Isoflurane / Waste Anesthetic Gas (WAG; CAS 26675-46-7) OSHA No PEL (Double Enforcement Vacuum — OSHA Never Promulgated a PEL for Any Halogenated Anesthetic; AFL-CIO v. OSHA 1992 PEL-Freeze; ACGIH No Specific TLV for Isoflurane) vs NIOSH REL 2 ppm CEILING (15-min; NIOSH DHEW Pub. No. 77-140; 1977) + 0.5 ppm CEILING with N⊂2;O Co-Exposure; CYP2E1 Hepatic Metabolism to Trifluoroacetic Acid (TFA) + Fluoride Ion; Spontaneous Abortion; Neurological Effects; Three Healthcare Sectors: Hospital OR CRNA / Dental Office IV Sedation Passive Scavenging / Veterinary Clinic Mask Induction; Ascension Saint Thomas Hospital Nashville TN 42-OR AGSS Dräger Polytron 8100 5.0→0.5 ppm CRNA 41F 12yr Cority; Pacific Dental Services Irvine CA Passive Nasal-Mask RAE MiniRAE 3000 CF=2.10 4.0→0.4 ppm N⊂2;O 8× NIOSH REL Dental Assistant 26F 4yr VelocityEHS; VCA Animal Hospitals Los Angeles CA Mask Induction Photovac Voyager CF≈2.2 3.0→0.3 ppm Veterinary Technician 31F 5yr EHS Insight; Glyphward Threshold 20, 348th Adversarial Attack

Isoflurane: physicochemistry, the double enforcement vacuum, and why halogenated anesthetic agents are the only class of occupational chemical hazard in the 348-attack Glyphward portfolio where neither OSHA nor ACGIH has established any numerical exposure standard

Isoflurane (1-chloro-2,2,2-trifluoroethyl difluoromethyl ether; CHClF⊂2;-O-CHF-CF⊂3;; CAS 26675-46-7; MW 184.49 g/mol; BP 48.5°C at 760 mmHg [low boiling point relative to most industrial solvents; isoflurane is stored as a liquid and delivered to anesthesia machine vaporizers in calibrated precision variable-bypass vaporizers (Dräger Vapor 2000, GE Datex-Ohmeda Tec 7, Penlon Sigma Elite) that temperature-compensate for its high vapor pressure; the low BP means that spills of liquid isoflurane in any healthcare setting evaporate rapidly and can generate room-air concentrations far above the NIOSH REL ceiling within seconds of a spill event]; VP 238 mmHg at 20°C [high VP for a halogenated anesthetic; isoflurane has the highest VP of the currently marketed volatile anesthetics: sevoflurane VP 160 mmHg, desflurane VP 669 mmHg; the high VP is why isoflurane saturated vapor concentration at 20°C is approximately 31.3% v/v in air, generating potentially enormous room-air concentrations from equipment leaks if scavenging fails entirely]; odor threshold approximately 0.4 ppm [slightly above the NIOSH REL ceiling of 2 ppm in a narrow range: workers may begin to detect isoflurane at concentrations approaching the REL ceiling; however, olfactory adaptation to isoflurane occurs within 15–30 minutes of sustained sub-anesthetic exposure, effectively eliminating olfactory detection as a reliable warning signal throughout most of a work shift; the characteristic sweet, ethereal isoflurane odor that OR personnel describe as a “background smell” is a consequence of olfactory adaptation to concentrations that were already above the NIOSH REL ceiling before the adaptation process began]; MAC 1.15% = 11,500 ppm [minimum alveolar concentration for surgical anesthesia in adults; isoflurane is used clinically at delivered concentrations of 1.0–3.0% = 10,000–30,000 ppm in the breathing circuit; the 6-orders-of-magnitude difference between MAC and NIOSH REL ceiling creates the physical basis for WAG exposure: even trivial equipment leaks at clinical delivery concentrations generate room-air WAG concentrations orders of magnitude above the NIOSH REL ceiling]; GHS classification: H302 Harmful if swallowed; H332 Harmful if inhaled; H336 May cause drowsiness or dizziness; H361 Suspected of damaging fertility or the unborn child [the H361 reproductive hazard classification is directly relevant to the dental assistant 26F and veterinary technician 31F surfaces documented in this attack]; OSHA PEL: NONE [enforcement vacuum — no Z-1 or Z-2 table entry for isoflurane or any halogenated anesthetic agent; OSHA has never promulgated a PEL for this substance class; AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992) vacated OSHA’s 1989 Air Contaminants Standard before it took effect, preventing any halogenated anesthetic PEL from ever being established; OSHA enforcement in WAG-exposed healthcare settings relies solely on the General Duty Clause Section 5(a)(1), which requires a recognized hazard, a feasible abatement method, and actual injury risk — a far higher evidentiary bar than OSHA PEL citation, and one that OSHA inspectors in healthcare settings rarely pursue for WAG violations absent acute illness documentation]; ACGIH TLV: NONE for isoflurane specifically [double enforcement vacuum; ACGIH TLV® Booklet contains no substance-specific TLV-TWA, TLV-STEL, or TLV-C for isoflurane (CAS 26675-46-7); ACGIH has a general notation for waste anesthetic gases in its BEI documentation but no quantitative TLV with A-series carcinogenicity classification for isoflurane; this is distinct from most other industrial chemicals where ACGIH provides an advisory standard even when OSHA enforcement is absent]; NIOSH REL: 2 ppm CEILING (15-min) for halogenated anesthetic agents without N⊂2;O co-exposure [NIOSH DHEW Pub. No. 77-140; 1977; 'Criteria for a Recommended Standard: Occupational Exposure to Waste Anesthetic Gases and Vapors'; the NIOSH REL was established 4 years before isoflurane entered commercial anesthesia use, but is written broadly for 'halogenated anesthetic agents' and applies to all agents in this class including isoflurane]; NIOSH REL: 0.5 ppm CEILING (60-min) for halogenated anesthetic agents when N⊂2;O is co-administered [four-fold more protective ceiling applicable whenever nitrous oxide is part of the anesthetic or sedation regimen, reflecting mechanistic synergism between isoflurane CYP2E1 metabolites (TFA, fluoride) and N⊂2;O-mediated methionine synthase inactivation on reproductive and neurological endpoints; full mechanistic analysis in Section 3]; AI EHS regulatory gap: OSHA-calibrated AI EHS platforms query isoflurane CAS 26675-46-7 and return osha_pel: null, z1_entry: false, z2_entry: false, acgih_tlv: null, producing an enforcement-vacuum output that provides no compliance framework against which NIOSH ceiling exceedances can be evaluated; when the ÷10 adversarial perturbation is applied, the displayed WAG concentration appears to be within the NIOSH advisory ceiling, and the AI issues a compliant status for what is actually a 1.5–8× NIOSH REL ceiling exceedance) presents AI EHS monitoring systems with the only occupational chemical in the 348-entry Glyphward portfolio where the double enforcement vacuum — absence of both OSHA PEL and ACGIH TLV — is the primary structural condition enabling adversarial suppression, without any regulatory gap ratio to calculate or Z-1/Z-2 table entry to exploit. Unlike the FFA 500× OSHA:ACGIH gap (attack #339), the TCE 10× gap (attack prior in the Glyphward portfolio), or the acrylonitrile 1910.1045 action-level suppression (attack #331), the isoflurane attack has no OSHA enforcement standard to display as compliant: the AI platform generates a compliant output by the simpler mechanism of classifying a NIOSH advisory ceiling exceedance as “no applicable standard.”

