The Rumack-Matthew nomogram: plotting a single acute acetaminophen ingestion against serum concentration and time to decide whether N-acetylcysteine is indicated
Acetaminophen (paracetamol) is the most commonly ingested drug in intentional and unintentional overdose worldwide, and it is uniquely dangerous because early symptoms so poorly predict eventual liver injury. A patient who has just swallowed a massive supratherapeutic dose may look and feel entirely well for the first day. The single most important — and most frequently mishandled — step in management is establishing a reliable time of ingestion and drawing the serum acetaminophen concentration at the correct moment, because a level drawn too early is not just unhelpful, it can be actively misleading.
Acetaminophen is a small, highly water-soluble molecule that is normally absorbed rapidly from the stomach and proximal small bowel, reaching peak plasma concentration within 30–90 minutes after a therapeutic dose taken on an empty stomach. In overdose, however, absorption kinetics change in ways that make an early level unreliable:
• Pylorospasm and delayed gastric emptying — large doses of acetaminophen itself slow gastric motility, delaying delivery to the absorptive surface of the small intestine. • Co-ingestants — opioids, anticholinergics, and other agents commonly taken in combination overdoses (or present in combination products such as acetaminophen-opioid tablets) further delay gastric emptying and can push the true absorption peak out past 4 hours. • Formulation effects — extended-release and enteric-coated products are specifically designed to delay and prolong release, and can produce a second later absorption peak. • Tablet mass and bezoar formation — massive ingestions can form a concretion of tablets in the stomach that continues to dissolve and release drug over many hours.
Because of this, a concentration drawn at 1 or 2 hours after ingestion may substantially underestimate the eventual peak, understating true risk and falsely reassuring the clinician. The Rumack-Matthew nomogram was validated using levels drawn no earlier than 4 hours post-ingestion, and applying it to an earlier level is a recognized cause of nomogram misuse. If a level is drawn early and falls below the treatment line, standard practice is to redraw at the 4-hour mark (or later) before using it to rule out the need for treatment — never to accept an early low level as reassuring on its own.
The clinical course of untreated acetaminophen poisoning is classically divided into four stages, and the first of these is exactly why history and laboratory testing — not physical exam — must drive early management:
Stage I (0.5–24 hours): Many patients are entirely asymptomatic. Others report anorexia, nausea, vomiting, malaise, pallor, and diaphoresis — a nonspecific viral-illness-like picture easily dismissed or misattributed to anxiety, alcohol, or a co-ingestant. Laboratory studies, including liver enzymes, are usually still normal in this window even in patients who will go on to develop severe hepatotoxicity.
This dissociation between symptom severity and eventual outcome is the central teaching point of acetaminophen overdose management: a comfortable-appearing, minimally symptomatic patient at hour 6 can still be tracking toward fulminant hepatic failure by hour 72–96 if a toxic ingestion goes untreated. Conversely, transient vomiting from a modest, non-toxic ingestion does not imply liver injury. This is precisely why objective data — a verified time of ingestion and an appropriately timed serum level plotted against the nomogram — replaces symptom-based triage as the decision tool, and why every patient with a history concerning for acetaminophen ingestion needs a level, regardless of how well they look in the emergency department.
The Rumack-Matthew nomogram is a semi-logarithmic plot of serum acetaminophen concentration (log scale, y-axis) against hours since a single acute ingestion (linear scale, x-axis), with a single treatment line running from 150 µg/mL at 4 hours to 4.3 µg/mL at 24 hours. Any level plotting on or above that line — for a patient with a reliably known ingestion time between 4 and 24 hours — indicates a course predicted to risk hepatotoxicity, and N-acetylcysteine treatment is recommended. It remains, fifty years after its publication, the single most consequential decision tool in clinical toxicology.
