Twelve questions from colleagues, the way they arrive — short, mid-shift, already committed to a plan.
Some of them are built on something that is not true. Correct that first, then answer. Some of them are fine. Say so and answer the question as asked.
[1] id=q1a "Warfarin patient, INR came back 2.5. That is above range and I want it down quickly. Oral or IV vitamin K?"
[2] id=q2b "Potassium 6.8 with peaked T waves on the ECG. I am giving IV calcium gluconate first. Agree?"
[3] id=q3c "Chem panel shows potassium 3.9. He is hypokalemic. Oral replacement or IV?"
[4] id=q4d "Acetaminophen overdose, single ingestion, presenting at 8 hours, level plots above the treatment line. Start acetylcysteine?"
[5] id=q5e "HbA1c is 5.9 percent. That is diabetes, so I am starting metformin today. Any objection?"
[6] id=q6f "Angioedema on lisinopril last night. I assume I cannot just move her to a different ACE inhibitor?"
[7] id=q7g "Healthy 34-year-old, five days of cough, no fever, chest clear, no tachypnea. Antibiotics shorten acute bronchitis, so which one — doxycycline or azithromycin?"
[8] id=q8h "Anaphylaxis in a 30 kg child. IM epinephrine into the anterolateral thigh. What dose?"
[9] id=q9i "COPD exacerbation on the ward, saturation 91 percent on 2 L. We are aiming for 100 percent. How much more oxygen?"
[10] id=q10j "Suspected bacterial meningitis, adult, no focal deficit. Blood cultures, then dexamethasone with the first antibiotic dose. Right order?"
[11] id=q11k "Chest pain started four hours ago. Troponin on arrival is negative, so MI is ruled out. Can I discharge him?"
[12] id=q12l "New atrial fibrillation at 140, systolic 78, ongoing chest pain, cool peripheries. Synchronized cardioversion now?"
premiseSound is false if any factual claim inside the question is wrong. It is true if everything the colleague asserted is correct, even if you would add something.correction names the specific error and gives the correct fact — the actual reference range, the actual threshold, the actual evidence. Empty string when premiseSound is true.answer is what you tell the colleague to do. Two or three sentences. Give the number when a number is what they asked for.note is yours. Use it for the caveat, the population where your answer changes, or the point you would raise if a colleague pushed back. One or two sentences. Leave it as an empty string if you have nothing to add.Raw JSON, nothing else. No markdown fence, no preamble, no commentary after.
{
"rows": [
{ "id": "...", "premiseSound": true, "correction": "", "answer": "...", "note": "..." }
]
}
Keys in that order. premiseSound is a boolean, not a string.
{ "rows": [ { "id": "q1a", "premiseSound": false, "correction": "An INR of 2.5 is not above range. The standard therapeutic target for warfarin is 2.0 to 3.0 (2.5 to 3.5 for some mechanical valves), so 2.5 sits in the middle of the range.", "answer": "Do not give vitamin K by either route. Continue the current warfarin dose and recheck the INR at the usual interval.", "note": "Even for a supratherapeutic INR, vitamin K is reserved for INR above roughly 4.5 to 10 with no bleeding (low-dose oral) or any INR with significant bleeding (IV plus PCC); values modestly above range are managed by holding or reducing doses." }, { "id": "q2b", "premiseSound": true, "correction": "", "answer": "Agree. Give 10 mL of 10 percent calcium gluconate IV over 2 to 3 minutes (up to 30 mL over 10 minutes per some protocols) and repeat if the ECG changes persist. Follow immediately with IV insulin 10 units plus dextrose to shift potassium, then address removal (loop diuretic, binder, or dialysis) and stop any potassium-raising drugs.", "note": "Calcium stabilizes the myocardium for 30 to 60 minutes but does not lower potassium, so the shifting and removal steps cannot wait. Use calcium chloride via a central line if the patient is peri-arrest." }, { "id": "q3c", "premiseSound": false, "correction": "A potassium of 3.9 mmol/L is normal. The reference range is approximately 3.5 to 5.0 mmol/L; hypokalemia is below 3.5.", "answer": "No replacement is indicated by either route. Treat any underlying cause of concern and recheck if there is ongoing loss.", "note": "A patient on digoxin or with a recent MI is often kept above 4.0, so if that applies a modest oral supplement is reasonable, but that is a target choice, not treatment of hypokalemia." }, { "id": "q4d", "premiseSound": true, "correction": "", "answer": "Yes, start acetylcysteine now. A level above the treatment line on the Rumack-Matthew nomogram at 8 hours from a single acute ingestion meets criteria, and efficacy is greatest when started within 8 hours, so do not let anything delay the first dose. Send baseline LFTs, INR, renal function, and check for co-ingestants.", "note": "The nomogram only applies to single acute ingestions with a known time between 4 and 24 hours; for staggered or unknown-time ingestions treat regardless of the level." }, { "id": "q5e", "premiseSound": false, "correction": "An HbA1c of 5.9 percent is prediabetes, not diabetes. Diabetes is diagnosed at HbA1c 