dissolved ionic content. Because conductivity varies strongly with temperature and with absorbed carbon dioxide, the chapter uses three stages:

• Stage 1 (designed for online use): Conductivity is measured without temperature compensation and compared with a temperature-dependent limit table. If the reading is at or below the table limit for the measured temperature, the water passes.

• Stage 2 (equilibrated sample): If Stage 1 is not met, a sample is equilibrated with atmospheric carbon dioxide at a controlled temperature and compared with a single Stage 2 limit. USP's FAQ confirms that offline laboratory testing may begin directly at Stage 2 rather than Stage 1.

• Stage 3 (pH-dependent limit): If Stage 2 is not met, conductivity is evaluated against a pH-dependent limit. Failing Stage 3 means the water does not meet the conductivity requirement and points to ionic contamination.

Both <643> and <645> allow online and offline testing. Online analyzers give continuous assurance that purification steps—reverse osmosis, electrodeionization, distillation—are performing chemically as designed. Passing these tests confirms that dissolved ions and bulk oxidizable organics are controlled; it does not demonstrate that the loop is microbiologically sound. Read numeric limits and table values from the official USP–NF text rather than circulating PDF copies, some of which reproduce superseded chapter versions.

Why Compendial Monographs Deliberately Stop at Chemistry

A frequent point of confusion during external audits is the absence of numerical bioburden limits in the USP monographs for Purified Water and Water for Injection. That silence is sometimes read as meaning microbial testing is optional once monograph chemistry passes. It is a deliberate compendial design choice, not implicit coverage by the chemistry tests.

USP's water FAQ explains that microbial requirements were not included in the bulk water monographs because of the various uses of these waters. Purified Water may be used to granulate oral solids, formulate topical products or rinse equipment; Water for Injection may be used in parenteral formulation and final rinsing. A single monograph number strict enough for the most sensitive use would impose meaningless burden on the others, while one loose enough for the least sensitive use would be inadequate for the rest. FDA's 1993 guide made the same point from the inspection side: the quality control department should assess each product made with the water and set microbial action limits based on the most microbially sensitive product.

Microbial control therefore lives in the informational chapter USP <1231> (Water for Pharmaceutical Purposes) and in each manufacturer's own fitness-for-use levels, scaled to the products the water actually serves. When a customer auditor asks for “the spec,” the defensible answer names two documents: the monograph for chemistry and the site's microbial alert/action procedure for microbiology.

The Microbiology Track: Alert Levels, Action Levels, and 'Unfit for Use'

In contrast to the pass/fail attribute tests in the monographs, pharmaceutical water microbiology is managed through process-control levels: alert levels and action levels. FDA's 1993 guide is explicit that none of the limits for water are pass/fail limits; all are action limits. Understanding what each level obliges a site to do is what turns a monitoring program into an audit-ready one.

An alert level is an early-warning signal that the system may be drifting from its normal operating state. EU GMP Annex 1 (paragraph 6.13) says alert levels should be based on initial qualification data and periodically reassessed using re-qualification, routine monitoring and investigation data. WHO TRS 1033 Annex 3 adds that alert and action levels should be based on historical data and that results should be trended statistically at defined intervals (for example monthly, quarterly and annually) against defined control limits such as 3 sigma. An alert excursion does not by itself mean the water is unusable; Annex 1 (6.14) expects it to be documented, reviewed and investigated to decide whether it is an isolated event or an adverse trend.

An action level marks the point at which the system can no longer be shown to be under control. USP's FAQ gives the general <1231> action levels as:

• Purified Water action level: generally 100 CFU/mL. FDA's guide treats any purified-water action limit above 100 CFU/mL as unacceptable and notes that products without an effective preservative system, such as some antacids, may need a tighter limit.

• Water for Injection action level: generally 10 CFU/100 mL (equivalent to 0.1 CFU/mL). FDA's guide adds that WFI sample volumes below 100 mL are unacceptable and that 100–300 mL is preferred.

Microbial results are only as meaningful as the sampling behind them. Point-of-use samples should be drawn the way manufacturing draws the water. FDA's guide states that if a hose is usually attached, the sample should be taken at the end of the hose, and if the procedure calls for flushing before use, the sample is taken after the flush. WHO TRS 1033 Annex 3 (12.5) says samples should be taken using the same methodology as production procedures, for example using a hose with a suitable flushing and drainage procedure. Annex 1 (6.13) adds that sampling programmes should cover all outlets and points of use at a specified interval, consider worst-case locations, and include at least one representative sample every day of the water used for manufacturing. A sample drawn from a clean valve while production dispenses through an unsampled hose produces flattering data—exactly the gap FDA cited in the warning letter discussed below.