The nomogram plots concentration on a logarithmic y-axis specifically because acetaminophen elimination approximates first-order kinetics: a constant fraction of the remaining drug is cleared per unit time, which produces a straight line when concentration is log-transformed and plotted against a linear time axis. The treatment line is fully defined by two anchor points:
• 4 hours post-ingestion → 150 µg/mL (≈993 µmol/L) • 24 hours post-ingestion → 4.3 µg/mL (≈28 µmol/L)
Between these points, the line follows log10(concentration) = log10(150) + [(log10(4.3) − log10(150)) / 20] × (hours − 4), which corresponds to a concentration halving roughly every 4 hours along the line — chosen to approximate the elimination half-life of acetaminophen at toxic, glutathione-depleted doses (which is prolonged compared with the roughly 2-hour half-life seen at therapeutic doses). Any point plotting on or above this line, for a level drawn between 4 and 24 hours after a single acute ingestion, falls in the zone where NAC treatment is recommended; points below the line fall in a zone where hepatotoxicity is not expected and treatment is not required on nomogram grounds alone.
The original 1975 publication actually depicted two lines: a lower line at 200 µg/mL at 4 hours (labeled "possible toxicity") and an upper line at 300 µg/mL at 4 hours (labeled "probable toxicity"), both converging toward the same low-concentration tail near 24–30 hours. Recognizing that waiting to distinguish "possible" from "probable" delayed treatment in patients who would benefit from it, a lower single line — the modern 150 µg/mL-at-4-hours line used throughout this simulation — was adopted in the United States with an approximate 25% safety margin below the original "possible toxicity" line, intentionally trading specificity for sensitivity.
The nomogram was derived retrospectively from a relatively small cohort — on the order of 100 patients with known single acute ingestions and measured outcomes — correlating admission concentrations against the subsequent development of hepatotoxicity (conventionally defined as peak AST/ALT >1000 IU/L). Because the line was deliberately drawn below the concentrations actually associated with observed liver injury in that cohort, it functions with very high sensitivity for hepatotoxicity risk (few true toxic ingestions are missed) at the cost of lower specificity (some patients treated will never have developed injury even without NAC).
Key limitations to keep in mind whenever the nomogram is applied clinically:
• Requires a single acute ingestion, not staggered or chronic supratherapeutic dosing. • Requires a reliably known time of ingestion; an estimated or uncertain time invalidates the plot. • Only valid from 4 hours (earliest reliable level) to 24 hours (beyond this, use clinical judgment, transaminases, and acetaminophen level trend rather than the graphical line). • Was derived from and validated primarily in adults and older children with immediate-release products; delayed-release/extended-release kinetics can fall outside the model's assumptions. • Was never intended to predict absence of ALL toxicity — it predicts risk of the specific hepatotoxicity outcome it was validated against.
Despite these caveats, the nomogram's core strength — converting a chaotic, high-stakes decision into a reproducible graphical rule — has made it one of the most durable and widely taught tools in all of emergency medicine and toxicology.
Acetaminophen hepatotoxicity is a story of metabolic pathway saturation. At therapeutic doses, the liver comfortably clears acetaminophen through two safe conjugation pathways. In overdose, a minor oxidative pathway becomes proportionally more important, generating a reactive metabolite — N-acetyl-p-benzoquinone imine (NAPQI) — faster than the liver's glutathione reserves can neutralize it. Once those reserves are exhausted, NAPQI is free to covalently bind hepatocellular proteins, triggering the centrilobular (zone 3) necrosis that defines acetaminophen liver injury.
At therapeutic dosing, acetaminophen is handled almost entirely by two phase II conjugation pathways in hepatocytes:
• Glucuronidation (UGT enzymes): roughly 50–55% of a therapeutic dose, producing an inactive, water-soluble glucuronide conjugate excreted renally. • Sulfation (SULT enzymes): roughly 30–44% of a therapeutic dose, similarly producing an inactive sulfate conjugate. Sulfation capacity is limited by available sulfate cofactor and saturates at relatively low doses. • Oxidation (CYP2E1, with contributions from CYP1A2 and CYP3A4): only about 5–10% of a therapeutic dose is routed through this pathway, generating the reactive electrophile NAPQI. Under normal conditions, NAPQI is immediately conjugated by hepatic glutathione to form a non-toxic mercapturate, and only trace amounts ever accumulate.
In overdose, the glucuronidation and especially the sulfation pathways saturate — there simply is not enough UDP-glucuronic acid and inorganic sulfate to keep pace — so a larger absolute (and proportional) amount of the ingested dose is shunted through CYP-mediated oxidation. NAPQI production rises accordingly. Glutathione is consumed to conjugate it, and because glutathione synthesis cannot keep pace with this accelerated demand, hepatic glutathione stores begin to fall. Clinically significant hepatotoxicity becomes likely once glutathione stores are depleted by roughly 70% of baseline — beyond this point, free reactive NAPQI accumulates faster than it can be cleared.