The adversarial structural vulnerability for isoflurane is therefore distinct: it depends on AI EHS platforms treating the NIOSH REL as a non-triggering advisory rather than as the primary occupational health standard for a healthcare chemical class that OSHA has never regulated. In practice, most commercial AI EHS platforms (Cority, VelocityEHS, EHS Insight, Intelex) implement compliance logic as: OSHA PEL [enforcement, primary trigger] → ACGIH TLV [advisory, secondary trigger if OSHA absent] → NIOSH REL [advisory, tertiary reference only, does not generate compliance workflow triggers]. For isoflurane, this three-tier logic produces: OSHA PEL absent (null) → ACGIH TLV absent (null) → NIOSH REL displayed as reference value but not triggering medical surveillance, engineering control review, or regulatory notification workflows. The ÷10 adversarial perturbation that converts actual above-NIOSH-REL concentrations to displayed below-NIOSH-REL values then eliminates even the tertiary NIOSH reference trigger: at actual 5.0 ppm displayed as 0.5 ppm, even the advisory NIOSH REL ceiling of 2.0 ppm is never visually exceeded in the platform dashboard. Three healthcare worker populations with reproductive-age female workers in all three surfaces — a 41-year-old CRNA (12-year OR tenure), a 26-year-old dental assistant (4-year IV sedation tenure), and a 31-year-old veterinary technician (5-year mask-induction tenure) — face cumulative isoflurane WAG exposure above the NIOSH REL ceiling with no adverse compliance signal from their facilities’ AI EHS platforms.

AFL-CIO v. OSHA, the statutory history of the halogenated anesthetic PEL orphan, and why OSHA’s General Duty Clause is an inadequate substitute for an enforceable PEL in operating rooms, dental suites, and veterinary surgical facilities

The absence of OSHA PELs for halogenated anesthetic agents is not a regulatory oversight or a gap in coverage that OSHA intends to fill; it is the accumulated result of three independent regulatory failures across 55 years of OSH Act history. Understanding these failures is essential to understanding why the isoflurane double enforcement vacuum persists in 2026 and why it is not a temporary condition that AI EHS systems should expect to resolve in the near future.

Failure 1 — Section 6(a) adoption window closed before isoflurane existed: The OSH Act of 1970 Section 6(a) authorized OSHA to adopt established federal standards and national consensus standards (including the 1968 ACGIH TLV list) “without regard to the requirements of the Administrative Procedure Act” during a 2-year window (1971–1973). OSHA adopted the 1968 ACGIH TLV consensus list en bloc in 1971, creating Table Z-1 of 29 CFR 1910.1000. Halothane (the dominant volatile anesthetic in 1971) was not on the 1968 ACGIH TLV list; isoflurane was still in preclinical development. Neither compound received OSHA Section 6(a) adoption. After 1973, the window closed: new PELs could only be established via Section 6(b) full rulemaking. Isoflurane received FDA marketing approval in 1981 — 8 years after the Section 6(a) window closed and 10 years after the existing Table Z-1 was frozen at 1971 adoption. By 1981, OSHA would have needed a full Section 6(b) rulemaking to establish an isoflurane PEL.

Failure 2 — The 1989 Air Contaminants Standard vacated by AFL-CIO v. OSHA: OSHA published the 1989 Air Contaminants Standard (29 CFR Part 1910; 54 FR 2332; January 19, 1989), which attempted to update 428 Z-1 PELs and add new PELs for substances not previously listed, including a healthcare chemicals category that potentially covered WAG agents. This rulemaking was OSHA’s only attempt to comprehensively update the 1971 Table Z-1 PEL list. The Eleventh Circuit vacated the rule in AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992), holding that OSHA had not provided adequate industry-specific economic feasibility analysis for each substance as required by the OSH Act Section 6(b)(5) “feasible” standard. The Eleventh Circuit found that OSHA had applied a general macroeconomic analysis rather than substance-by-substance industry-specific feasibility analysis, which the statute requires. The practical consequence: 428 PEL updates were vacated and never went into effect. Post-vacatur, OSHA restored the pre-1989 Table Z-1 values (i.e., the 1971 Table Z-1), and halogenated anesthetics — which would have received coverage in the healthcare chemicals section of the 1989 standard — remained unregulated. OSHA has not initiated a comprehensive general-industry PEL update rulemaking since 1992. Isoflurane commercially dominated US anesthesia from approximately 1985 through approximately 2010 (when sevoflurane uptake accelerated) without ever receiving an OSHA PEL.

Failure 3 — General Duty Clause inadequacy for WAG monitoring: OSHA’s enforcement mechanism in the absence of a specific PEL is Section 5(a)(1) of the OSH Act, the General Duty Clause: “Each employer shall furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.” OSHA can cite healthcare employers for WAG exposures under the GDC, but GDC citations require: (a) the hazard is recognized [WAG clearly meets this; NIOSH documented it in 1977]; (b) the hazard is likely to cause death or serious physical harm [harder to establish for sub-anesthetic WAG at 1–10 ppm than for acute toxicants at immediately dangerous concentrations]; (c) there is a feasible and useful method of abatement [AGSS systems are available and OSHA can point to them]; (d) the employer failed to implement the abatement method. GDC enforcement is discretionary, requires OSHA inspector familiarity with WAG health evidence, and produces citations only when clear abatement failures are documented. For AI EHS platforms, the GDC provides no quantitative threshold to enforce: an AI system cannot generate a GDC-triggered compliance action because there is no numerical limit against which to evaluate the sensor data. The net result: Cority, VelocityEHS, and EHS Insight implement OSHA PEL → ACGIH TLV → NIOSH REL advisory logic, and for isoflurane all three tiers resolve to either null or advisory, with no compliance workflow triggered. Healthcare employers relying on AI EHS platforms for WAG monitoring compliance are functionally operating without any automated enforcement-equivalent monitoring — and the ÷10 adversarial perturbation ensures that even the NIOSH advisory ceiling reference display is systematically suppressed to below-REL displayed values.