Why Passing Chemistry Does Not Predict Bioburden: Three Failure Modes

Three mechanisms explain why chemical instrument readings and microbial contamination can diverge, and why a long run of flat chemistry trends is not a substitute for counts:

1. Ionic Invisibility of Bacterial Cells

Conductivity responds to dissolved mobile ions such as sodium, chloride and bicarbonate. Microorganisms are intact cells, not dissolved ions, so a conductivity reading is not designed to register them. That is the mechanism behind FDA's statement that, because distilled and deionized water have similar ionic quality, conductivity meters cannot be used to monitor microbiological quality. A conductivity trend can show that a deionization or reverse osmosis stage is failing; it cannot show whether organisms are growing downstream of it.

2. Sessile Biofilm Architecture and Intermittent Sloughing

In distribution systems, many organisms are not free-floating in the water column but attached to pipe and tank walls as biofilm—a risk Annex 1 (6.9) addresses by calling for turbulent flow to minimize microbial adhesion and subsequent biofilm formation. FDA's guide explains that biofilm continuously sloughs off organisms, so contamination is not uniformly distributed and a sample may not be representative; it notes that a count of 10 CFU/mL in one sample and 100 or even 1,000 CFU/mL in a subsequent sample would not be unrealistic. An online TOC or conductivity sensor on the circulation return reports on the bulk loop; it does not see a colonized outlet, hose or dead leg.

3. Bioburden versus Bacterial Endotoxin Divergence

In WFI systems, endotoxin testing is sometimes treated as a surrogate for bioburden. FDA's guide rejects that shortcut: because WFI can pass the LAL endotoxin test and still fail the microbial action limit, it is important to monitor WFI systems for both endotoxins and microorganisms. The divergence has a biological basis. Bacterial endotoxin is lipopolysaccharide from the outer membrane of Gram-negative bacteria, so Gram-positive contamination does not register in an endotoxin assay, while endotoxin from killed Gram-negative cells can remain after a viable count has fallen. Bioburden, endotoxin and chemistry are three different questions; none answers the others.

EU GMP Annex 1: Continuous Chemistry and the Two-Gate Release Framework

EudraLex Volume 4, Annex 1 (published August 2022) is the EU GMP text for sterile medicinal products. Its water-system paragraphs pair continuous chemical monitoring with ongoing microbial control rather than letting one replace the other:

• Paragraph 6.10 — WFI production and storage: WFI should be produced by distillation or by a purification process equivalent to distillation (which may include reverse osmosis coupled with techniques such as electrodeionization, ultrafiltration or nanofiltration), and stored and distributed to minimize microbial growth—for example, by constant circulation above 70 °C.

• Paragraph 6.13 — Ongoing chemical and microbial monitoring: Regular ongoing chemical and microbial monitoring should be performed to ensure the water continues to meet compendial expectations. Alert levels are based on qualification data and periodically reassessed, and sample plans cover all outlets and points of use with at least one representative sample every day of the water used for manufacturing.

• Paragraph 6.14 — Excursion management: Alert-level excursions should be documented, reviewed and investigated to establish whether they are isolated events or adverse trends; each action-limit excursion should be investigated for probable root cause and potential impact on products and processes.

• Paragraph 6.15 — Continuous chemistry: WFI systems should include continuous monitoring such as TOC and conductivity, because these may give a better indication of overall system performance than discrete sampling. Sensor locations should be based on risk.

• Paragraphs 6.16–6.17 — Pure steam: Pure steam generators should produce steam meeting defined chemical and endotoxin levels. Where pure steam directly sterilizes materials or product-contact surfaces, its condensate should meet the current WFI monograph, and Annex 1 states that microbial testing is not mandatory for steam condensate.

Read together, 6.13 and 6.15 answer the argument that real-time chemistry makes microbial sampling redundant. The same annex that prefers continuous TOC and conductivity as a performance indicator also requires ongoing microbial monitoring with daily representative samples. It treats the two tracks as complements, not substitutes.

