CMC practice questions, on the November 2026 plan

Exertional syncope with sinus tachycardia in severe AS comes from a fixed outflow obstruction. The heart cannot raise output to match the vasodilation of exercise, so cerebral perfusion falls.

Complete heart block would show on telemetry as AV dissociation, not sinus tachycardia.

A diuretic may add to low preload, but the exertional timing and severe stenosis point to the fixed obstruction.

Vasovagal syncope involves a trigger, a prodrome and bradycardia. It does not fit exertion with tachycardia.

BMI and LDL are not metabolic syndrome criteria, and a BP of 128/78 is below the threshold. This set meets none.

A waist of 98 cm is under the male cutoff and LDL is not a criterion. Only the BP of 134/86 counts, which is one of three.

Metabolic syndrome needs 3 of 5 criteria: waist over 102 cm in men, triglycerides of 150 or more, HDL below 40 in men, BP of 130/85 or more, and fasting glucose of 100 or more. A waist of 106, triglycerides of 180 and HDL of 35 give three.

Glucose of 96, HDL of 52 and triglycerides of 140 are all on the normal side of the cutoffs.

Hyperphosphatemia is the problem of untreated kidney failure, which CRRT removes. After 48 hours of therapy, phosphate runs low.

High potassium is one of the problems CRRT corrects. Without replacement, potassium tends to fall, not rise.

Continuous hemodiafiltration steadily removes small solutes, including phosphate, so hypophosphatemia is common after a day or two unless it is replaced. It can weaken the respiratory muscles and delay weaning.

Serum sodium tracks the sodium in the replacement fluid. CRRT does not drive it upward.

Atropine acts on the AV node. It does not help infranodal block and is not given to a patient with a normal heart rate.

Amiodarone treats tachyarrhythmias and can worsen conduction disease.

A new bifascicular block during an anterior MI carries a real risk of complete heart block.

A new RBBB with left axis deviation, or bifascicular block, during an anterior STEMI shows septal ischemia that can progress suddenly to complete heart block. Transcutaneous pacing pads are placed in advance.

Amiodarone loading does not prevent atherectomy-related bradycardia and can worsen it.

Lidocaine prophylaxis for ventricular arrhythmias is not standard and does not address the bradycardia risk.

Rotational atherectomy of the RCA often causes bradycardia and heart block from debris and the right-sided conduction supply. Temporary pacing on standby is the key preparation.

Routine IABP is not required for atherectomy. It is reserved for selected high-risk anatomy or poor LV function.

Controlled BP is acceptable. Uncontrolled hypertension above about 180/110 would be the problem.

EECP squeezes the legs to push blood back toward the heart, which can dislodge a recent DVT. Active or recent DVT is a contraindication.

Prior bypass surgery is common in patients referred for EECP and is not a contraindication.

Compensated HF with an EF of 40 percent is usually acceptable. Decompensated HF is the concern.

Closing the defect once pulmonary vascular disease is fixed causes acute RV failure. It is not done in Eisenmenger syndrome.

Extra IV fluid does not raise SpO₂ in a fixed shunt and can strain the right ventricle.

Chronic cyanosis from the shunt does not correct with oxygen, and a baseline SpO₂ in the 80s is expected. Chasing 95 percent adds risk without benefit.

With right-to-left shunting through the VSD, any air bubble in a vein can cross into the systemic circulation and cause a stroke. In-line air filters on every IV line prevent paradoxical air embolism.

An SVR of 1100 is within the normal range of about 800 to 1200, so afterload is not the main problem.

A high PAOP with a low stroke volume, while SVR and heart rate are near normal, means the ventricle is filled but not ejecting. That is a contractility problem, as in systolic heart failure.

Hypovolemia would show a low PAOP. A wedge of 22 mm Hg means the left heart is full.

High preload alone would come with a normal or high stroke volume. A stroke volume of only 40 mL despite high filling pressure points to pump failure.

Vericiguat and riociguat both stimulate soluble guanylate cyclase. Taking them together risks severe hypotension and is contraindicated.

Vericiguat was studied on top of standard therapy that included sacubitril/valsartan.

Vericiguat can be given with an SGLT2 inhibitor such as dapagliflozin.

An MRA such as spironolactone is part of standard HFrEF therapy and is compatible with vericiguat.