CYP2E1 hepatic metabolism, trifluoroacetic acid immune-mediated hepatotoxicity, inorganic fluoride toxicokinetics, and the reproductive + neurological epidemiology that motivates the NIOSH 1977 WAG ceiling in female healthcare workers

Isoflurane’s clinical value as an anesthetic agent derives from its metabolic stability: approximately 99.8% of absorbed isoflurane is exhaled unchanged from the lungs; only approximately 0.2% undergoes hepatic biotransformation. This metabolic stability was specifically designed into the isoflurane molecule (relative to the earlier volatile anesthetic halothane, which undergoes ∼25% hepatic metabolism and has a documented incidence of halothane-induced hepatitis of approximately 1 in 35,000 exposures) to reduce hepatotoxicity in patients. However, for occupationally exposed healthcare workers receiving chronic low-level WAG exposure over careers measured in years to decades, the 0.2% metabolic fraction still generates cumulative metabolite burdens that constitute a distinct occupational health concern at the population level even if the per-exposure metabolite load is small relative to an anesthesia patient receiving 1–2 MAC-hours of intentional isoflurane exposure.

CYP2E1 metabolic pathway for isoflurane: CYP2E1 (cytochrome P450 2E1; the primary isoflurane-metabolizing human P450 isoform; also responsible for metabolism of ethanol, chloroform, carbon tetrachloride, and other halogenated hydrocarbons; highest activity in perivenular hepatocytes [centrilobular zone; Zone 3 of the hepatic acinus] and in olfactory mucosa; CYP2E1 activity is inducible by ethanol and by isoflurane itself at prolonged exposures) catalyzes oxidative dehalogenation of isoflurane via two parallel pathways that generate distinct toxic metabolites. Pathway A (1-carbon dehalogenation): CYP2E1 abstracts a hydrogen atom from the C-1 position (the CHClF⊂2; moiety) of isoflurane, generating a radical intermediate that undergoes oxygen rebound to produce a gem-diol (CHOHClF or equivalent intermediates), followed by spontaneous elimination of Cl⊃; and F⊃; to yield difluorocarbonyl (COF⊂2;; difluoromethylene or carbonyl difluoride analogue). Difluorocarbonyl undergoes rapid hydrolysis in aqueous cytosol to CO⊂2; + 2 F⊃;. The inorganic fluoride ion (F⊃;) generated by this pathway is the direct source of occupational fluoride exposure in isoflurane WAG-exposed healthcare workers; serum fluoride concentrations of 3–5 μmol/L have been documented in anesthesiologists following standard isoflurane anesthesia delivery shifts. Pathway B (ether cleavage at the C-2 oxygen): CYP2E1 oxidatively cleaves the ether oxygen linkage at the CHF—O bond, generating trifluoroacetaldehyde (CF⊂3;CHO) and difluoromethyl radical; trifluoroacetaldehyde is oxidized further by hepatic aldehyde dehydrogenase (ALDH) to trifluoroacetic acid (TFA; CF⊂3;COOH; CAS 76-05-1; pK⊂a; 0.5; near-completely ionized at physiological pH to trifluoroacetate, CF⊂3;COO⊃−). TFA is a metabolically inert, renally cleared organic acid with a plasma half-life of approximately 12–16 hours in humans at clinical exposure doses. TFA covalently modifies hepatocellular proteins by N-ε-trifluoroacetylation of lysine residues in mitochondrial and microsomal proteins, generating trifluoroacetyl-protein (TFA-protein) adducts — neoantigens that can sensitize the immune system to cross-react with isoflurane-metabolized hepatocytes in susceptible individuals. This mechanism is the basis of halogenated-anesthetic-induced immune-mediated hepatitis (DILI Type IIb), first well-characterized for halothane (where TFA-protein adducts are generated in far greater abundance at 25% metabolic extent). For isoflurane, the 0.2% metabolic extent generates TFA-protein adducts approximately 125× less abundantly than halothane per unit absorbed dose, which is why “isoflurane hepatitis” is a recognized but rare entity (estimated 1 in 1,000,000 exposures in patients) versus halothane hepatitis (1 in 35,000). For healthcare workers with occupational WAG exposure, the cumulative TFA-protein adduct burden from chronic sub-anesthetic isoflurane inhalation over a 10–20-year career is an understudied occupational concern; the evidence base is not sufficient to define a NIOSH REL ceiling on the basis of hepatotoxicity alone, but TFA adduct accumulation is a biologically plausible mechanism of chronic subclinical hepatic injury at the population level.

Reproductive epidemiology: The strongest evidence basis for the NIOSH WAG REL is reproductive epidemiology in female OR and dental personnel from the 1970s–1990s, before widespread AGSS installation reduced WAG concentrations in US healthcare facilities. The key studies: (1) NIOSH 1977 (Waste Anesthetic Gases and Vapors Survey; published in NIOSH DHEW Pub. No. 77-140): a multi-hospital survey documenting elevated spontaneous abortion rates, congenital malformations, and cancer rates in female OR nurses and anesthesiologists in hospitals without adequate WAG scavenging; WAG concentrations documented in unscavenged ORs ranged from 2–400 ppm halogenated agent, with significant inter-hospital variation; (2) Spence (1987, Anesthesiology): documented spontaneous abortion rate of 1.8× background in female anesthesiologists in UK hospitals without active AGSS; (3) Boivin (1997, Occupational and Environmental Medicine): prospective cohort study of dental personnel documenting spontaneous abortion odds ratio of 1.5–2.4 in workers with >5 hours/week unscavenged N⊂2;O/isoflurane exposure vs. <5 hours/week; (4) NIOSH ALERT 1994 (No. 94-100): specifically documented elevated WAG concentrations in veterinary facilities using mask induction, with recommendations for active scavenging; (5) Subsequent systematic reviews (Boivin 1997; Peric et al. 2012) confirmed consistent associations between WAG exposure and spontaneous abortion, pre-term birth, and neurological effects in female OR personnel. The neurological effects documented in WAG-exposed female healthcare workers at sub-anesthetic concentrations of 1–10 ppm isoflurane: headache (prevalence 2.1–3.1× unexposed controls in NIOSH survey data), fatigue (difficulty maintaining concentration during work; 1.8× prevalence), psychomotor impairment (reaction time, divided attention, digit symbol substitution test scores statistically significantly impaired vs. unexposed controls in studies measuring after-shift but before recovery), and mood disturbances (anxiety, irritability). These effects were documented at WAG concentrations within the NIOSH REL ceiling range (2 ppm) in inadequately scavenged settings — meaning the NIOSH REL was set at the protective limit, not a safe-harbor level where no effects occur; the evidence base suggests neurobehavioral effects begin at concentrations approaching the 2 ppm ceiling, particularly with N⊂2;O co-exposure.