The Split-Gate Structure of Paragraph 6.12

Paragraph 6.12 is the most operational water requirement in Annex 1. It calls for sterilization, disinfection or regeneration on a predetermined schedule and as a remedial action after out-of-limit results, a validated rinse after chemical disinfection, and testing after the intervention. The results then feed two separate decisions:

flowchart TD
    A["Water System Disinfection or Regeneration"] --> R["Validated Rinse After Chemical Disinfection"]
    R --> B["Post-Intervention Testing"]

    subgraph Gate_1 ["Gate 1: Water System Return to Use"]
      B --> C["Chemical Testing (TOC and Conductivity)"]
      C --> D{"Chemical Results Approved?"}
      D -- "Yes" --> E["System Returned to Use"]
      D -- "No" --> F["System Stays Out of Use and Is Investigated"]
      E --> G["Manufacturing With the Water Resumes"]
    end

    subgraph Gate_2 ["Gate 2: Batch Certification and Release"]
      B --> H["Microbiological and Endotoxin Testing"]
      G --> I["Batches Made With the Water Await Results"]
      H --> J{"Microbial and Endotoxin Results Within Specification and Approved?"}
      J -- "Yes" --> M["Batches Can Be Considered for Certification or Release"]
      J -- "No" --> L["Excursion Investigation and Product Impact Assessment"]
      I --> M
      L --> N["Batch Disposition Decided by Quality Unit"]
    end

    style Gate_1 fill:#f8fafc,stroke:#0284c7,stroke-width:2px
    style Gate_2 fill:#f8fafc,stroke:#16a34a,stroke-width:2px

As illustrated above, Gate 1 allows the water system to return to use once chemical testing confirms that the system has been rinsed and chemistry is back within limits. Gate 2 is separate: batches manufactured with that water are not considered for certification or release until microbiological and endotoxin results are verified within specification and approved. Culture-based microbial methods take days to read, so the practical effect is that production can resume on chemistry while the batches made in that window wait for the microbiology.

Global Boundaries: WHO TRS 1033 and EMA Quality Guidelines

International guidance reinforces the split. The World Health Organization's good manufacturing practices for water for pharmaceutical use (WHO Technical Report Series No. 1033, Annex 3, 2021) starts from the way water differs from other ingredients:

Water is usually drawn from an on-demand system and is not subject to testing and batch or lot release before use, so its microbiological and chemical quality must be controlled throughout production, storage and distribution. WHO (12.3) expects flow, pressure and temperature to be monitored with online instruments—as well as conductivity and TOC where possible—with periodic offline testing to confirm the online results, and it allows other parameters to be monitored offline. Offline testing covering physical, chemical and microbiological attributes should follow a predetermined programme (12.4). WHO also states that bulk purified water and bulk WFI should each have alert and action limits for both chemical and microbiological purity, and that filtration should not be used in distribution loops or at take-off user points (10.3), so a point-of-use filter cannot stand in for a controlled loop. WHO guidance is not binding law, and national adoption varies.

In parallel, the European Medicines Agency's guideline on the quality of water for pharmaceutical use (EMA/CHMP/CVMP/QWP/496873/2018) sets the minimum water grade for each product type in marketing-authorisation dossiers:

• WFI where the product demands it: WFI (Ph. Eur. 0169) is the minimum grade for parenteral products—including haemofiltration, haemodiafiltration and peritoneal dialysis solutions—irrigation solutions and biologics. Sterile ophthalmic, nasal/ear and cutaneous preparations list Purified Water as the minimum, with the guideline noting that a risk assessment may call for higher-quality water.

• The final-rinse rule: In general, the final rinse of equipment, containers and closures should use the same quality of water as the related manufacturing stage. Passing PW chemistry does not make PW an acceptable final rinse where the product stage requires WFI.

• Membrane WFI: Since 1 April 2017, revised Ph. Eur. monograph 0169 has allowed WFI production by a purification process equivalent to distillation, such as reverse osmosis coupled with other appropriate techniques. The EMA guideline says it should be read together with an inspectors' Q&A on reverse osmosis, biofilms and control strategies—a reminder that the production method changes the biofilm risk profile, not the need for microbial evidence.

The US CGMP Hooks Regulators Actually Cite in Warning Letters

When FDA writes Form 483 observations or warning letters about water monitoring, the citations are to the CGMP regulations in 21 CFR Part 211 rather than to USP informational chapters. Four sections matter most:

• 21 CFR 211.160(b) — Scientifically sound controls: Laboratory controls must include scientifically sound and appropriate specifications, standards, sampling plans and test procedures designed to assure that components, in-process materials and drug products conform to appropriate standards. A monitoring plan that treats conductivity as evidence of microbial quality would be hard to defend as scientifically sound, and FDA has cited this section for unrepresentative water sampling (see the case below).

