ABPS Hospital Medicine common mistakes
ABPS Hospital Medicine · Understand the traps and how to avoid them.
Ordering the next step in a resuscitation sequence before committing to the underlying diagnosis and severity (e.g., reaching for a fluid bolus or pressor in undifferentiated shock without first classifying it by volume status, cardiac output, and perfusion findings).
Shock vignettes are written to reward the first decision that changes management, and candidates under time pressure pattern-match to a familiar drug or fluid instead of building a physiology-based classification from the stem's hemodynamic and exam data.
What to do instead
Before selecting any intervention, explicitly write down the shock category (hypovolemic, distributive, cardiogenic, obstructive) with the two or three data points that support it; then choose the intervention that addresses that mechanism, and verify it is the single most informative next step rather than a downstream one.
Treating a single abnormal vital sign or lab value in isolation rather than interpreting it in the context of the full clinical picture (e.g., acting on a low bicarbonate without calculating the anion gap, delta gap, and checking respiratory compensation).
Isolated numbers are salient and easy to recall, so candidates anchor on the value rather than performing the multi-step interpretation the question is testing; acid-base and electrolyte items almost always require a sequence of calculations.
What to do instead
For every acid-base or electrolyte stem, run a fixed checklist: pH and primary disorder, expected compensation, anion gap, delta-delta, and osmolar gap where relevant; only then answer, and confirm the chosen answer is consistent with all computed values.
Choosing the most aggressive or broadest therapy when the question asks for the next appropriate step in antimicrobial management (e.g., escalating to broad-spectrum coverage instead of narrowing, de-escalating, or addressing source control when the culture data and clinical trajectory support it).
Candidates equate 'sicker patient' with 'broader antibiotics' and under-weight stewardship principles, source control, and the specific question wording ('most appropriate next step' vs 'empiric coverage').
What to do instead
For each infectious disease item, identify whether the question is about empiric initiation, de-escalation, duration, or source control; select the answer matching that phase, and prefer the option that incorporates culture data and the least broad effective regimen consistent with the clinical scenario.
Missing the perioperative medication and timing implications in a surgical patient (e.g., failing to hold or bridge an anticoagulant, continuing a medication that should be stopped, or overlooking a postoperative complication masquerading as a routine finding).
Perioperative items layer a surgical context over internal medicine knowledge, and candidates focus on the procedure rather than the medication timeline, the patient's baseline risk, and expected postoperative physiology.
What to do instead
For any perioperative stem, list the patient's medications with perioperative implications, the planned procedure's bleeding and thrombosis risk, and the timing of the last dose; then answer with the management that matches that timeline, and screen for the specific postoperative complication the vignette is hinting at.
Failing to distinguish delirium from other causes of altered mental status and defaulting to a workup or treatment for a single presumed cause (e.g., ordering antipsychotics or assuming a urinary source without a structured delirium evaluation).
Altered mental status stems contain distractor findings that suggest one cause, and candidates accept the first plausible explanation instead of systematically excluding reversible contributors and confirming the syndrome.
What to do instead
For altered mental status, first confirm the syndrome (delirium vs dementia vs psychiatric vs focal deficit), then systematically review the reversible causes (hypoxia, hypoglycemia, infection, medications, withdrawal, metabolic, structural); answer with the step that addresses the most likely reversible cause identified by the stem's specific findings.
Under-recognizing time-critical cardiovascular emergencies because the presentation is atypical (e.g., missing an acute coronary syndrome, aortic dissection, or pulmonary embolism when the stem emphasizes a non-classic symptom or a normal initial test).
Textbook presentations are over-represented in study materials, and candidates discount diagnoses when a single test (e.g., an initial ECG or troponin) is normal, forgetting that timing and serial testing matter.
What to do instead
For any chest pain, dyspnea, or syncope stem, explicitly screen for the can't-miss emergencies first (ACS, dissection, PE, tamponade, tension pneumothorax, esophageal rupture); if the question hinges on a normal early test, choose the answer that reflects appropriate serial testing or further risk stratification rather than reassurance.
Misclassifying respiratory failure and selecting the wrong oxygen or ventilatory strategy (e.g., applying high-flow oxygen or noninvasive ventilation to a patient who needs immediate intubation, or failing to recognize hypercapnic vs hypoxemic failure).
Candidates conflate 'low oxygen' with 'needs more oxygen' and do not consistently classify the failure type or assess for impending respiratory fatigue and failure of noninvasive support.
What to do instead
For every respiratory stem, classify hypoxemic vs hypercapnic vs mixed failure, assess work of breathing and mental status, and identify contraindications to noninvasive support; choose the intervention that matches the classification, and if the stem shows fatigue or worsening despite support, select escalation to invasive ventilation.
Ignoring the systems, safety, and quality lens in items that ask about error prevention, handoffs, or process improvement (e.g., answering with an individual clinician's action when the question is testing a system-level intervention or a root-cause analysis).
Candidates are trained to think clinically and default to individual-level fixes, missing that quality and safety items reward system-based, standardized, and measurement-driven answers.
What to do instead
For quality, safety, and systems items, identify whether the question is asking for root-cause analysis, a standardized process, a measure, or a communication tool; prefer the answer that changes the system or process and includes a measurable outcome, rather than one that relies on individual vigilance.
Failing to reconcile conflicting data in nephrology and fluid/electrolyte stems (e.g., choosing a fluid or diuretic strategy without integrating volume status, urine output, and the trend in renal function, or misinterpreting a fractional excretion or urine sodium).
Renal vignettes present multiple data streams, and candidates latch onto one (often the creatinine) instead of synthesizing volume assessment with the urine studies and the clinical trajectory.
What to do instead
For renal and fluid items, build a one-line synthesis: volume status by exam, urine output and trend, urine chemistry interpretation, and the likely mechanism (prerenal, intrinsic, postrenal, or a specific tubular disorder); answer with the intervention that matches that synthesis and recheck that it is consistent with all provided data.
Missing the endocrine or metabolic emergency behind a nonspecific presentation (e.g., not recognizing thyroid storm, adrenal crisis, DKA/HHS, or a drug-induced metabolic derangement when the stem emphasizes a subtle or atypical finding).
Endocrine emergencies present with overlapping, nonspecific features, and candidates do not maintain a high index of suspicion or fail to connect a medication, infection, or recent event to the metabolic picture.
What to do instead
For any stem with tachycardia, altered mental status, hypotension, or unexplained metabolic derangement, explicitly screen for the endocrine emergencies and their triggers; answer with the diagnostic or therapeutic step that addresses the most likely emergency, and confirm the choice accounts for the precipitant the vignette provides.