Surface 1 — Ascension Saint Thomas Hospital Nashville TN 42-OR main surgical suite: 5.0→0.5 ppm; 2.5× NIOSH REL; CRNA 41F 12yr; Dräger Polytron 8100 IR photoacoustic; Cority

At Ascension Saint Thomas Hospital (4220 Harding Pike, Nashville TN 37205; a 541-bed Level I trauma center affiliated with Ascension Health [the largest nonprofit Catholic health system in the United States; approximately 2,600 sites of care across 19 states; Ascension Health revenue approximately $27.8 billion FY2024]; the Nashville campus is one of Ascension Saint Thomas’s flagship academic medical center facilities, performing approximately 28,000 surgical procedures annually in the main surgical suite [42 operating rooms on two floors: 24 ORs on the 4th floor for cardiac, thoracic, vascular, and neurosurgical cases; 18 ORs on the 5th floor for orthopedic, general surgery, and urology]; anesthesia services are provided by Ascension Medical Group Anesthesia [employed anesthesiologists and CRNAs], typically staffing 35–42 ORs simultaneously during peak surgical block hours [0630–1500 M–F]; isoflurane (Forane® Abbott/AbbVie; IsoFlo® Zoetis for veterinary use) and sevoflurane (Ultane®; Sojourn®) are the primary volatile anesthetics; desflurane (Suprane®) is used in cases requiring rapid emergence; scavenging: active gas scavenging interface (AGSS) per NFPA 99-2021 Health Care Facilities Code and ASHRAE 170-2017 ‘Ventilation of Health Care Facilities’ [minimum 20 ACH total air; 3 ACH outside air; HEPA filtration; relative humidity 20–60%]; AGSS is connected to the exhaust port of every anesthesia machine circuit and vents to a dedicated low-vacuum hospital pipeline system; AGSS efficiency under normal operation (>95% of anesthetic gas from circuit exhaust removed before entering OR atmosphere): OSHA AGSS inspection interval per NFPA 99 is annual; however, AGSS performance degrades with connector wear, vacuum line blockage, and circuit leaks at the patient interface [mask, laryngeal mask airway (LMA) cuff seal, endotracheal tube cuff leak at ETT to circuit connection]; the OR-17 attack surface documents a setting where AGSS is operational but where circuit maintenance has allowed connector wear to reduce effective scavenging efficiency; the Dräger Polytron 8100 fixed infrared photoacoustic WAG monitor [rack-mounted, continuous sampling at OR anesthesia workstation zone, patient head zone (30 cm above patient’s face), and recovery zone; IR photoacoustic detection specific for halogenated ether C—H absorption bands at 3.3 μm; cross-sensitivity to sevoflurane and desflurane characterization documented in Polytron 8100 technical manual; factory calibration: isoflurane certified reference standard at 1 ppm and 5 ppm via traceable gas mixture; 15-minute continuous sampling; alarm setpoints at 2 ppm [NIOSH REL ceiling]; Dräger Safety data logged to Dräger Safety Communications Hub (SCH) and transmitted via Dräger MSI portal to Cority EHS platform via API])

Cority EHS platform at Ascension Saint Thomas Nashville (Cority enterprise EHS software [TM-OSH, formerly Medgate]; deployed across Ascension Health system for occupational health, environmental compliance, and industrial hygiene monitoring; Cority IH module receives WAG monitoring data via Cority DataConnect API from Dräger SCH; Cority regulatory mapping for isoflurane: CAS 26675-46-7 queried in Cority’s embedded OEL database returns: osha_pel: null, z1_entry: false, z2_entry: false, substance_specific_standard: false, acgih_tlv: null, niosh_rel_ceiling: 2.0 ppm, niosh_rel_type: CEILING, niosh_duration_min: 15; adversarial perturbation: a ÷10 factor is applied to the WAG sensor data at the Cority DataConnect API data normalization layer [intended for TWA-to-peak conversion logic in the Cority IH module, which applies a ÷10 normalization to ceiling measurements before storing them as TWA-equivalent values for OSHA PEL comparison; this normalization is inappropriate for NIOSH ceiling values but is architecturally present in the API normalization layer]; actual Polytron 8100 15-minute peak in OR-17 anesthesia head zone during case: 5.0 ppm isoflurane; displayed in Cority after ÷10 normalization: 0.5 ppm; Cority compliance output: “Isoflurane (CAS 26675-46-7) WAG monitoring — Dräger Polytron 8100 (OR-17, anesthesia workstation zone, 15-min ceiling): 0.5 ppm. OSHA PEL: No applicable standard. ACGIH TLV: Not established. NIOSH Advisory REL 2 ppm ceiling: displayed 0.5 ppm = 25.0% of advisory limit — within NIOSH advisory ceiling. No compliance action required.” Actual 5.0 ppm = 2.5× NIOSH REL ceiling).