Atropine at bedtime is not a treatment for OSA-related bradyarrhythmias and has anticholinergic harms.

An EP study is not needed when there is a clear reversible cause and the daytime ECG is normal.

Pauses and AV block only during sleep in untreated severe OSA come from vagal surges during apneas. CPAP treats the cause and usually abolishes them.

A pacemaker is not needed for asymptomatic nocturnal block that resolves with CPAP.

The effect develops over weeks to months, not within 24 hours.

Home BP monitoring is needed to judge the response and adjust medications.

Renal denervation lowers BP modestly and gradually, as an adjunct. Current BP medications are continued, with home BP monitoring guiding any later change.

Stopping all medications after a week overestimates a modest effect and risks rebound hypertension.

Nesiritide is a natriuretic peptide that dilates veins and arteries, lowering preload. Its intended effect in warm-and-wet HF is a fall in PAOP and improved dyspnea.

Nesiritide is not a negative chronotrope, and a large fall in HR is not its target.

A rise in SVR is the opposite of nesiritide's arterial vasodilation.

A rising cardiac index with unchanged PAOP describes an inotrope, not a vasodilator.

A faster VVI rate paces more often and worsens the syndrome.

Capture is not the problem. The symptoms come from lost AV synchrony.

Ventricular pacing with intact AV conduction causes atrial contraction against closed valves, which produces cannon a waves, neck pulsations and a BP fall. This is pacemaker syndrome, and restoring AV synchrony with dual-chamber pacing fixes it.

A beta-blocker slows the sinus rate and increases ventricular pacing.

500 to 600 mL/h is an intermittent dialysis pace that a patient with acute HF and AKI cannot tolerate.

300 to 400 mL/h outpaces refilling and invites hypotension and kidney hypoperfusion.

20 to 40 mL/h barely dents a CVP of 18 and delays decongestion.

For an 80-kg patient with a MAP of 70, a starting net ultrafiltration rate of about 1 to 2 mL/kg/h, or 100 to 150 mL/h, removes fluid while the vascular space refills. It is adjusted to hemodynamic tolerance.

AV nodal blockers can suppress inducibility during an EP study. They are usually held for about five half-lives so the arrhythmia can be induced and mapped.

Amiodarone has a long half-life and would suppress the arrhythmia for weeks.

Taking both drugs the morning of the study can make the tachycardia impossible to induce.

Diltiazem also slows the AV node, so continuing it defeats the purpose.

Overt hyperthyroidism needs a high free T4 or T3, and these are normal.

Subclinical hypothyroidism is a high TSH with a normal free T4. This choice reverses the direction.

Nonthyroidal illness usually shows a low T3 with variable TSH in a sick inpatient. It does not fit an isolated low TSH in a patient with AF.

A low TSH with normal free T4 and T3 is subclinical hyperthyroidism. A TSH below 0.1 in an older adult with AF is usually treated.

Rehab exercise does not explain the cold intolerance, constipation and high TSH.

Ezetimibe rarely causes myopathy and is not mentioned as part of the regimen.

Hypothyroidism raises the risk of statin myopathy and itself causes muscle aches and a high CK. Cold intolerance, constipation and a high TSH point to it.

Low-dose aspirin does not increase the risk of statin myopathy.

Extra oxygen all shift treats a resting state that is already adequate. The fall comes from demand, not low arterial oxygen.

Starting an inotrope for a short, activity-related dip that resolves with rest goes too far.

Transfusion raises oxygen content, but nothing suggests anemia, and ScvO₂ recovers on its own at rest.

A brief fall in ScvO₂ during activity that recovers with rest shows higher oxygen consumption that a limited cardiac output cannot match. Spacing activities with rest periods prevents repeated oxygen debt.

Transmural scar is abnormal at rest and does not improve with dobutamine.

Hibernating myocardium starts out hypokinetic at rest, improves at low dose and then worsens at peak dose, which is the biphasic response.

A wall that is normal at rest and low dose, then becomes hypokinetic at peak stress, has adequate flow at rest but a flow-limiting stenosis under demand. This is inducible ischemia.

A normal response is increasing contraction at higher doses, not new hypokinesis.

Isoproterenol raises heart rate and oxygen demand. Chest pain with ST depression in a patient with CAD is ischemia, so the infusion is stopped.

