Board Blind Spots
Twenty board-style questions on the concepts internal-medicine residents most often miss on their first ABIM certification exam. Choose an answer to reveal what the question is really testing, why residents miss it, and a primary-source reference. Written at the PGY-3 level.
0 of 20 answered · 0 correct
A new serum assay for a chronic infection has a sensitivity of 95% and a specificity of 90%. You apply it as a screen in a population where the disease prevalence is 1%, and a patient tests positive. Which statement best characterizes this result?
Predictive value depends on prevalence
At a prevalence of 1%, even a sensitive (95%) and specific (90%) test has a positive predictive value of only about 9% — so most positives are false positives (B). Choice A confuses sensitivity with the post-test probability, the single most common error; the NPV here is roughly 99.9%, higher than the PPV, so C is false; sensitivity and specificity are independent properties of the test, so D is wrong; and a low-PPV screen is not confirmatory (E).
Why residents miss it: Predictive values depend on prevalence — a counterintuitive dependency residents routinely ignore (base-rate neglect). Biostatistics is the best-documented internal-medicine knowledge gap and is rarely reinforced after medical school.
An 80-year-old nursing-home resident with mild baseline dementia is referred for a urinalysis obtained because of cloudy, malodorous urine. She has no fever, dysuria, frequency, suprapubic or flank pain, and no change from her baseline mental status. Urinalysis shows positive leukocyte esterase and nitrites with 50–100 WBC/hpf; culture grows greater than 100,000 CFU/mL of E. coli. What is the most appropriate management?
Asymptomatic bacteriuria — when not to treat
This is asymptomatic bacteriuria — a positive culture with pyuria but no urinary symptoms. In non-pregnant adults (including older adults with dementia and no acute urinary symptoms) it should not be treated (D). Cloudy, malodorous urine and pyuria do not change that. Antibiotics (A, B, E) cause harm without benefit, and a test of cure (C) is not indicated. Bacteriuria alone does not explain delirium unless other signs of infection are present.
Why residents miss it: The reflex to "treat the lab, not the patient" — reinforced by defensive medicine — overrides stewardship. Pyuria is mistaken for infection, and the older patient with nonspecific findings is the classic over-treatment trap.
A 44-year-old woman has a blood pressure of 158/98 mmHg despite adherence to maximal-dose lisinopril, amlodipine, and hydrochlorothiazide. She is asymptomatic. Serum potassium is 3.2 mmol/L and her eGFR is normal. What is the most appropriate next step?
Screening for primary aldosteronism
Resistant hypertension with spontaneous hypokalemia meets criteria to screen for primary aldosteronism — the most common secondary cause of hypertension. Screen with the aldosterone-to-renin ratio (B), confirm biochemically, and only then image. A treats before diagnosing, and a mineralocorticoid-receptor antagonist distorts the ratio; C screens the wrong axis (pheochromocytoma); D images before biochemical confirmation, where common nonfunctioning adenomas mislead; E under-evaluates a guideline-clear scenario.
Why residents miss it: Adrenal workups are rare in training and usually handed to consultants, so residents pattern-match to empiric therapy or jump to imaging. ABIM's own knowledge-gaps report flags "adrenal hormonal excess — testing" as a most-missed item.
A 74-year-old man reports fatigue. TSH is 7.8 mIU/L (reference 0.4–4.5) with a normal free T4. He has no goiter and no prior thyroid disease. Repeat testing at 8 weeks shows TSH 7.2 with a normal free T4; TPO antibodies are negative. What is the most appropriate next step?
Subclinical hypothyroidism — when not to treat
This is mild subclinical hypothyroidism (TSH in the 7–8 range with a normal free T4) in an older adult. The TRUST trial showed that levothyroxine produces no symptomatic benefit here; treatment is generally reserved for a TSH of 10 or higher, marked symptoms, or pregnancy. So observation is correct (E). A and B over-treat to "normalize the number"; C adds testing that would not change management; D looks for a nodule that is not suggested.
Why residents miss it: The reflex to correct an abnormal TSH overrides the evidence for doing less. The TSH-below-10, older-adult nuance is undertaught, and "less is more" runs against the instinct to add therapy.
A patient with 2 days of vomiting and an acute febrile illness has the following: pH 7.40, sodium 140, chloride 95, bicarbonate 24 mmol/L (albumin normal). What best describes the acid-base status?
Mixed acid-base disorders and the delta-delta
The bicarbonate (24) and pH (7.40) look normal, but the anion gap is 21 — clearly elevated. The elevated gap signals a high-anion-gap metabolic acidosis; the delta-delta shows the bicarbonate is higher than expected for that gap, revealing a concurrent metabolic alkalosis (here from vomiting) — answer C. Stopping at the normal pH and bicarbonate gives the wrong answer E; seeing only the gap gives A.