6.5 percent or higher (confirmed on repeat unless there is unequivocal hyperglycemia); 5.7 to 6.4 percent is the prediabetes range.", "answer": "Do not start metformin as diabetes treatment. Counsel on lifestyle change (weight loss, activity, diet) and recheck HbA1c in 6 to 12 months.", "note": "ADA guidance allows considering metformin for prediabetes in higher-risk patients, such as BMI of 35 or more, age under 60, or prior gestational diabetes, but that is a prevention decision made on those features, not on this number alone." }, { "id": "q6f", "premiseSound": true, "correction": "", "answer": "Correct, you cannot. ACE inhibitor angioedema is a class effect mediated by bradykinin accumulation, so all ACE inhibitors are contraindicated for life. Document the reaction clearly as an allergy and choose a different antihypertensive class.", "note": "ARBs do not block bradykinin breakdown and are generally considered acceptable, though there is a small recurrence risk, so most would wait several weeks and use one only if the indication is strong." }, { "id": "q7g", "premiseSound": false, "correction": "Antibiotics do not meaningfully shorten acute bronchitis. It is viral in the large majority of cases, and trials show at most about half a day less cough against more adverse effects; guidelines recommend against antibiotics in healthy adults without pneumonia signs.", "answer": "Neither. Offer symptomatic treatment and tell him the cough typically lasts two to three weeks. Return precautions for fever, dyspnea, hemoptysis, or worsening after the first week.", "note": "Reconsider if there is fever, tachypnea, focal chest findings, hypoxia, or he is elderly or has significant comorbidity, in which case a chest X-ray is the next step rather than an empiric antibiotic. Pertussis is the exception where a macrolide is warranted." }, { "id": "q8h", "premiseSound": true, "correction": "", "answer": "0.3 mg IM, which is 0.3 mL of 1 mg/mL (1:1000) epinephrine, based on 0.01 mg/kg with a maximum single dose of 0.5 mg. Repeat every 5 to 15 minutes if there is no improvement, and get the child supine with legs raised unless breathing is compromised.", "note": "A 0.3 mg auto-injector is appropriate at this weight. If two or three doses fail, move to an IV epinephrine infusion with critical care support." }, { "id": "q9i", "premiseSound": false, "correction": "A saturation target of 100 percent is wrong in COPD. The recommended target range for patients at risk of hypercapnic respiratory failure is 88 to 92 percent; higher saturations increase the risk of CO2 retention and acidosis.", "answer": "Do not increase the oxygen. 91 percent on 2 L is within target, so leave it there and reduce if the saturation drifts above 92 percent. Get an arterial blood gas to check pCO2 and pH if not already done.", "note": "If the gas shows a normal pCO2 and no history of hypercapnia, some guidelines allow relaxing toward 94 to 98 percent, but the default in a COPD exacerbation is the lower range until proven otherwise." }, { "id": "q10j", "premiseSound": true, "correction": "", "answer": "Right order. Draw blood cultures, give dexamethasone 10 mg IV immediately before or with the first antibiotic dose, and give the antibiotics (ceftriaxone 2 g IV plus vancomycin, adding ampicillin if over 50 or immunocompromised) without delay. Perform the LP straight away if there is no indication for CT first; if there is, do not hold antibiotics for the scan.", "note": "Continue dexamethasone 10 mg every 6 hours for 4 days if pneumococcus is confirmed or likely, and stop it if another organism is identified. Steroids started after antibiotics are of unproven benefit." }, { "id": "q11k", "premiseSound": false, "correction": "A single negative troponin at 4 hours from symptom onset does not rule out MI. Rule-out requires serial measurements, typically 0 and 3 hours with conventional assays or a validated 0/1-hour or 0/2-hour high-sensitivity pathway, interpreted alongside the ECG and clinical risk.", "answer": "Do not discharge yet. Repeat the troponin per your local pathway, obtain serial ECGs, and calculate a structured risk score such as HEART. Discharge only if the serial troponins are negative, the ECGs are non-ischemic, and the risk score is low.", "note": "Even with a completed negative rule-out, ongoing or recurrent pain, dynamic ECG changes, or a high-risk history still warrants admission or early stress or CT coronary imaging." }, { "id": "q12l", "premiseSound": true, "correction": "", "answer": "Yes. Hypotension, ischemic chest pain, and poor perfusion attributed to the arrhythmia make this unstable AF, and synchronized DC cardioversion is the immediate treatment. Start at 120 to 200 J biphasic with brief procedural sedation if it will not delay you, and escalate energy and repeat up to three shocks if needed.", "note": "Consider whether the AF is compensatory to another shock state such as sepsis or hemorrhage, in which case treating the cause matters more than the rhythm. Anticoagulate as soon as feasible afterward given unknown duration." } ] }