Increasing it to induce SVT worsens the ischemia. Reaching the study goal does not outweigh demand ischemia.

Nitroglycerin while continuing the cause treats the symptom and keeps the trigger running.

This is sinus tachycardia driven by the drug, not SVT. Adenosine would only briefly slow it.

Cryoablation near the right pulmonary veins can injure the right phrenic nerve. A weakening diaphragm contraction during phrenic pacing means injury is starting, and freezing must stop at once.

Transient sinus bradycardia can occur with vagal stimulation during ablation and is usually brief.

Inferior ST elevation can come from air embolism, which is a concern, but it is not what the phrenic pacing check is designed to find.

A falling esophageal temperature signals esophageal cooling, which matters with left-sided ablation near the esophagus, not phrenic injury.

With a CVP of 3 and PAOP of 8, no congestion remains. More diuretic worsens the volume depletion.

Adding another inotrope does not fix an empty ventricle. Dobutamine also vasodilates.

Raising the milrinone dose adds vasodilation and deepens the hypotension.

Low CVP and PAOP after a large diuresis mean the hypotension comes from too little preload plus milrinone's vasodilation. Pausing the milrinone and giving a small fluid bolus restores filling.

If the pump slipped back into the aorta, the placement signal would show an aortic waveform, not a ventricular one.

Thrombosis raises motor current and causes hemolysis. It does not change the placement signal to ventricular.

A ventricular placement signal with a flat motor current means both the inlet and outlet are in the LV. The pump is too deep and must be pulled back.

Correct position shows an aortic placement signal with a pulsatile motor current. A ventricular signal is never the correct position.

Immediate cath lab activation is reserved for STEMI, cardiogenic shock, recurrent arrhythmia or ongoing ischemia after ROSC.

A balloon pump supports a failing circulation. This patient is hemodynamically stable and needs no circulatory support.

For a comatose but stable survivor without ST elevation, trials such as COACT and TOMAHAWK found no benefit from immediate angiography. Angiography is deferred while care focuses on temperature, oxygen and blood pressure.

Fibrinolysis has no role without a STEMI or a confirmed PE and adds bleeding risk after CPR.

A long PR and 2:1 block after slow-pathway ablation suggest injury to normal AV conduction, a known complication. It is reported, and pacing is prepared.

Heart block is not an expected result of AVNRT ablation. It is a complication.

Adenosine blocks AV conduction further and can cause complete block.

Metoprolol slows AV conduction and could worsen the block.

Continuing the same rate leaves the patient overanticoagulated and at bleeding risk.

A bolus with a higher rate is the response to a low aPTT. This reverses the direction.

Protamine is for serious bleeding or a full reversal need. A high aPTT without bleeding is managed by dose adjustment.

An aPTT far above range without bleeding is handled by the nomogram: hold the infusion for about an hour, then restart at a lower rate and recheck in about 6 hours.

An access-site hematoma causes local swelling or bleeding, not a new murmur and pulmonary edema.

Sudden pulmonary edema with a new loud holosystolic apical murmur hours after edge-to-edge repair suggests clip detachment or leaflet tear with acute severe mitral regurgitation.

A small residual atrial septal defect from the transseptal puncture is common and rarely causes acute pulmonary edema.

Tamponade causes hypotension and muffled heart sounds with JVD, not a new loud regurgitant murmur.

A higher dose is not an accepted way to overcome caffeine and adds side effects.

Caffeine blocks adenosine receptors and blunts the coronary vasodilation that regadenoson produces, which can make the scan falsely negative. Caffeine is avoided for 12 to 24 hours before, so the test is rescheduled.

Holding a beta-blocker does not reverse caffeine's blockade of adenosine receptors.

Caffeine does affect regadenoson: it competes at the same adenosine receptors. Proceeding risks a falsely reassuring scan.

Paced beats have a pacing spike and a fixed rate. Nothing suggests a pacemaker.

A wide beat with RBBB pattern ending a short cycle after a long one is Ashman phenomenon. The right bundle is still refractory after the long cycle and conducts aberrantly. The absent compensatory pause also favors a supraventricular beat.

PVCs usually have no consistent long-short trigger and often show other morphologies. They are not tied to cycle-length changes.

Bigeminy means every other beat is ectopic. This pattern is occasional and tied to cycle length.