Why residents miss it: Residents stop at the pH and bicarbonate and never calculate the anion gap or delta-delta, so mixed disorders with a "normal" bicarbonate are missed. Acid-base is conceptually dense and often deferred to consultants.
Reference: Seifter JL. Integration of acid-base and electrolyte disorders. N Engl J Med. 2014.
A 70 kg woman with 3 weeks of thiazide use presents with a serum sodium of 110 mmol/L and mild confusion but no seizures. Over the first 24 hours of treatment her sodium rises to 122 mmol/L. What is the most appropriate action?
Overcorrection of hyponatremia and osmotic demyelination
A rise of 12 mmol/L in 24 hours exceeds the safe limit (about 6–8 mmol/L per 24 hours) in chronic hyponatremia and risks osmotic demyelination. The correct maneuver is to re-lower the sodium with hypotonic fluid, often with desmopressin (A). B wrongly accepts overcorrection; C and E would push the sodium higher; D does not address the rate already exceeded.
Why residents miss it: Attention fixes on raising the sodium rather than on the rate, and the danger of overcorrection (osmotic demyelination) is underappreciated. The counterintuitive step of "undo it by re-lowering the sodium" is rarely taught.
Reference: Sterns RH. Disorders of plasma sodium — causes, consequences, and correction. N Engl J Med. 2015.
A 58-year-old man has an acutely swollen, red, exquisitely tender first metatarsophalangeal joint for 1 day. Serum urate is 5.8 mg/dL (reference below 6.8). He is afebrile. What is the most appropriate next step to establish the diagnosis?
Acute gout — a normal urate does not exclude it
A normal serum urate during an acute flare does not exclude gout — urate often falls during attacks — and the presentation cannot be distinguished from septic arthritis without fluid. Arthrocentesis with polarized-light microscopy (negatively birefringent, needle-shaped crystals) is the answer (D). A risks missing septic arthritis; B misreads urate kinetics; starting allopurinol during a flare (C) can worsen it and is not diagnostic; radiographs (E) are not diagnostic acutely.
Why residents miss it: Gout feels like a common, easy diagnosis (a Dunning-Kruger blind spot), so residents anchor on it and skip the tap — missing both the normal-urate-in-flare pitfall and septic arthritis.
Reference: Mikuls TR. Gout. N Engl J Med. 2022.
A 35-year-old man needs cefazolin for surgical prophylaxis. His chart lists a penicillin allergy: as a child he developed a maculopapular rash on day 5 of amoxicillin, with no hives, no anaphylaxis, and no mucosal involvement. What is the most appropriate approach?
Penicillin-allergy labels and beta-lactam cross-reactivity
A benign delayed maculopapular rash in childhood is a low-risk history. Cephalosporin cross-reactivity with penicillin is very low — and cefazolin shares no side chain with the penicillins — so cefazolin can be given (B). More than 90% of patients labeled "penicillin allergic" are not. Blanket avoidance (A, E) worsens outcomes and stewardship; skin testing (C) is not required for low-risk histories; a guarded test dose (D) is unnecessary here.
Why residents miss it: The penicillin-allergy label is taken at face value, and an overstated fear of cephalosporin cross-reactivity drives reflexive avoidance — with real costs to antibiotic stewardship.
Reference: Shenoy ES, et al. Evaluation and Management of Penicillin Allergy: A Review. JAMA. 2019.
A 24-year-old woman develops diffuse urticaria, wheezing, and lightheadedness 10 minutes after a bee sting. Her blood pressure is 88/54 mmHg. What is the most appropriate immediate management?
Anaphylaxis — epinephrine first, by the right route
This is anaphylaxis (skin, respiratory, and cardiovascular involvement). Intramuscular epinephrine into the anterolateral thigh is the immediate, first-line, life-saving treatment (A). Antihistamines and steroids (B) are adjuncts only and do not treat the reaction; albuterol and fluids (C) are supportive; an IV bolus (D) risks arrhythmia and is not the initial route; subcutaneous absorption (E) is unreliable.
Why residents miss it: Under pressure, residents reach for diphenhydramine and steroids instead of epinephrine, or use the wrong route — failing to recognize that prompt IM epinephrine is the only step that changes the outcome.
Three weeks after starting allopurinol, a patient develops fever to 39°C, a diffuse morbilliform eruption with facial edema, and a transaminitis (ALT six times the upper limit of normal). The CBC shows eosinophilia and atypical lymphocytes. What is the most likely diagnosis?
DRESS versus a benign drug eruption
The long latency (2–6 weeks), fever, facial edema, eosinophilia, atypical lymphocytes, and hepatitis define DRESS (C); allopurinol is a classic trigger. SJS (A) brings mucosal erosions, blistering, and a positive Nikolsky sign; AGEP (B) produces sterile pustules with rapid onset; a benign morbilliform eruption (D) lacks the organ involvement and eosinophilia — the dangerous trap here; SSSS (E) is a different entity.