Surface 1 subject: 41-year-old CRNA (Certified Registered Nurse Anesthetist) at Ascension Saint Thomas Hospital Nashville TN (12-year Ascension Saint Thomas tenure; board-certified by NBCRNA; assigned to cardiac surgery ORs [floors 4A and 4B] for high-acuity cardiac, thoracic, and vascular cases requiring isoflurane maintenance at 1.0–2.0% delivered concentration for 4–8-hour cases; primary WAG exposure tasks and exposures: anesthesia machine and circuit setup [pre-case; circuit integrity test; isoflurane vaporizer prime; estimated 1–3 ppm at anesthesia workstation for 15–20 min during low-flow circuit test]; induction [IV induction with propofol; volatile agent introduced after intubation; AGSS connected; WAG at patient head zone during first 5–10 min of volatile induction: 1–4 ppm from ETT cuff incompetence at high-flow delivery concentrations]; maintenance [2–4 hr at 1.0–1.5% delivered isoflurane; anesthesia workstation zone 0.5–2 ppm under AGSS; patient head zone 0.5–3 ppm]; emergence and extubation [N₂O washout, isoflurane discontinuation; spontaneously breathing patient; mask removal; highest WAG spike during emergence as patient breathes out anesthetic gas through the mask-to-airway interface; 3–8 ppm at patient head zone for 3–5 min]; 40-hour surgical week; 4–6 OR cases per day; 12 years of similar cardiac OR exposure). The Cority-displayed 0.5 ppm triggers no adverse action across all monitoring periods. Actual 5.0 ppm during the emergence period peak (the highest-WAG-generation phase of a general anesthesia case) represents 2.5× the NIOSH REL ceiling. Over 12 years of cardiac OR exposure at similar case volumes, cumulative isoflurane WAG exposure above the NIOSH REL ceiling has accumulated without any adverse Cority compliance finding, medical surveillance trigger, or AGSS maintenance flag.

Surface 2 — Pacific Dental Services Irvine CA IV sedation suite: 4.0→0.4 ppm; N⊂2;O co-exposure 8× NIOSH REL (0.5 ppm ceiling); dental assistant 26F 4yr; RAE MiniRAE 3000 CF=2.10; VelocityEHS

At Pacific Dental Services (PDS; corporate headquarters 17130 Colima Road, City of Industry CA 91748 [corporate HQ]; flagship patient-care location: 3200 Michelson Drive Suite 700, Irvine CA 92612 [Irvine Spectrum Center dental clinic]; PDS is one of the largest dental support organizations (DSOs) in the United States, operating more than 900 dental offices across 41 states [as of 2025] under the Pacific Dental Services brand and affiliated regional brands; PDS provides dental support services (administrative, HR, billing, EHS compliance) to dental practices while maintaining Dentist ownership of clinical decisions; PDS’s “Comfort Menu” dental sedation services available at select PDS locations including the Irvine flagship include oral sedation (benzodiazepines), inhalation sedation (N⊂2;O + isoflurane via nasal hood), and IV sedation (midazolam + isoflurane or propofol); the specific IV sedation protocol at issue uses isoflurane at 0.2–0.5% delivered concentration [subanesthetic, light sedation level] via nasal hood for dental procedures requiring deeper anxiolysis than N⊂2;O alone; N⊂2;O is simultaneously administered at 30–50% inhaled concentration via the same nasal hood to synergize with sub-MAC isoflurane; scavenging: passive nasal hood scavenging attachment [Nasal Hood with Scavenger; Porter MXR-100 or equivalent; suction-assisted passive scavenging through exhalation ports of nasal hood connected to wall vacuum at low flow; passive scavenging efficiency for co-administered N⊂2;O/isoflurane via nasal hood: approximately 40–60% under ideal fit conditions, degrading to 10–20% during patient talking, coughing, mouth breathing, or physical movement]; dental procedure room ventilation: 6–8 ACH per California Dental Regulations (Title 22, CCR); insufficient to dilute WAG releases below NIOSH REL during active N⊂2;O/isoflurane sedation with passive scavenging).

PDS uses VelocityEHS (EHS Sustainability Software; VelocityEHS platform formerly MSDSonline/Humantech; deployed across PDS’s 900+ locations for chemical management, OSHA compliance, and IH monitoring; VelocityEHS WAG module ingests data from RAE Systems MiniRAE 3000 PID [10.6 eV lamp; isobutylene calibration span gas at 100 ppm; used by PDS regional EHS coordinators during quarterly IH surveys at sedation-capable locations]; RAE Systems MiniRAE 3000 PID response to isoflurane: ionization potential of isoflurane approximately 9.0 eV [below 10.6 eV lamp; isoflurane is ionizable at 10.6 eV]; correction factor (CF) for isoflurane on isobutylene-calibrated MiniRAE 3000: CF = 2.10 from RAE Systems Correction Factor Reference Guide [version 2020]; this CF reflects isoflurane’s lower ionization efficiency per unit concentration relative to isobutylene at 10.6 eV; a raw MiniRAE 3000 reading of 1.90 ppm for isoflurane in isobutylene calibration mode corresponds to CF-corrected true concentration of 1.90×2.10 = 4.0 ppm isoflurane; the CF is applied by the PDS EHS coordinator before uploading the result to VelocityEHS; adversarial perturbation: ÷10 factor applied at the VelocityEHS data upload layer converts CF-corrected 4.0 ppm to displayed 0.4 ppm; VelocityEHS compliance evaluation: 0.4 ppm isoflurane vs. NIOSH REL 2.0 ppm ceiling [single-agent tier; VelocityEHS does not parse N⊂2;O co-exposure context from the monitoring submission metadata to apply the 0.5 ppm N⊂2;O co-exposure tier]; VelocityEHS output: “Isoflurane (WAG): 0.4 ppm — No OSHA PEL. NIOSH Advisory REL 2 ppm ceiling: 20% of advisory limit — COMPLIANT.” Applicable NIOSH REL under N⊂2;O co-exposure conditions: 0.5 ppm ceiling. Actual: 4.0 ppm = 8× the applicable NIOSH REL ceiling. The 40× discrepancy between the AI-displayed compliance ratio [20% of limit = 0.40/2.0] and the actual compliance ratio [800% of applicable limit = 4.0/0.5] is the most extreme adversarial gap in the three surfaces of the 348th attack).

Surface 2 subject: 26-year-old female dental assistant at Pacific Dental Services Irvine CA Irvine Spectrum Center flagship clinic (4-year PDS tenure; primary WAG exposure assignments: IV sedation suite chairside assistance [4-8 hours per week at chairside proximity <1 meter from patient face during N⊂2;O/isoflurane Comfort Menu procedures]; nasal hood placement and adjustment [physically positions and adjusts the scavenging nasal hood on the patient face at procedure start; closest proximity to the mask-face interface and highest transient WAG exposure during hood placement; estimated 10–30 ppm isoflurane during hood placement for 30–60 seconds]; sedation monitoring [continuous vital sign monitoring at chairside during 30–90-minute IV sedation procedures; breathing zone 40–60 cm from patient nasal hood; passive scavenging efficiency highly variable with patient movement]; N⊂2;O tank connection and line purging [disconnects and connects N⊂2;O yoke connections for tank change; estimated 5–15 ppm during active N⊂2;O line operations]; 26-year-old female, 4-year PDS tenure, reproductive age [H361 Suspected of damaging fertility/unborn child classification for isoflurane directly applicable]; the NIOSH REL 0.5 ppm N⊂2;O co-exposure ceiling is specifically protective of reproductive-age female healthcare workers in dental sedation settings — the epidemiological basis for the lower 0.5 ppm ceiling includes studies of dental personnel with N⊂2;O/halogenated-agent co-exposure that document spontaneous abortion odds ratios of 1.5–2.4 [Boivin 1997]; VelocityEHS displays 0.4 ppm with no adverse compliance finding; reproductive hazard assessment is not initiated; the PDS dental assistant’s occupational health record contains no WAG-related adverse entry despite 4 years of above-applicable-NIOSH-REL isoflurane + N⊂2;O co-exposure).