• 21 CFR 211.165(b) — Objectionable microorganisms: Requires appropriate laboratory testing, as necessary, of each batch of drug product required to be free of objectionable microorganisms. In-spec water chemistry provides no evidence about objectionable organisms, which FDA's 1993 guide defines as organisms that can cause infections when the product is used as directed or that can grow in the product.

• 21 CFR 211.165(c) and (e) — Written plans and validated methods: Sampling and testing plans must be described in written procedures that are followed, and the accuracy, sensitivity, specificity and reproducibility of test methods must be established and documented.

• 21 CFR 211.48 — Plumbing baseline: Potable water must meet EPA Primary Drinking Water Regulations (40 CFR part 141) and be supplied under continuous positive pressure in a plumbing system free of defects. It is a baseline for potable water, not a monitoring program for PW or WFI, and should not be over-read.

Enforcement Pattern: FDA Warning Letter 573233

One illustration is Warning Letter 573233 (US Pharmaceuticals, Inc., June 6, 2019), issued after a late-2018 inspection of a New Jersey manufacturer of over-the-counter (OTC) products. FDA cited 21 CFR 211.192 because the firm frequently lacked investigations into failing microbiological and chemical monitoring results from its purified water system, including total aerobic count results above its limit, too-numerous-to-count results, more than ten uninvestigated chemical testing failures, and TOC exceedances in water lots used to manufacture OTC products released for distribution. Microbiological deviations dated back to 2016.

Two details map directly onto this article's question. First, the firm argued that conforming finished-product microbial testing showed its batches were acceptable; FDA responded that microbial testing cannot be relied upon as the sole justification to release drug product batches, because contamination is not uniformly distributed. Second, FDA cited 21 CFR 211.160(b) because samples were not representative: they were collected following system sanitization, a timing FDA said gave no meaningful picture of water used through the week, and directly from the system valve without the discharge hose attached—although that hose, used to dispense water for manufacturing, carried visible black and brown residue. FDA asked for revised procedures requiring daily water samples for microbiological counts and microbial identification, a program covering both total counts and objectionable microbes, and three years of microbial monitoring results. These were firm-specific requests in a single enforcement action, not incidence data or a universal requirement.

Decision Matrix: Chemistry vs. Microbiology by Water Grade

The matrix below consolidates the sources above. It separates what each result can prove from what it cannot, by water grade, and labels whether the basis is a monograph test, an informational chapter or GMP guidance:

Two reading notes. Annex 1's two-gate rule is written for sterile manufacture under EU GMP; for non-sterile uses of PW, the site's own procedures—informed by FDA's expectation that action-limit excursions trigger a product-impact assessment—determine how microbial results feed batch decisions. And the numeric action levels are informational-chapter process controls, not monograph specifications, so a site may need tighter levels for sensitive products and should not present them to an auditor as compendial pass/fail limits.

Answering the Auditor: What Each Result Is Allowed to Prove

Back to the scenario: months of flat online TOC and conductivity, and an auditor asking why microbial sampling continues. A source-backed answer can be organized around five statements:

• Chemistry proves chemistry. In-spec TOC and conductivity show that the water meets the monograph's chemical attributes and that purification is performing. They are not designed to detect organisms, and FDA's guide says conductivity cannot be used to monitor microbiological quality.

• The monograph is silent on microbes by design. USP left microbial requirements out of the PW and WFI monographs because the waters' uses vary; microbial control sits in <1231> action levels and the site's own fitness-for-use procedure.

• Contamination is local and intermittent. Biofilm sloughs organisms unevenly, so point-of-use samples drawn the way production draws water—hose attached, after the specified flush—are the direct evidence of what reaches product.

• Regulators expect both tracks. Annex 1 requires ongoing chemical and microbial monitoring with daily representative samples, WHO expects chemical and microbiological alert and action limits, and in a 2019 warning letter FDA rejected finished-product microbial testing as the sole justification for releasing batches made with out-of-limit water.

• Each result controls a different decision. Chemistry can return a system to use; microbiology and endotoxin results gate batch certification after interventions in EU sterile manufacture, and action-level excursions trigger investigation under each framework reviewed here.

This is an operational framing for monitoring-program design and audit preparation, not a determination about any specific site's compliance or a basis for releasing product. Jurisdiction matters: USP chapters, US CGMP regulations, EU GMP guidance and WHO guidance carry different legal status in different markets.