Once glutathione conjugation capacity is overwhelmed, unconjugated NAPQI covalently binds cysteine residues on hepatocellular proteins — including mitochondrial proteins — impairing mitochondrial respiration, generating oxidative stress, and ultimately triggering hepatocyte necrosis. This injury is concentrated in zone 3 (centrilobular/perivenular) hepatocytes because these cells have the highest constitutive CYP2E1 expression and receive the least-oxygenated blood in the hepatic sinusoid, making them both the greatest NAPQI producers and the most vulnerable to the resulting oxidative and mitochondrial injury.
The untreated clinical course unfolds in four classic stages:
Stage I (0.5–24 h): Asymptomatic or nonspecific nausea, vomiting, malaise, diaphoresis; transaminases typically still normal. Stage II (24–72 h): Right-upper-quadrant abdominal pain and tenderness as hepatocellular injury develops; AST/ALT begin rising, sometimes dramatically (into the thousands); bilirubin and INR may begin to rise; some patients develop oliguria/renal function changes. Stage III (72–96 h): Peak hepatotoxicity — AST/ALT often peak here (frequently >10,000 IU/L in severe cases); this is the window in which fulminant hepatic failure, coagulopathy, hepatic encephalopathy, lactic acidosis, and acute kidney injury manifest; this stage carries the highest mortality risk if the injury is severe enough. • Stage IV (4 days–2 weeks): Either recovery, with resolution of hepatic dysfunction and histologic regeneration over 1–3 months, or progression to multiorgan failure and death in the most severe cases.
NAC administered early — ideally well before glutathione depletion is complete — interrupts this cascade by directly replenishing the substrate the liver needs to detoxify NAPQI before mitochondrial injury and necrosis become established.
N-acetylcysteine is one of the great success stories of clinical toxicology: a cheap, widely available antidote that, given early enough, essentially prevents acetaminophen hepatotoxicity. It works by directly restoring the liver's capacity to detoxify NAPQI, and — even once significant injury has begun — continues to provide benefit through antioxidant and hemodynamic mechanisms. Multiple validated regimens exist, differing mainly in route, infusion schedule, and how they balance treatment duration against the risk of infusion-related adverse reactions.
NAC acts through several complementary mechanisms:
• Glutathione repletion: NAC is deacetylated to cysteine, the rate-limiting substrate for glutathione synthesis, directly restocking the pool available to conjugate NAPQI. • Direct NAPQI conjugation: NAC itself can act as an alternate glutathione surrogate/sulfhydryl donor, directly binding and neutralizing NAPQI even before full glutathione resynthesis occurs. • Antioxidant effects: NAC scavenges reactive oxygen species generated during NAPQI-mediated mitochondrial injury, limiting oxidative damage independent of conjugation chemistry. • Hemodynamic/microcirculatory support: in established hepatic failure, NAC improves oxygen delivery and microvascular blood flow, which is why NAC is continued and shown to improve outcomes even in patients who present late, after hepatotoxicity is already established — it is not purely a "prevention window" drug.
The efficacy of NAC is strongly time-dependent: treatment started within 8 hours of ingestion is associated with a hepatotoxicity rate approaching zero, treatment between 8–10 hours shows measurably reduced but still excellent efficacy, and efficacy continues to decline the later treatment is started — one of the strongest arguments for rapid recognition, appropriately timed level draws, and prompt nomogram interpretation described in earlier stages.
Several validated NAC regimens are in clinical use:
IV three-bag ("Prescott") protocol — 21 hours, 300 mg/kg total: • Bag 1 (loading): 150 mg/kg in dextrose 5% infused over 60 minutes • Bag 2: 50 mg/kg infused over the next 4 hours • Bag 3: 100 mg/kg infused over the following 16 hours
IV two-bag (simplified) protocols — increasingly adopted to reduce dosing errors and infusion-reaction rates, typically: • Bag 1: 200 mg/kg infused over 4 hours • Bag 2: 100 mg/kg infused over the following 16 hours (total again 300 mg/kg over 20 hours; the slower initial rate compared with the classic 60-minute Prescott load reduces peak plasma NAC concentration and the associated histamine-release reactions)
Oral protocol — 72 hours, higher cumulative dose: • Loading dose: 140 mg/kg by mouth • Maintenance: 70 mg/kg every 4 hours for 17 additional doses • Total cumulative dose ≈1330 mg/kg over 72 hours; requires a tolerated oral/enteral route and is less commonly used where IV NAC is available, owing to palatability (sulfurous odor/taste), vomiting, and the longer treatment course.