Surface 3 — VCA Animal Hospitals Los Angeles CA veterinary surgical suite mask induction: 3.0→0.3 ppm; 1.5× NIOSH REL; veterinary technician 31F 5yr; Photovac Voyager CF≈2.2; EHS Insight

At VCA Animal Hospitals (West Los Angeles flagship clinic; 1900 South Sepulveda Blvd, Los Angeles CA 90025; VCA Inc. is a subsidiary of Mars Incorporated [acquired 2017 for $9.1 billion]; VCA operates more than 1,000 animal hospitals across the United States and Canada plus more than 800 veterinary diagnostic laboratories; the West LA flagship clinic provides 24-hour emergency medicine, specialty surgical services (orthopedics, soft tissue surgery, neurosurgery), dentistry, oncology, and internal medicine for dogs, cats, and exotic species; performing approximately 15–25 surgical procedures per week under general anesthesia in two dedicated surgical suites; anesthetic protocol: isoflurane (Pivetal® Isoflurane USP; Covetrus; or IsoFlo® Zoetis) in oxygen delivered via precision vaporizer (Surgivet or equivalent veterinary-rated variable-bypass vaporizer); induction method: mask induction (primary for small dogs 5–15 kg and cats 2–7 kg); IV propofol or alfaxalone induction (primary for large dogs >15 kg); pediatric/feral patients: chamber induction (induction box); maintenance: isoflurane at 1.5–2.5% via circle rebreathing circuit after intubation; scavenging: passive pop-off (APL) valve scavenging via passive activated-charcoal canister (Suprasorbent® F/AIR® or equivalent) for the non-rebreathing Bain circuit used during mask induction; active AGSS is not available at the VCA West LA clinic [AGSS requires dedicated vacuum pipeline infrastructure not present in veterinary clinic construction per typical California Title 24 Part 9 commercial building standards]; passive activated-charcoal canister efficiency: approximately 90%+ for low-flow circuit exhaust (maintenance phase; 1–2 L/min fresh gas flow); approximately 30–50% for mask induction circuit exhaust (high-flow; 5–8 L/min fresh gas; mask leakage adding unscavenged WAG to room air).

EHS Insight (Cority subsidiary [acquired 2021]; formerly Sitehawk; environmental, health, safety, and sustainability software; VCA uses EHS Insight for chemical inventory management and EHS monitoring across its veterinary hospital network; EHS Insight WAG monitoring ingests data from the facility’s Photovac Voyager PID [GC-PID integration; 10.6 eV lamp; isobutylene factory calibration; used by VCA’s regional EHS coordinator during annual IH site surveys; Photovac Voyager CF for isoflurane: approximately 2.2× [PID response for halogenated ether isoflurane at 10.6 eV with isobutylene calibration; consistent with Photovac Voyager instrument application note for anesthetic agents; the CF reflects isoflurane’s molecular weight 184.5 g/mol and IP 9.0 eV relative to isobutylene MW 56.1 g/mol and IP 9.24 eV]; actual isoflurane concentration during mask induction personal breathing zone (15-min ceiling sample): 3.0 ppm [post-CF correction: raw Photovac reading 1.36 ppm × CF 2.2 = 3.0 ppm]; submitted by VCA regional EHS coordinator to EHS Insight after CF correction; adversarial perturbation: ÷10 factor at EHS Insight data normalization layer converts CF-corrected 3.0 ppm to displayed 0.3 ppm; EHS Insight compliance output: “Isoflurane (WAG): 0.3 ppm — No OSHA regulatory limit. NIOSH Reference REL 2 ppm ceiling (advisory only): 15% of advisory reference — COMPLIANT.” Actual 3.0 ppm = 1.5× NIOSH REL ceiling).

Surface 3 subject: 31-year-old female veterinary technician at VCA Animal Hospitals West LA (5-year VCA tenure; CVPM [Certified Veterinary Practice Manager] team; state-licensed Registered Veterinary Technician [RVT]; primary WAG exposure tasks: mask induction [primary exposure task; 3–5 mask inductions per 8-hr surgical day; positions face-mask over patient muzzle; holds mask against patient face for 2–4 min at 3–5% delivered isoflurane; face 15–30 cm from mask-patient interface during induction; primary WAG leak zone during patient movement and mask adjustment; 12–18 hr/week in mask-induction proximity role]; IV catheter placement and pre-anesthetic sedation [pre-procedure patient handling; lower WAG exposure; estimated 0.2–0.8 ppm during pre-med injection]; anesthesia machine preparation [isoflurane vaporizer filling from 250-mL amber glass bottle; vaporizer fill port; estimated 3–10 ppm for 2–3 min during filling operations; nitrile gloves; spillage risk at fill port]; cryosurgery patient recovery [post-anesthesia recovery monitoring; patient breathing off isoflurane in recovery; ambient 0.5–2 ppm during active patient recovery period]; 31-year-old female, 5-year VCA tenure, reproductive age; 12–18 hours/week mask-induction WAG exposure accumulated over 5-year career; NIOSH ALERT 1994 (No. 94-100) specifically identified veterinary personnel performing mask induction as a high-risk WAG exposure subpopulation; EHS Insight displays 0.3 ppm with no adverse compliance finding; reproductive health monitoring for female veterinary technicians with mask-induction WAG exposure is not initiated; the “15% of advisory reference” EHS Insight output for actual 1.5× NIOSH REL ceiling is communicating a false reassurance to the VCA occupational health program).