Regulatory and Compendial Horizon: 2026–2027 Developments

Three developments affect endotoxin and water-chapter references. None changes the chemistry/microbiology split, and each has a different document status that should be confirmed before it is cited in a procedure:

1. FDA Pyrogen and Endotoxins Testing Q&A, Edition 2 (March 2026)

FDA announced a level 2 revision of its June 2012 final guidance (announcement content current as of March 18, 2026), reissuing it as Pyrogen and Endotoxins Testing: Questions and Answers, Edition 2. FDA describes the changes as minor updates; notably, certain references to Limulus Amebocyte Lysate (LAL) testing were removed to accommodate a broader scope of recombinant reagents, and sponsors using recombinant reagents should verify that the assay method is suitable for its intended purpose. FDA ties the change to USP <86>, Bacterial Endotoxins Test Using Recombinant Reagents, published in May 2025.

2. USP <86> Recombinant Endotoxin Testing and Water Monographs

USP's water FAQ states that USP <86> is official but was not yet listed in the water and steam monographs, and that a revision first proposed in Pharmacopeial Forum PF 52(1) would add it, with an anticipated official date of April 1, 2027. Until a revised monograph becomes official, the currently official monograph text—which references USP <85>—governs WFI and steam endotoxin testing; confirm the current status in USP–NF before changing a water method. The change would add a test option; it would not merge endotoxin and bioburden testing or alter the microbial action-level logic.

3. USP <1231> Revision Proposed in PF 51(4)

A revision to informational chapter USP <1231> was proposed in Pharmacopeial Forum PF 51(4), with public comments closing on September 30, 2025. As summarized by ECA Academy, the draft proposed:

• Online measurement preferred: Recommending online TOC and conductivity measurement over offline testing in the sampling section, and, where online measurement is not used, ensuring that water samples are representative of the water actually used.

• Stage 2 entry and TOC terminology: Clarifying that offline conductivity testing may start directly at Stage 2, and explaining the difference between the TOC target limit response and the actual limit, including the need for official USP reference standards.

• Nitrosamines: Adding a section on the low risk of nitrosamine contamination, noting that water treatment can remove nitrosamines and precursors but cannot guarantee it, so a control strategy and risk-based assessment may be appropriate. USP's FAQ separately links nitrosamines in PW and WFI to source-water concentration.

Sources

• U.S. Food and Drug Administration. Guide to Inspections of High Purity Water Systems (7/93). Office of Regulatory Affairs, Rockville, MD (Content current as of September 2016).

• United States Pharmacopeia. FAQs: Water for Pharmaceutical and Analytical Purposes. Official Compendial Clarifications on Monographs, <643>, <645>, and <1231>, Rockville, MD.

• European Commission. EudraLex Volume 4, EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use, Annex 1: Manufacture of Sterile Medicinal Products. Brussels, Belgium (August 2022).

• World Health Organization. WHO Technical Report Series No. 1033, Annex 3: WHO Good Manufacturing Practices: Water for Pharmaceutical Use. Geneva, Switzerland (2021).

• European Medicines Agency. Guideline on the Quality of Water for Pharmaceutical Use (EMA/CHMP/CVMP/QWP/496873/2018). Committee for Medicinal Products for Human Use, Amsterdam, Netherlands (2020).

• U.S. Food and Drug Administration. 21 CFR 211.165: Testing and Release for Distribution. Electronic Code of Federal Regulations, Title 21, Part 211 (Current as of September 2026).

• U.S. Food and Drug Administration. 21 CFR 211.160: General Requirements (Laboratory Controls). Electronic Code of Federal Regulations, Title 21, Part 211 (Current as of September 2026).

• U.S. Food and Drug Administration. 21 CFR 211.48: Plumbing. Electronic Code of Federal Regulations, Title 21, Part 211 (Current as of October 2026).

• U.S. Food and Drug Administration. FDA Clarifies Current Thinking on Pyrogen and Endotoxins Testing (Edition 2 Notice). Center for Drug Evaluation and Research, Silver Spring, MD (Content current as of March 18, 2026).

• U.S. Food and Drug Administration. Warning Letter: US Pharmaceuticals Inc. (CMS 573233). Division of Pharmaceutical Quality Operations I, Philadelphia, PA (June 6, 2019).

• ECA Academy. The USP Has Revised General Chapter <1231> on Pharmaceutical Water (PF 51(4) Draft Notice). European Compliance Academy, Heidelberg, Germany (July 2025).