Carbon monoxide gives a falsely normal or high SpO₂ and cherry-red blood, not a fall to 85 percent with brown blood.

Motion artifact does not change blood color or persist with a stable waveform.

PE lowers the PaO₂. Here the arterial oxygen is normal.

Benzocaine can oxidize hemoglobin to methemoglobin. SpO₂ drifts toward 85 percent and does not respond to oxygen while the PaO₂ is normal, and the blood looks chocolate-brown. Co-oximetry confirms it, and methylene blue treats it.

Hypokalemia causes U waves and an apparent long QT, not a short ST segment.

A shortened ST segment and QTc in a patient with cancer who is confused, constipated and polyuric point to hypercalcemia. Malignancy is a common cause.

Hypocalcemia lengthens the ST segment and QT. This option reverses the calcium-QT relationship.

Low magnesium lengthens the QT and predisposes to torsades.

Fluid adds to an already high left-sided pressure and pulmonary flow.

In a post-MI VSD with low output, IABP lowers afterload so more blood flows forward and less is shunted left to right. It bridges the patient to closure.

A beta-blocker reduces contractility in a low-output state and can cause collapse.

Phenylephrine raises afterload and increases the left-to-right shunt.

A change from LBBB to RBBB paced morphology means the lead tip is now pacing the LV, so it has passed through the septum or RV wall. Loss of capture and pleuritic pain support perforation.

Lead maturation slightly raises thresholds over weeks. It does not change the paced QRS morphology or cause pain.

Hyperkalemia can raise capture thresholds but does not flip the paced bundle-branch pattern or cause pleuritic pain.

A temporary pulse generator's battery does not change paced morphology, and it is replaced at the bedside.

A normal exam, oxygen saturation and BNP do not support an emergency PE work-up.

Aspirin is part of DAPT after a stent. Stopping it alone is not the patient's decision.

Stopping ticagrelor two weeks after a stent risks stent thrombosis for a benign side effect.

Mild, brief dyspnea is a known ticagrelor side effect, with normal oxygen saturation, lungs and BNP. The patient keeps taking it and reports it at follow-up, because stopping risks stent thrombosis.

Continuing unchanged ignores the label's LVEF threshold for interrupting therapy.

Mavacamten is a myosin inhibitor that can reduce systolic function. An LVEF below 50 percent with new HF symptoms calls for holding it and rechecking echo, as its safety program requires.

Increasing the dose would further reduce contractility in a patient whose LVEF has already fallen.

Verapamil also depresses contractility, and adding it does not address the systolic dysfunction.

Glucose shifts most during rewarming, so checks are needed then.

Raising insulin during rewarming adds to the risk of hypoglycemia.

Insulin sensitivity rises and glucose falls as the body rewarms, so hypoglycemia is a risk. Glucose is checked more often and the insulin dose is reduced.

Keeping the same plan ignores the expected change in insulin sensitivity.

Cultures are reasonable, but sending them without treating the fever leaves the brain exposed to hyperthermia.

Re-cooling to 33 °C is not indicated. The goal is to prevent fever, not to restart hypothermia.

Fever control continues to matter after the device phase ends. Ignoring a temperature of 38.3 °C in a comatose patient risks further brain injury.

Fever after the temperature-control phase is linked to worse neurologic outcome. In a comatose patient, active fever prevention is resumed, typically keeping temperature at or below 37.5 °C.

The BP is normal, so this is not dialysis hypotension, and a saline bolus does not address cerebral edema.

Headache, nausea, restlessness and confusion during a first dialysis run with a very high BUN suggest dialysis disequilibrium. Rapid urea removal leaves the brain relatively hyperosmolar and water shifts in. The run is slowed or stopped and the provider notified.

The glucose is 112 mg/dL, which rules out hypoglycemia as the cause of the confusion.

A faster blood pump removes urea even faster and worsens the osmotic shift into the brain.

Reprogramming the lower rate does not restore a depleting battery and can drain it faster.

Waiting for symptoms risks loss of pacing, especially in a pacemaker-dependent patient.

Leads are working, with normal capture and sensing. The battery, not the lead, needs replacing.

A magnet rate that falls to the manufacturer's elective replacement value means the battery is near the end of its life. A planned generator change is scheduled before pacing becomes unreliable.

Stopped output with signs of tamponade means a blocked drain, not a resolved effusion. Removing it takes away the route for drainage.