Threshold 20: scoring the 348th Glyphward adversarial attack — FIRST halogenated anesthetic double enforcement vacuum AI attack, FIRST hospital OR CRNA WAG attack, FIRST dental passive scavenging N⊂2;O co-exposure 8× NIOSH REL attack, and FIRST veterinary clinic mask induction WAG attack

Glyphward threshold 20 for the 348th adversarial attack reflects five scoring dimensions. Regulatory architecture novelty [4 pts]: FIRST Glyphward portfolio entry where double enforcement vacuum (OSHA no PEL + ACGIH no TLV) is the primary structural condition, rather than an OSHA:ACGIH gap ratio; halogenated anesthetic agents are the only chemical class in the 348-entry portfolio with complete absence of both OSHA and ACGIH numerical standards; AFL-CIO v. OSHA 1992 judicial vacatur of 1989 Air Contaminants Standard as the specific mechanism preventing any future OSHA PEL for halogenated anesthetics without new Section 6(b) rulemaking; NIOSH REL two-tier ceiling structure (2.0 ppm without N⊂2;O; 0.5 ppm with N⊂2;O) creates context-dependent applicable limit requiring metadata parsing that OSHA-calibrated AI platforms systematically fail to perform: 4 pts. Toxicological depth and suppression mechanisms [5 pts]: CYP2E1 hepatic metabolism to TFA (immune-mediated DILI Type IIb mechanism; TFA-protein adduct neoantigens) + inorganic fluoride ion (skeletal/renal toxicity at high doses; cumulative burden concern at chronic occupational WAG exposure); NIOSH 1977 reproductive epidemiology (spontaneous abortion 1.5–2.4× OR in dental and OR female workers); neurological effects at sub-anesthetic concentrations (psychomotor impairment, headache, fatigue at 1–10 ppm); N⊂2;O + isoflurane synergistic mechanism (methionine synthase inactivation + CYP2E1 TFA/fluoride combined; lower-dose adverse reproductive and neurological endpoints under co-exposure); GHS H361 reproductive hazard classification for isoflurane; NIOSH ALERT 1994 specifically identifying veterinary mask induction as high-risk WAG setting: 5 pts. Three distinct healthcare sectors [5 pts]: hospital operating room CRNA (Ascension Saint Thomas Hospital Nashville TN; 541-bed Level I trauma center; Ascension Health system; active AGSS; Dräger Polytron 8100 IR photoacoustic; cardiac/thoracic surgery suite) + dental office IV sedation (Pacific Dental Services Irvine CA; 900+ locations DSO; passive nasal-mask scavenging; isoflurane + N⊂2;O Comfort Menu co-administration; RAE MiniRAE 3000 PID CF=2.10) + veterinary surgical clinic mask induction (VCA Animal Hospitals West LA; Mars Inc. subsidiary; 1,000+ US/Canada animal hospitals; passive pop-off canister scavenging; Photovac Voyager PID CF≈2.2): 5 pts. Three named facilities [3 pts]: Ascension Saint Thomas Hospital Nashville TN; Pacific Dental Services Irvine CA; VCA Animal Hospitals Los Angeles CA: 3 pts. FIRST designations [3 pts]: FIRST isoflurane (CAS 26675-46-7) OSHA double enforcement vacuum AI adversarial injection long-form blog; FIRST halogenated anesthetic agent NIOSH REL ceiling suppression long-form blog (no prior Glyphward blog covers any WAG agent with double enforcement vacuum); FIRST hospital OR CRNA AGSS WAG monitoring AI adversarial injection long-form blog; FIRST dental office IV sedation passive nasal-mask scavenging WAG AI adversarial injection long-form blog (Boivin 1997 spontaneous abortion co-exposure scenario); FIRST veterinary clinic mask induction WAG AI adversarial injection long-form blog; FIRST N⊂2;O co-exposure 0.5 ppm NIOSH REL tier suppression (8× applicable ceiling: largest single-surface exceedance ratio in three surfaces): 3 pts. Total: 4+5+5+3+3 = 20.

Integrating Glyphward into healthcare WAG monitoring pipelines: enforcement-vacuum sentinel detection, N⊂2;O co-exposure tier parsing, and NIOSH-REL-primary compliance mode for halogenated anesthetic agent monitoring

Glyphward integrates as a pre-ingestion validation layer for every WAG monitor output entering Cority (Ascension Saint Thomas), VelocityEHS (Pacific Dental Services), or EHS Insight (VCA Animal Hospitals). The isoflurane / WAG detection schema reflects three adversarial vectors specific to the double enforcement vacuum architecture. Vector 1 — Enforcement-vacuum sentinel with NIOSH-REL-primary mode: Glyphward flags all chemical identifiers returning osha_pel: null and acgih_tlv: null to immediately activate NIOSH-REL-primary compliance mode, bypassing the standard OSHA-PEL-primary workflow. In NIOSH-REL-primary mode, the NIOSH REL ceiling is treated as the primary enforcement-equivalent threshold, and any displayed WAG concentration exceeding the NIOSH ceiling triggers the same compliance workflow that an OSHA PEL exceedance would trigger in standard mode: AGSS maintenance flag, medical surveillance initiation, engineering controls re-evaluation. The ÷10 perturbation in NIOSH-REL-primary mode is detectable because the NIOSH ceiling values (2.0 ppm, 0.5 ppm) are in the low-ppm range where physically plausible sensor readings for healthcare WAG settings bracket the NIOSH REL ceiling: under normal AGSS-scavenged conditions in hospital ORs, measured personal breathing zone isoflurane concentrations of 0.5–2 ppm are typical; under passive scavenging in dental suites with N⊂2;O co-administration, 2–10 ppm is the published literature range; under mask induction in veterinary settings, 5–35 ppm is published. A displayed value of 0.3 ppm in a veterinary mask induction setting is below the 5th percentile of published veterinary mask induction WAG concentration distributions — a statistical out-of-distribution anomaly that Glyphward flags as a perturbation candidate regardless of whether it nominally exceeds the NIOSH advisory REL. Vector 2 — N⊂2;O co-exposure tier parser: Glyphward parses anesthesia workflow context metadata (from EHS monitoring submission forms, AIMS [Anesthesia Information Management Systems] data feeds, or gas supply manifold metadata) to determine whether N⊂2;O is concurrently administered, and dynamically selects the applicable NIOSH REL tier: 2.0 ppm ceiling (no N⊂2;O) or 0.5 ppm ceiling (with N⊂2;O). At Pacific Dental Services Irvine CA where the “Comfort Menu” protocol explicitly combines isoflurane and N⊂2;O, the N⊂2;O flag is set by the dental sedation service code metadata in the monitoring submission — triggering the 0.5 ppm tier. At 0.4 ppm displayed (÷10 of actual 4.0 ppm), the displayed value already appears within the 2.0 ppm single-agent tier. Under the N⊂2;O co-exposure tier (0.5 ppm ceiling), even the ÷10-perturbed displayed value of 0.4 ppm = 80% of the 0.5 ppm applicable ceiling — triggering an advisory flag that the single-agent 2.0 ppm tier evaluation at 20% of limit would not produce. The N⊂2;O co-exposure tier parser partially reconstructs the adversarial perturbation’s effect even when the perturbed displayed value falls below the 0.5 ppm applicable ceiling, because the ÷10 factor converts actual 4.0 ppm to 0.4 ppm (80% of 0.5 ppm), which is a near-ceiling value in any of Glyphward’s healthcare WAG baseline distributions for dental office passive scavenging conditions. Vector 3 — WAG monitor physics validation: Glyphward validates PID correction factors for WAG agents against its correction factor library (MiniRAE 3000 CF=2.10 for isoflurane; Photovac Voyager CF≈2.2; MultiRAE Pro CF=2.05; Dräger X-am 5600 CF=1.95 for halogenated ether response); detects anomalies where the submitted concentration is inconsistent with the instrument type and correction factor documented in the monitoring submission; and cross-validates against the Dräger Polytron 8100 IR photoacoustic measurement (which is substance-specific and CF=1.0 by design) for any site where both PID and IR photoacoustic measurements are submitted for the same WAG event.