Why residents miss it: DRESS is dismissed as a simple drug rash because the skin looks morbilliform; the long latency and systemic features (eosinophilia, transaminitis, facial edema) are the clues residents overlook.
Reference: Kroshinsky D, et al. Drug Reaction with Eosinophilia and Systemic Symptoms. N Engl J Med. 2024.
A 68-year-old woman with chronic venous insufficiency has 5 days of bilateral lower-leg erythema, warmth, scaling, and itching to the mid-shin. She is afebrile with normal vital signs and a normal white-cell count. What is the most appropriate management?
Bilateral "cellulitis" is usually stasis dermatitis
Bilateral lower-extremity "cellulitis" is almost always stasis dermatitis (or contact dermatitis), not infection — cellulitis is overwhelmingly unilateral and usually febrile. The chronicity, scaling, itching, and normal vitals point to dermatitis, treated with a topical steroid, emollients, and compression (E). Antibiotics (A, B, D) treat a non-infection; simultaneous bilateral cellulitis (C) is rare.
Why residents miss it: Red, warm legs are reflexively called cellulitis, driving large amounts of unnecessary antibiotics; the distinguishing clues (bilateral, afebrile, itchy, scaly, chronic) are overlooked.
Reference: Raff AB, Kroshinsky D. Cellulitis: A Review. JAMA. 2016.
A 70-year-old man with nonvalvular atrial fibrillation (CHA2DS2-VASc 3) on warfarin needs an elective colonoscopy with anticipated polypectomy. He has no prior stroke and no mechanical valve. What is the most appropriate periprocedural anticoagulation plan?
Periprocedural bridging in atrial fibrillation
In atrial fibrillation without high thromboembolic risk (no recent stroke, no mechanical valve), the BRIDGE trial showed that forgoing heparin bridging is noninferior for thromboembolism and roughly halves major bleeding. So interrupt warfarin without bridging (B). Bridging (C, E) causes net harm in this population; continuing warfarin through a polypectomy (A) invites bleeding; D is illogical.
Why residents miss it: Bridging dogma persists even though the evidence reversed it; the decision requires weighing thrombotic against bleeding risk, and residents often defer it to consultants and never build the framework.
A patient with prior tuberculous pericarditis has progressive dyspnea, an elevated JVP with a prominent y-descent, a pericardial knock, and a Kussmaul sign. Echo shows respiratory variation in mitral inflow and a septal "bounce," with normal ventricular wall thickness. Which finding most specifically distinguishes this from restrictive cardiomyopathy?
Constrictive pericarditis versus restrictive cardiomyopathy
Constriction and restriction share many features, but enhanced ventricular interdependence — discordant respiratory variation of LV and RV systolic pressures — is the most specific sign of constriction (D). Elevated JVP with a rapid y-descent (A) and near-equalized diastolic pressures (C) occur in both and are less specific; BNP (B) tends to be higher in restriction; increased wall thickness (E) suggests an infiltrative or restrictive process such as amyloidosis.
Why residents miss it: These cases are uncommon, and the classic teaching point (pressure equalization) is nonspecific; residents conflate it with the truly specific finding, ventricular interdependence.
Which patient most clearly meets current USPSTF criteria for annual low-dose CT lung cancer screening?
Lung cancer screening eligibility (USPSTF)
The USPSTF recommends annual low-dose CT for adults aged 50–80 with at least a 20 pack-year history who currently smoke or quit within the past 15 years. Only C (age 55, 25 pack-years, current smoker) meets all three. A falls short on pack-years; B quit more than 15 years ago; D is over age 80; E is under age 50.
Why residents miss it: The precise cutoffs — age 50–80, 20 pack-years, and a 15-year quit window — are easily blurred, and the 2021 update lowered the age and pack-year thresholds from the prior version.
A 78-year-old man with early dementia refuses a recommended below-knee amputation for wet gangrene. He can state his diagnosis, that surgery is advised, and that without it he could die, and he explains that he would "rather die with both legs than live as an amputee." His daughter (his health-care proxy) wants the surgery. What is the most appropriate action?
Capacity is decision-specific
Capacity is decision-specific and is present when a patient can understand the information, appreciate how it applies to them, reason about the options, and express a choice — all of which he does — regardless of a dementia diagnosis. His refusal stands (A). A surrogate acts only when the patient lacks capacity, so B and D are wrong; capacity is a clinical determination that does not require a court (C) or a psychiatrist (E), and "competence" is a separate legal term.
Why residents miss it: A diagnosis of dementia is wrongly equated with global incapacity; residents do not apply the four-part, decision-specific capacity standard and default to families or psychiatry.
Reference: Appelbaum PS. Assessment of Patients' Competence to Consent to Treatment. N Engl J Med. 2007.
