ABIM Critical Care common mistakes
ABIM Critical Care · Understand the traps and how to avoid them.
Choosing the single most abnormal lab value as the diagnosis and treating it in isolation, rather than first constructing a unifying acid-base or electrolyte explanation (e.g., labeling a low bicarbonate as metabolic acidosis without calculating the anion gap and delta gap).
The exam presents busy, multi-abnormality stems; candidates anchor on the most visually striking number and skip the stepwise compensation/anion-gap framework they know but do not routinely apply under time pressure.
What to do instead
For every acid-base or electrolyte stem, force a written sequence before answering: pH, primary disorder, expected compensation (Winter's formula, etc.), anion gap with albumin correction, and delta-delta. Practice until this becomes an automatic checklist applied to every such item.
Answering hemodynamic questions from a single number (e.g., a low central venous pressure or a normal blood pressure) instead of integrating the full profile of preload, cardiac output, and mixed/central venous oxygen saturation to classify the shock state.
Candidates memorize threshold cutoffs but not the physiologic relationships among them, so a distractingly normal or abnormal single value drives the answer.
What to do instead
Build a mental shock-classification table (hypovolemic, cardiogenic, distributive, obstructive) with the expected direction of CVP, CO, SVR, and ScvO2 for each, and rehearse assigning any presented profile to a category before reading the options.
Treating ventilator questions as recall of mode names rather than analyzing the patient-ventilator interaction: failing to distinguish high peak from high plateau pressure, misattributing a dyssynchrony or auto-PEEP to the wrong cause, or not applying lung-protective and weaning criteria to the specific scenario.
Mode nomenclature is easy to memorize and feels like the tested content, while the pressure-time and flow-time reasoning that actually drives management is less practiced.
What to do instead
Drill a fixed interpretive algorithm: check plateau vs peak to localize resistance vs compliance, compute driving pressure, assess auto-PEEP and patient effort, then map findings to the correct adjustment. Apply it to every ventilator stem regardless of how the question is phrased.
Managing suspected sepsis with reflexive broad-spectrum antibiotics and fluids but neglecting the time-critical bundle elements and source control, or conversely delaying antibiotics to await cultures in an unstable patient.
Candidates split into two failure patterns: protocol-driven answers that ignore the individual source, or diagnostic-caution answers that violate the principle that cultures should not delay therapy in shock.
What to do instead
For every sepsis stem, explicitly sequence: recognize organ dysfunction, draw cultures without delaying antibiotics, give appropriate empiric coverage plus source control, and reassess perfusion. Practice stems where the correct answer is 'antibiotics now, cultures after' versus 'drain the source.'
Failing to recognize and emergently manage the neurologic deterioration patterns (declining consciousness, pupillary change, herniation syndromes, status epilepticus, elevated intracranial pressure) because the answer is framed around imaging or subspecialty consultation rather than immediate bedside intervention.
Neurocritical care stems are information-dense and candidates default to 'get a scan' or 'call neurology' instead of the immediate stabilizing maneuver the question is testing.
What to do instead
Memorize a tiered emergency response for each neuro pattern (airway/ventilation, osmotherapy, seizure termination, blood pressure targets) and answer the 'what do you do first' question with the bedside action, reserving imaging and consultation for the subsequent step.
Confusing the resuscitative management of specific toxin and overdose syndromes-choosing the wrong antidote, missing the indication for dialysis or alkalinization, or applying decontamination that is contraindicated or time-irrelevant.
Toxicology rewards pattern recognition of toxidromes and antidote tables, but candidates mix up antidotes across agents and forget that some interventions (e.g., urinary alkalinization, extracorporeal removal) hinge on specific toxin properties.
What to do instead
Build and repeatedly self-test a toxidrome-to-antidote-to-enhanced-elimination matrix, and for each overdose stem identify the toxin's key property (protein binding, volume of distribution, dialyzability) before selecting an intervention.
Answering procedural and resuscitation questions by describing the correct technique but selecting an option that omits the safety or confirmation step (e.g., ultrasound confirmation, post-procedure imaging, checklist pause, or reversal of anticoagulation) that the scenario is actually testing.
Candidates focus on the mechanical act of the procedure and overlook that exam items frequently hinge on the peri-procedural safety sequence.
What to do instead
For every procedure stem, run a pre/intra/post checklist (indication, consent, coagulation and anticoagulation status, site and technique, confirmation of placement, complication surveillance) and choose the option that includes the critical safety element.
Defaulting to maximal life-sustaining treatment in end-of-life or goals-of-care scenarios, or conversely assuming a surrogate's stated preference overrides documented patient wishes, instead of applying decision-making hierarchy and substituted judgment.
Candidates conflate 'what is medically possible' with 'what is ethically and legally appropriate,' and are unfamiliar with the hierarchy of decision-makers and the weight of advance directives.
What to do instead
Study the decision-making hierarchy (patient with capacity, then advance directive, then surrogate using substituted judgment, then best interests) and practice answering goals-of-care stems by identifying whose preference controls and whether the proposed action aligns with documented wishes.
Treating quality-improvement and patient-safety items as generic management trivia rather than applying a structured framework-misidentifying the correct QI tool, the right level of a systems error, or the appropriate root-cause versus improvement approach for the described problem.
These domains feel peripheral to clinical knowledge, so candidates under-prepare and answer from intuition rather than from the frameworks (e.g., model for improvement, error taxonomy, human-factors analysis) the items test.
What to do instead
Review the core QI and safety frameworks and practice mapping each stem to the correct tool or error category, distinguishing system failures from individual failures and choosing interventions that address the identified level.
Managing endocrine, nutritional, and metabolic emergencies with a single corrective measure-e.g., giving insulin in diabetic ketoacidosis without addressing volume, potassium, and the precipitant, or correcting sodium or glucose too rapidly-rather than sequencing the full corrective plan and its complications.
These crises involve coupled derangements (volume, potassium, osmolality, precipitating illness), and candidates fixate on the headline abnormality while neglecting order of operations and correction-rate limits.
What to do instead
For each metabolic emergency, rehearse a sequenced plan (fluids first, then electrolyte/insulin therapy with rate limits, identify and treat the precipitant, monitor for the classic complication of overcorrection) and answer by selecting the option that reflects correct ordering and safe rates.