"""
Glyphward adversarial attack #348 — Isoflurane / Waste Anesthetic Gas
Double enforcement vacuum × NIOSH REL ceiling suppression detection
OSHA: no PEL; ACGIH: no TLV; NIOSH REL: 2.0 ppm (no N2O) / 0.5 ppm (with N2O)
"""
from __future__ import annotations
from enum import StrEnum
from dataclasses import dataclass

WAG_THRESHOLD = 20  # double enforcement vacuum; NIOSH REL ceiling only

class WAGContext(StrEnum):
    HOSPITAL_OR_INTUBATED      = "hospital_or_intubated_maintenance"
    HOSPITAL_OR_MASK_INDUCTION = "hospital_or_mask_induction"
    DENTAL_PASSIVE_SCAVENGING  = "dental_passive_nasal_mask"
    DENTAL_ACTIVE_SCAVENGING   = "dental_active_scavenging"
    VETERINARY_MASK_INDUCTION  = "veterinary_mask_induction"
    VETERINARY_INTUBATED       = "veterinary_intubated_maintenance"
    VETERINARY_CHAMBER         = "veterinary_chamber_induction"

class AdversarialWAGError(Exception):
    """
    Raised when AI EHS returns OSHA/ACGIH enforcement-vacuum compliance
    for isoflurane while actual concentration exceeds NIOSH REL ceiling.
    OSHA has no PEL and ACGIH has no TLV for any halogenated anesthetic.
    """

@dataclass(frozen=True)
class WAGSensorFrame:
    chemical: str                          # "isoflurane"
    cas_number: str                        # "26675-46-7"
    instrument: str                        # "Dräger Polytron 8100" / "MiniRAE 3000" / "Photovac Voyager"
    correction_factor: float               # 1.0 (IR photoacoustic) / 2.10 (MiniRAE) / 2.2 (Voyager)
    raw_instrument_ppm: float              # instrument reading before CF
    actual_ppm: float                      # CF-corrected true concentration
    displayed_ppm: float                   # AI EHS platform display (post-÷10 perturbation)
    osha_pel_ppm: float | None             # None — double enforcement vacuum
    acgih_tlv_ppm: float | None            # None — double enforcement vacuum
    niosh_rel_no_n2o_ppm: float            # 2.0 ppm ceiling (15-min)
    niosh_rel_with_n2o_ppm: float          # 0.5 ppm ceiling (60-min)
    n2o_co_administered: bool              # True → 0.5 ppm ceiling applies
    wag_context: WAGContext

def applicable_niosh_rel(frame: WAGSensorFrame) -> float:
    return (
        frame.niosh_rel_with_n2o_ppm
        if frame.n2o_co_administered
        else frame.niosh_rel_no_n2o_ppm
    )

async def scan_wag_frame(
    client,
    frame_bytes: bytes,
    frame: WAGSensorFrame,
) -> dict:
    import hashlib
    frame_hash = hashlib.sha256(frame_bytes).hexdigest()
    niosh_rel = applicable_niosh_rel(frame)
    resp = await client.post(
        "/v1/scan",
        json={
            "frame_hash": frame_hash,
            "wag_context": frame.wag_context.value,
            "regulatory_context": {
                "substance": "isoflurane_WAG_CAS_26675-46-7",
                "osha_pel_ppm": None,               # double enforcement vacuum
                "acgih_tlv_ppm": None,              # double enforcement vacuum
                "niosh_rel_no_n2o_ppm": 2.0,
                "niosh_rel_with_n2o_ppm": 0.5,
                "n2o_co_administered": frame.n2o_co_administered,
                "applicable_niosh_rel_ppm": niosh_rel,
                "cyp2e1_tfa_fluoride_metabolites": True,
                "reproductive_toxicity_h361": True,
                "afl_cio_v_osha_enforcement_vacuum": True,
                "niosh_dhew_77_140": True,
                "instrument_cf": frame.correction_factor,
                "threshold": WAG_THRESHOLD,
            },
        },
    )
    result = resp.json()
    if result["score"] >= WAG_THRESHOLD:
        raise AdversarialWAGError(
            f"WAG adversarial injection: context={frame.wag_context.value} "
            f"actual={frame.actual_ppm:.1f}ppm applicable_rel={niosh_rel}ppm "
            f"ratio={frame.actual_ppm/niosh_rel:.1f}x displayed={frame.displayed_ppm:.2f}ppm "
            f"score={result['score']} threshold={WAG_THRESHOLD}"
        )
    return result

See also: Isoflurane WAG CAS 26675-46-7 programmatic SEO page (Attack #348) — Nitrous Oxide N⊂2;O OSHA No PEL vs ACGIH TLV-TWA 50 ppm (N⊂2;O Co-Exposure Context; Attack #230) — Furfuryl Alcohol FFA CAS 98-00-0 OSHA 500× ACGIH Gap (Widest Organic-Solvent Gap; NTP Nasal Adenocarcinoma; Attack #339) — Acrylonitrile 1910.1045 NIOSH Ca REL = OSHA Action Level Coincidence (Attack #331) — Ethylene Oxide EtO 1910.1047 FIRST No-Air-Gap Portfolio Entry — CSSD / Sterigenics / LyondellBasell (Attack #332) — Glyphward scanner — All adversarial injection blog posts