Furosemide lowers preload in a patient whose heart is compressed, which can cause collapse.

Fluid can support preload briefly, but it does not relieve tamponade.

Drain output that stops abruptly, with hypotension, tachycardia and pulsus paradoxus, suggests the drain is blocked and fluid is reaccumulating. An urgent echo confirms tamponade and guides drainage.

Turning off pacing can be dangerous in a dependent patient. Mode changes are programmed per protocol.

MRI with a cardiac device is possible when the generator and leads are MR-conditional. The first step is to confirm the device's status and then program it as the manufacturer requires.

A magnet does not reliably make a device MRI-safe, and the magnetic field itself can trigger unpredictable modes.

Many devices are MR-conditional, so canceling the scan for good denies a useful test.

Platelet clumping on the smear with no bleeding suggests EDTA-induced pseudothrombocytopenia. A repeat count in a citrate tube confirms the true count.

Holding aspirin for a falsely low count can increase ischemic risk in a patient with chest pain.

There has been no heparin exposure, so HIT is not a reasonable concern.

Transfusing for a lab artifact exposes the patient to risk with no benefit.

B12 deficiency also causes macrocytic anemia and neuropathy. It does not explain high-output failure.

High-output heart failure with warm extremities and lactic acidosis in a malnourished drinker suggests wet beriberi. Thiamine deficiency causes vasodilation and impaired aerobic metabolism, and IV thiamine reverses it quickly.

Folate deficiency causes macrocytic anemia, not a high-output state with lactic acidosis.

Cardiac output is already 9.2 L/min. An inotrope adds nothing to a high-output state caused by vasodilation.

NSAIDs and colchicine treat pericarditis, which this choice confuses with immune-mediated myocarditis.

Continuing the checkpoint inhibitor allows a potentially fatal myocarditis to progress. GDMT alone does not treat the inflammation.

Conduction disease may need temporary pacing if it progresses, but a permanent device is not the first step in an inflammatory process that may reverse.

Rising troponin with new conduction disease weeks after starting a checkpoint inhibitor, with normal coronaries, means immune checkpoint inhibitor myocarditis. It is treated urgently with high-dose IV corticosteroids and withdrawal of the checkpoint inhibitor.

Arterial embolism makes the limb pale, cold and pulseless, but without massive swelling.

Cellulitis causes red, warm skin, not deep cyanosis with reduced arterial pulses.

Massive swelling of the entire leg, intense pain, purple-blue color and fading pulses mean extensive venous thrombosis is now blocking arterial inflow. This is phlegmasia cerulea dolens, a limb-threatening emergency often linked to cancer.

Uncomplicated DVT causes swelling and pain but does not compromise arterial flow or turn the leg blue.

For symptomatic paroxysmal AF in a younger patient with normal LV function, catheter ablation can be a first-line rhythm-control option under the 2023 ACC/AHA AF guideline.

Ablation works best in paroxysmal AF, not permanent AF.

Drug failure is not required before ablation in selected patients.

Anticoagulation is guided by stroke risk score after ablation, not by rhythm success.

Waiting until January stretches the maintenance interval to 9 months.

May would be only one month after the April dose, which is far too soon.

Inclisiran is given at day 0 and 3 months, then every 6 months. After doses in January and April, the next dose is due in October.

July would be the 3-month interval repeated, which applies only to the loading phase.

Arrhythmic death occurs at night too, and the device must be worn during sleep.

The device decides on its own when a shock is needed. No button delivers a shock on request.

Feeling better does not reduce early post-MI arrhythmic risk. The device is worn until LVEF is reassessed.

A wearable defibrillator protects only while it is worn, and sudden death risk is highest early after an MI. It is worn day and night, removed only briefly for bathing.

Patients with COPD and hypercapnia during a breathing trial are at high risk of failing extubation. Planned NIV applied right after extubation reduces reintubation.

An SpO₂ goal of 96 percent or more in COPD risks hyperoxic hypercapnia. The usual target is about 88 to 92 percent.

High-flow oxygen through a nonrebreather in hypercapnic COPD can worsen CO₂ retention and does not support ventilation.

Waiting for distress before starting NIV misses the preventive benefit and delays reintubation if it fails.

Ezetimibe is commonly combined with statins without this interaction.

