ABP Med Toxicology common mistakes
ABP Med Toxicology · Understand the traps and how to avoid them.
Treating a positive urine immunoassay drug screen as a diagnosis - e.g., letting an opiate-positive screen explain a depressed patient and stopping the workup there, or declaring a synthetic-opioid overdose 'negative' because the screen is negative.
Screens are fast, familiar, and emotionally satisfying as a 'test that proves the story'; training in cross-reactivity (e.g., quinolones, poppy seeds, dextromethorphan) and in the structural blind spots for fentanyl analogues and novel benzodiazepines is thin, so the assay gets more epistemic weight than it deserves.
What to do instead
Before every screen is interpreted, write the one-line rule: immunoassay = exposure history only, never rule-in or rule-out. Anchor management on the history, toxidrome, and quantitative serum tests (acetaminophen, salicylate, ethanol, lithium, iron, digoxin, carboxyhemoglobin, methemoglobin) sent with a documented draw time. Ask explicitly in practice questions: what does this result change?
Plotting an acetaminophen concentration on the Rumack-Matthew nomogram using the wrong time zero, or applying the nomogram at all to a repeated supratherapeutic ingestion, a chronic alcohol user, or a presentation beyond 24 hours.
The nomogram is the single most memorised image in toxicology, and candidates reach for it reflexively; the qualifying conditions (single acute ingestion, known time, 4–24 h window) get compressed into a footnote in most study resources, and time zero defaults to arrival rather than ingestion.
What to do instead
Adopt a fixed three-question gate before plotting: single ingestion? time known and reliable? between 4 and 24 hours post-ingestion? If any answer is no, switch deliberately to the non-nomogram pathway (NAC for repeated supratherapeutic ingestion, undetectable/unknown time, or late presentation with any measurable level plus injury markers). Always state the assumed time zero out loud in explanations.
Under-titrating atropine in organophosphate poisoning because of a reflex fear of tachycardia, and giving a single dose then waiting for improvement rather than doubling doses until secretions dry.
Textbook doses (0.5–1 mg) come from cardiac/anticholinergic teaching, not from poisoning, where requirements can reach hundreds of milligrams; the physiological endpoint (dry lungs, not heart rate) is under-taught, and bradycardia as the monitoring target is over-learned.
What to do instead
Memorise the endpoint before the dose: titrate to clearing pulmonary secretions and resolved bronchorrhea, doubling the dose every few minutes; tachycardia is expected and not a reason to stop. Pair this with the two adjunct rules tested alongside it: an oxime early for nicotinic features, and avoid succinylcholine for intubation because phosphorylated pseudocholinesterase prolongs paralysis.
Sending an opioid-tolerant or long-term benzodiazepine user home from the emergency department after naloxone reversal or a few hours of observation, or stopping a benzodiazepine abruptly in a patient admitted for another problem.
The dramatic reversal rewards a 'fixed' framing, and withdrawal is treated as a soft, non-emergent problem compared with the ingestion that brought the patient in; the prolonged half-life of buprenorphine, methadone, and diazepam relative to naloxone or to a short observation window is easily overlooked.
What to do instead
Build a standing discharge checklist for any opioid- or sedative-dependent patient: precipitated withdrawal risk, naloxone kit and education, and an explicit plan for the home benzodiazepine dose continued in hospital. For withdrawal risk, use symptom-triggered scoring (CIWA-type) with a written taper rather than fixed-dose or abrupt cessation, and state the delayed-onset agents (methadone, buprenorphine, long-acting benzodiazepines, gabapentinoids) in every answer.
Applying a constriction band and pressure immobilisation, or performing early fasciotomy, for a North American pit viper bite; or ruling out amatoxin mushroom poisoning because symptoms did not appear within the first six hours when the actual ingestion history is unreliable.
Pressure immobilisation is drilled for elapid bites and Australian/Asian practice, and candidates generalise it; the 6-hour rule is remembered backwards in the 'late = benign' direction, while the true lesson is that early (<6 h) gastrointestinal symptoms argue against amatoxin but late presentation does not exclude it.
What to do instead
Store region-specific rules explicitly: pressure immobilisation for elapid snakes only, never for viperids; splint, keep the limb at heart level, mark the leading edge of swelling every 15 minutes, and give antivenom for progressive systemic or haematological toxicity rather than waiting for a compartment syndrome. For mushrooms, ask about the specific 8–12 h delay with hepatic phase and treat empirically with silibinin/NAC plus charcoal when the history fits, regardless of early GI symptoms.