On hospital day 6 of unfractionated heparin for a DVT, a patient's platelet count falls from 250 to 70 x10^9/L, and a new lower-extremity arterial thrombus is found. The 4T score is high. What is the most appropriate management?
Heparin-induced thrombocytopenia
This is heparin-induced thrombocytopenia with thrombosis. All heparin must stop and a non-heparin anticoagulant (such as argatroban, fondaparinux, or a DOAC) must start, because the prothrombotic state persists (E). Warfarin alone (B) can precipitate venous limb gangrene and must wait until the platelets recover; simply stopping heparin (C) leaves the patient unprotected; LMWH (D) cross-reacts; and platelet transfusion (A) is avoided, with heparin continued.
Why residents miss it: The misconception that "stopping heparin is enough" — plus the warfarin-alone and platelet-transfusion errors — are common; HIT management is counterintuitive and usually run by hematology.
A 62-year-old hypertensive man has sudden vertigo, diplopia, dysarthria, and gait ataxia. A non-contrast head CT is normal. What is the most appropriate next test?
Posterior-circulation (basilar) stroke and imaging
Vertigo, diplopia, dysarthria, and ataxia together point to the posterior (vertebrobasilar) circulation. Non-contrast CT is insensitive in the posterior fossa, so a normal scan is falsely reassuring; CT (or MR) angiography is needed to evaluate the basilar and vertebral arteries, and basilar occlusion is time-critical (C). Repeating CT (A) wastes time; carotid duplex (B) images the wrong territory; echocardiography (D) is a later step; pursuing peripheral vertigo (E) is the classic miss.
Why residents miss it: Posterior-circulation strokes are mistaken for benign peripheral vertigo, and the insensitivity of non-contrast CT in the posterior fossa is underappreciated, so the wrong vascular imaging is ordered.
Reference: Merwick Á, Werring D. Posterior circulation ischaemic stroke. BMJ. 2014.
A 67-year-old woman has completed 5 years of oral alendronate for osteoporosis (initial hip T-score −2.8). She has had no fractures; her current hip T-score is −2.2 and her FRAX is not in the high-risk range. What is the most appropriate next step?
Bisphosphonate drug holidays
After about 5 years of an oral bisphosphonate in a patient who is now at modest, no-longer-high fracture risk, a drug holiday with periodic reassessment is reasonable, balancing efficacy against the small risks of atypical femur fracture and osteonecrosis (B). Indefinite therapy (A) and escalation to denosumab (C) or teriparatide (D) are not warranted at this risk level — and denosumab rebounds if later stopped. Stopping with no monitoring for 5 years (E) abandons follow-up.
Why residents miss it: The drug-holiday concept and the need to re-stratify fracture risk are poorly taught, so residents reflexively continue or escalate therapy.
Reference: Black DM, Rosen CJ. Postmenopausal Osteoporosis. N Engl J Med. 2016.
A patient with proteinuric diabetic kidney disease is started on lisinopril. Two weeks later the serum creatinine has risen from 1.6 to 1.9 mg/dL (about 19%), with a stable potassium of 4.8 mmol/L. What is the most appropriate action?
Creatinine rise after starting an ACE inhibitor
A creatinine rise up to about 30% that then stabilizes, with an acceptable potassium, is an expected hemodynamic effect of starting a renin-angiotensin-system inhibitor in proteinuric CKD and should not prompt stopping the drug — the long-term benefit is renoprotection (D). Discontinuing (A), dose-reducing (B), or switching to an ARB (C, which acts the same way) all forfeit that benefit; renal-artery imaging (E) is reserved for a larger rise or hyperkalemia.
Why residents miss it: The expected creatinine bump is misread as nephrotoxicity, so a renoprotective drug is stopped — one of the most common avoidable errors in CKD care.
An asymptomatic 50-year-old man has a single hemoglobin A1c of 6.6% (diagnostic threshold 6.5% or higher) on a routine screen. He has no symptoms of hyperglycemia. What is the most appropriate next step to establish a diagnosis of diabetes?
Confirming a diagnosis of diabetes
In an asymptomatic patient, a single abnormal result does not establish diabetes; the diagnosis requires two abnormal results — either repeating the same test or obtaining a second, different test (A). Diagnosing or treating off one value (B, C) is premature; a 3-year interval (D) applies to a normal screen, not a near-threshold abnormal one; and a random glucose (E) is diagnostic only at 200 mg/dL or higher with symptoms.
Why residents miss it: The shorthand "A1c of 6.5 means diabetes" skips the requirement for confirmation, and the caveats of A1c (anemia, hemoglobinopathy, CKD) are easily forgotten because residents mostly manage already-diagnosed diabetes.
Preparing for the boards and want a second set of eyes on your weak areas? Dr. Bray mentors residents and international medical graduates at no cost.
Get in touch →