Fenofibrate has much less effect on statin levels. It is the fibrate used with a statin when one is needed.

Prescription omega-3 products can be safely added to a statin.

Gemfibrozil blocks statin clearance and raises the risk of myopathy and rhabdomyolysis. It is contraindicated with simvastatin.

Internal trigger is asynchronous and meant for cardiac arrest or no cardiac output. It is not used with a paced rhythm.

When pacing spikes are counted as R waves, the balloon times erratically. In a 100 percent paced rhythm, the pacer trigger mode uses the spike to time inflation reliably.

Lowering augmentation volume does not address timing errors from a bad trigger.

Reducing the ratio to 1:3 cuts support without fixing the trigger problem.

Adding a vasodilator drops BP more and does not treat a complicated dissection.

Rising pain, falling BP and falling urine output despite medical therapy mark a complicated type B dissection with malperfusion or impending rupture. Urgent endovascular repair is indicated.

Lowering BP further worsens malperfusion when pressure is already falling.

Continuing the same therapy for 4 more hours delays repair while organs are already underperfused.

A CT trip for an unstable patient comes after initial stabilization and may be unsafe.

A large fluid bolus can overdistend a failing right ventricle in PE. Only cautious volume is used.

Sudden dyspnea, hypoxemia and hypotension in a patient with DVT suggest a massive PE. Oxygen comes first while the team prepares for reperfusion.

Raising the legs does nothing for hypoxemia and may push more clot.

A prior bleeding peptic ulcer is a major bleeding risk factor during dual antiplatelet therapy and favors a proton pump inhibitor.

Smoking and a prior MI raise ischemic, not bleeding, risk.

Diabetes raises ischemic risk more than bleeding risk.

Age 66 with controlled BP is not high bleeding risk. Age 75 or older is the usual threshold.

A conscious wearer who hears the alarm presses and holds the response buttons. That tells the device the patient is awake and prevents an unneeded shock.

Turning the monitor off removes protection and does not fix the cause of the alarm.

Removing the vest while it is alarming takes away protection if a real arrhythmia is under way.

Lying down to wait for a shock accepts a painful, unnecessary shock when the patient is conscious and well.

NAPA, procainamide's active class III metabolite, is cleared almost entirely by the kidney. In stage 4 CKD it builds up and lengthens the QT, so a NAPA level, read with the parent-drug level, shows accumulation.

Thyroid tests matter for amiodarone, not procainamide. This choice confuses two antiarrhythmics.

Procainamide toxicity in CKD comes from renal accumulation of NAPA. Liver tests do not measure it.

Potassium and magnesium are worth checking in any long-QT patient, but they cannot show drug accumulation.

Unstable angina has ischemic symptoms without troponin rising above the 99th percentile. Here troponin rose to 30.

Ischemic symptoms with a hs-troponin that rises from below to above the 99th percentile, and no ST elevation, meet the definition of NSTEMI.

Chronic myocardial injury shows a stable, persistently high troponin, not a rise from 4 to 30.

Pericarditis causes positional pain and diffuse ST elevation, not localized ST depression with a troponin rise.

In permanent AF, rapid conduction competes with pacing, so biventricular pacing falls. AV node ablation ensures near-100 percent pacing when rate control fails.

Diltiazem depresses contractility in HFrEF and is avoided.

The LV lead works when it paces. The problem is how often it paces, not where it sits.

Generator replacement does not change pacing percentage in AF.

An LVEF of 55 percent is still normal and reflects the left ventricle, not the right.

TAPSE measures how far the tricuspid annulus moves toward the apex in systole. A fall from 20 mm to 13 mm, below the normal of about 17, shows declining RV systolic function.

A smaller IVC that collapses more suggests lower right atrial pressure. If anything, that is the opposite of a failing RV.

RV wall thickening is hypertrophy, an adaptation to pressure load. It describes wall structure, not how well the RV contracts.

A magnet stops shocks but would also disable therapy for a real VT storm. It is used only after the rhythm shows the shocks are inappropriate.

An anxiolytic may help after the cause is known, but it does not tell whether the shocks were appropriate.

Several shocks in a conscious patient may be appropriate therapy for an electrical storm or inappropriate shocks for AF, noise or lead failure. The first step is to look at the rhythm, which decides what comes next.

Potassium and magnesium matter for storm, but checking the rhythm comes first.