Starting multiple-dose activated charcoal, urinary alkalinisation, or haemodialysis on the basis of a drug name or a single number, without checking the agent-specific indication, the contraindications in this patient, and the lab prerequisite (e.g., alkalinisation in a patient with hypokalaemia, or MDAC in a patient with an unprotected airway).
Decontamination and enhanced elimination are taught as a menu of options attached to drug lists, so the drug triggers the intervention; contraindications and prerequisites (potassium before alkalinisation, ileus and aspiration risk before charcoal, EXTRIP criteria before dialysis) are learned separately and fail to fuse under exam pressure.
What to do instead
For each of charcoal, alkalinisation, and dialysis, memorise a two-column card: indications (drug, timing, severity criteria) and stop/hold criteria (airway, ileus, K+, pH, QRS, renal function). Answer in the order: does this patient meet criteria now, what could this harm, what number do I need before and during. For alkalinisation, state the target urine pH and the potassium replacement plan in the same breath as the bicarbonate.
Reassuring a parent that a single tablet swallowed by a toddler is harmless without calculating mg/kg against the agent's toxic threshold, and dosing an antidote (naloxone, NAC, atropine, glucagon) from the adult dose rather than weight.
One tablet sounds trivially small, and the child looks well at presentation; the concept of 'one-pill killers' (sulfonylureas, clonidine, calcium channel blockers, tricyclics, opioids, chloroquine) is taught as a list rather than as a calculation habit, and antidote dosing is memorised in adult vials.
What to do instead
Force the arithmetic every time: weight in kg, dose ingested in mg/kg, compared against a memorised threshold table for the specific agent, plus a decontamination window and observation duration tied to the drug's pharmacokinetics. For antidotes, write the paediatric dose as mg/kg with the maximum, and state the infusion rate and endpoint. 'Well-appearing' is never the discharge criterion; observed time past peak effect is.
Reaching for haemodialysis purely because a lithium, salicylate, methanol, or ethylene glycol level is high, without assessing the clinical and EXTRIP-type criteria (acidosis, altered mentation, seizures, renal failure, refractory symptoms, time since ingestion, and the drug's volume of distribution and protein binding).
Numbers feel objective and are easy to recall, whereas the criteria are multi-part and partly clinical; candidates also conflate dialysability (small volume of distribution, low protein binding) with the decision to dialyse, and forget that lithium dialyses well but redistributes.
What to do instead
Learn a single decision frame: dialysability properties, then the specific criteria, then the endpoint and the need for repeat sessions (lithium rebound, salicylate ongoing absorption). Practise saying 'level alone is not an indication' and naming the two or three criteria that would make it one for each agent. For lithium, pair dialysis with the post-dialysis level recheck at 4–6 hours.
Treating an elevated osmolar gap as proof of toxic alcohol ingestion, or dismissing the diagnosis because the osmolar gap is normal or the anion gap has not yet opened; giving ethanol instead of fomepizole without checking whether the patient already has ethanol on board.
The osmolar and anion gaps are taught as a two-stage story but candidates memorise the sequence loosely and forget that the gap depends on the timing of ingestion, the assay's ethanol contribution, and the patient's baseline osmolality; the 'gap closes as the anion gap opens' principle is the very point being tested.
What to do instead
Memorise the timeline explicitly: osmolar gap peaks early and falls as metabolism generates acid, so a normal osmolar gap with a rising anion gap and acidosis is late, not absent, disease. Calculate the gap with the correct formula and ethanol correction, and treat on suspicion when the history and acidosis fit rather than waiting for a methanol/ethylene glycol level. Confirm fomepizole dosing is adjusted for concurrent ethanol and for haemodialysis.
On risk-assessment items, computing a reference dose or safe exposure by dividing the NOAEL by a single default factor, or by using an uncertainty factor the question did not support, and then reporting the output as if it defined an individual patient's risk.
The mathematics is simple, so candidates rush it, and the conceptual distinctions (hazard versus risk, NOAEL versus LOAEL, interspecies versus intraspecies factors, the difference between population-based regulatory values and individual clinical probability) are the parts that actually get tested and are the parts most often skimmed.
What to do instead
Write out the formula and label each factor by its justification (10 for animal-to-human, 10 for human variability, extra for LOAEL, subchronic data, or severe effect) before calculating; if the stem does not justify an extra factor, do not invent one. Then state in one sentence what the number is (a population-based screening value with deliberate conservatism) and what it is not (a threshold for an individual exposure or a treatment trigger).