USMLE Step 2 CK common mistakes
USMLE Step 2 CK · Understand the traps and how to avoid them.
Commits to a diagnosis after the first abnormal finding in the stem and then reads every subsequent data point as confirmation, never asking what single result would disprove it.
Stems are built with a salient opener (a striking potassium, a suppressed TSH, a dropped hemoglobin), and under time pressure that first plausible label relieves uncertainty; confirmation bias then suppresses the contradictory detail the writer deliberately planted.
What to do instead
Before answering, state aloud the one finding that would break your diagnosis. If the stem contains it, change the diagnosis. Read the final question sentence first, then re-read the data filtered through what is actually being asked, so a discordant value cannot be skimmed past.
For a hypotensive or otherwise unstable medical patient, selects the definitive diagnostic study (urgent endoscopy, CT angiogram, biopsy) instead of the immediate stabilizing action.
Study habits reward diagnostic yield, so the 'best test' reflex outranks the ABC/stabilize-first sequence that the question writer is actually testing; unstable patients are also written to look like they need a diagnosis fast.
What to do instead
Apply a fixed screen before looking at options: is there an immediate threat to airway, breathing, or circulation (hypotension, hypoxemia, altered mentation, active hemorrhage)? If yes, the answer is resuscitation and the study comes later. Only choose a diagnostic test when the stem states the patient is stable.
Sends a physiologically deranged trauma or acute-abdomen patient to the CT scanner before resuscitation, or chooses the definitive single-stage operation rather than a damage-control approach.
Operative and imaging knowledge is memorized as procedures and diagnoses, while the physiology-driven decision rules (hypotension, acidosis, hypothermia, coagulopathy) get far less rehearsal and feel less like 'surgery.'
What to do instead
For every acute surgical stem, compute physiologic status first: blood pressure, heart rate, lactate/base deficit, temperature, and mentation. Deranged physiology means bedside or no imaging and an abbreviated damage-control plan; stable physiology means full imaging and definitive repair. Commit to the branch before reading the options.
Manages perioperative anticoagulation and antiplatelet therapy as all-or-nothing, either stopping and bridging every patient regardless of indication or continuing agents without weighing procedure bleeding risk against thrombotic risk.
The question is framed as a pharmacology fact, but it is really a two-competing-risks decision, and candidates reach for a memorized 'stop the drug' rule rather than stratifying by indication and procedure.
What to do instead
Answer in three steps: name the anticoagulant indication and its thrombotic risk (mechanical valve or recent event versus low-risk atrial fibrillation), name the procedure's bleeding risk (neuraxial or major versus minor), then decide hold versus bridge. Justify bridging only when thrombotic risk clearly exceeds bleeding risk.
Applies adult normal ranges and adult dosing to infants and children, judging heart rate, respiratory rate, hemoglobin, and drug doses by adult thresholds.
Adult medicine dominates clinical exposure, and pediatric norms shift with each age band and are rarely re-derived under time pressure; a number 'looks normal' to an adult-trained eye.
What to do instead
Before answering any pediatric item, state the age explicitly and recall the age-specific vital sign range and weight-based (mg/kg) dose with its maximum cap. If a value is given, compare it to the range for that exact age band, not to adult norms.
Treats all pregnancies the same regardless of gestational age, applying one threshold for tocolysis, antenatal corticosteroids, magnesium, or delivery timing across the whole of pregnancy.
Obstetric guidelines are organized into gestational-age windows, but the gestational age is usually buried mid-stem while the dramatic presentation (bleeding, hypertension, contractions) pulls attention away from it.
What to do instead
Circle the gestational age before reading the options and classify it into a window (under 20 weeks, 20 to 23 6/7, 24 to 33 6/7, 34 to 36 6/7, 37 or more). Eliminate any option whose intervention is invalid outside that window, even if the intervention is otherwise correct.
Manages hypertensive disorders of pregnancy by treating the blood pressure number alone, missing severe features, omitting magnesium seizure prophylaxis, and choosing delivery timing without reference to gestational age and maternal or fetal status.
The stem leads with a striking blood pressure and a headache, which cues antihypertensive therapy as 'the answer,' while the severity criteria and delivery decision require assembling several findings the candidate never ranks.
What to do instead
Ask three ordered questions: are severe features present (symptoms, platelets, liver enzymes, renal function, pulmonary edema)? Has maternal stabilization been done (magnesium, controlled blood pressure)? Does gestational age and maternal or fetal status mandate delivery now or permit expectant management? Answer each before choosing an option.
Names a primary psychiatric disorder without first excluding a medical, neurologic, or substance-induced cause, and applies DSM criteria without checking the required symptom count, duration, and functional impairment.
Psychiatric vignettes resemble classic textbook cases, so the syndrome is recognized and 'locked in'; duration criteria are easy to skim because they sit in a clause rather than in a lab value, and candidates want to resolve the ambiguity quickly.
What to do instead
Before committing, read for vitals, urine toxicology, new medications (steroids, stimulants, interferon), and recent illness or surgery, and eliminate any option whose required duration or symptom count the stem does not satisfy. If a medical cause is present, treat it first rather than labeling a primary psychiatric disorder.
Answers biostatistics items by comparing relative risks, odds ratios, or p-values instead of computing absolute risk reduction and number needed to treat, and declares a result meaningful from a p-value alone.
Journals and teaching emphasize relative effects and statistical significance, so the more impressive ratio feels like the stronger answer, and computing an absolute difference under time pressure is deliberately effortful.
What to do instead
Draw the 2x2 table first, then compute ARR as control event rate minus intervention event rate, NNT as 1/ARR, and use the confidence interval to judge precision. Only after that compare options, and reject any answer that rests solely on a p-value or a large relative risk.
Recommends a live attenuated vaccine to a pregnant or significantly immunocompromised patient, or withholds a routine vaccine because of a mild acute illness, low-grade fever, or current antibiotic use.
Contraindication lists are memorized as a blur of exceptions, and candidates apply caution indiscriminately rather than sorting vaccines by mechanism and matching the exception to the correct patient.
What to do instead
Classify every vaccine as live attenuated or inactivated before answering. Live vaccines are contraindicated in pregnancy and significant immunocompromise; mild acute illness, low-grade fever, and concurrent antibiotics are not contraindications to vaccination. Match the exception to the specific patient in the stem rather than applying blanket caution.