Duration is extended beyond the standard course, and transaminases/INR are trended, whenever a patient has a markedly elevated presenting level, detectable acetaminophen with elevated transaminases at completion of the standard course, or evidence of hepatotoxicity — NAC should not be stopped on a fixed clock if biochemical markers of ongoing injury persist.
Anaphylactoid reactions — flushing, urticaria, angioedema, bronchospasm, and rarely hypotension — occur in roughly 10–20% of patients receiving IV NAC, driven by non-immunologic (non-IgE-mediated) histamine release related to infusion rate, and are therefore most common during the rapid original 60-minute loading dose. Management is to slow or briefly pause the infusion, give antihistamines (and bronchodilators/epinephrine for severe reactions), and then resume at a slower rate rather than abandoning treatment — true NAC hypersensitivity requiring discontinuation is rare. This reaction profile is a major reason two-bag regimens with a slower initial infusion have gained favor.
The Rumack-Matthew nomogram was validated for exactly one scenario: a single acute ingestion with a reliably known time, plotted with a level drawn between 4 and 24 hours. Outside that scenario the graphical line is not just less accurate — it can be dangerously misleading. Chronic and staggered supratherapeutic ingestion, unknown time of ingestion, and extended-release formulations all require alternative decision frameworks, and patients who do progress to severe hepatotoxicity must be assessed against validated transplant-listing criteria.
Repeated supratherapeutic ingestion (RSI) — taking doses above the therapeutic maximum repeatedly over many hours to days, rather than as one bolus — is common (accidental overdosing on combination cold/flu products, chronic self-medication for pain) and cannot be plotted on the Rumack-Matthew nomogram at all, because there is no single "time zero" and no single dose to anchor a treatment line against. In this setting, decision-making instead relies on a combination of: any detectable acetaminophen concentration together with a recognized risk factor (large cumulative dose, chronic alcohol use, malnutrition/fasting state, concurrent enzyme-inducing medications), or any elevation of AST/ALT above the upper limit of normal, either of which should prompt NAC treatment empirically rather than waiting on a nomogram plot that does not exist for this pattern.
Similarly, when the time of ingestion is unknown or unreliable (unwitnessed overdose, altered/intoxicated patient, unclear history), the level cannot be validly located on the x-axis of the nomogram at all. Standard practice in this situation is to treat empirically with NAC while obtaining a second level and liver enzymes, rather than attempting to force an uncertain history onto a tool that requires precision on exactly that variable.
Extended-release acetaminophen products are designed to release drug over a prolonged period and can produce a delayed or biphasic absorption curve; a single early level that appears reassuringly low may not reflect a second later peak. Standard caution is to obtain a repeat level roughly 4 hours after the first (and to have a lower threshold for empiric NAC) whenever an extended-release product is suspected or confirmed, rather than relying on one single-timepoint nomogram plot as in an immediate-release ingestion.
Co-ingestants that slow gastrointestinal motility — opioids, anticholinergics, and other agents that delay gastric emptying — can likewise delay and blunt the acetaminophen absorption curve, again arguing for a repeat level before assuming a low early result is truly reassuring.
For the minority of patients who progress to severe acetaminophen-induced acute liver failure despite treatment, King's College Criteria are the most widely used prognostic tool for identifying patients who need urgent liver transplant evaluation:
• Arterial pH <7.30 (after adequate fluid resuscitation), OR • All three of: INR >6.5 (prothrombin time >100 seconds), serum creatinine >3.4 mg/dL (300 µmol/L), and grade III or IV hepatic encephalopathy.
Meeting either criterion identifies a cohort with historically very high mortality without transplantation, and should trigger immediate transfer to/consultation with a liver transplant center; NAC is continued throughout this evaluation, since it continues to provide hemodynamic and antioxidant benefit even at this late stage.