Anesthesia & Critical Care

At two in the morning an ambulance wheels in a young man from a car crash — pale, a bruise on his neck, all that is left of his consciousness a moan. The intern reaches to start an IV, the nurse calls for a CT scan first, but the attending grips his hand still and says, back to A first. This is the most beautiful thing about this lifesaving alphabet: the order is not rote memory but an arrangement by death speed. Airway obstruction kills within minutes, so it comes first; the inability to breathe kills within minutes, so it comes second; hemorrhagic shock kills within tens of minutes, so it comes third; an intracranial event can hold on a little longer; hypothermia and a missed wound worsen over hours, so they come last.

Airway is never only airway. For blunt trauma, falls, or car crashes with any complaint of neck pain or numbness, always apply a rigid cervical collar before moving the patient, because if a fractured cervical vertebra shifts during transfer, a patient who could once walk and move may never stand again — so airway and cervical spine are two faces of the same coin. The intubation line is drawn at a Glasgow Coma Scale of eight, because behind that line is the question of whether the patient can protect his own airway — coughing, swallowing, clearing secretions, all beyond reach below a score of eight. So even an eighteen-month-old who is still breathing on his own should be intubated once his score falls to six, because the airway can collapse at any moment; breathing does not equal a safe airway, and this is the most vicious trap in the exam hall.

When it comes to chest trauma, the two most lethal injuries never rely on imaging. Tension pneumothorax is a disease of pressure crushing both the heart and the opposite lung, so the affected side has no breath sounds, the trachea deviates toward the healthy side, the jugular veins are pushed full and distended by the pressure, and the blood pressure falls — see this combination and go straight to needle decompression; never wait for that chest X-ray film, because by the time it comes back the patient may already be nearly gone. Massive hemothorax looks very similar, with no breath sounds on the affected side as well, but the blood has all leaked into the chest cavity, so there is less blood returning and the jugular veins collapse instead — this is the one dividing line between the two, whether pressure has pushed the veins full or blood loss has let them collapse. In the trauma setting, cardiac tamponade is managed first by surgery; pericardiocentesis is only a temporary bridge, not the final stop. For blunt abdominal trauma with suspected bleeding, do a bedside ultrasound first, and only send a stable patient on to CT; abdominal X-ray is nearly useless for hemoperitoneum and is the exam's favorite distractor — do not take the bait.

The ethics section is easily misread. DNR does not mean do not treat — a DNR declines futile CPR, not life-sustaining treatment for an acute, reversible event, so a patient who has signed a DNR but develops respiratory failure from a car crash should still be intubated, because his current respiratory failure is reversible. Transfer to a trauma center also looks at physiologic and anatomic high-risk indicators; an isolated open tibial fracture, though it needs surgery, will not kill on the way there, so it does not meet the transfer threshold. Triage classification runs on the same logic — a respiratory rate of forty per minute is an immediate life threat and takes top priority, while an agitated patient who is still cursing at everyone still has intact cerebral perfusion and oxygenation and ranks lower. This looks counterintuitive until you reason it through death speed, and then it makes sense. The whole chapter comes down to one line to remember: the sequence is, at its core, death speed, and a deteriorating patient always goes back to A.

That young man from the car crash, back in the resuscitation bay: blood pressure one hundred ten, pulse one eighteen, respirations twenty-six, skin cold and clammy, mind anxious. The intern glances at the blood pressure and relaxes, saying it's fine; the attending shakes his head, because a normal systolic pressure doesn't mean he's fine — his pulse pressure has already been squeezed narrow. What makes hemorrhagic shock so fascinating is that the body does not collapse the moment it senses blood loss; it has an entire compensatory system that holds the line first and only lets the blood pressure fall once it can hold no longer. So if you look only at the systolic pressure, you will misjudge him as fine — the real skill is watching pulse pressure, heart rate, urine output, and mental status, the earlier alarms.

The four classes are not rote memory — they are a timeline of when compensation collapses. Once blood loss begins, the sympathetic system tightens the vessels and speeds the heart, so diastolic pressure is pushed up first and pulse pressure narrows first — the earliest fingerprint starting at Class II. By the time systolic pressure can no longer hold and starts to fall, the patient has already reached Class III, with blood loss exceeding thirty percent, and confusion means cerebral perfusion has finally run short. So a systolic pressure that is still holding cannot be taken to mean everything is fine — treat narrowing pulse pressure plus a racing heart plus falling urine output plus an anxious mind as the earlier compensatory alarms. The exam loves to hand you the pair urine output five to fifteen, heart rate one twenty, and ask you to pick the class — the answer is Class III, not Class II.

The most terrifying thing about massive traumatic hemorrhage is not how much blood is lost, but this vicious cycle of hypothermia plus acidosis plus coagulopathy — once the three lock together they multiply each other into a snowball, each one dragging the next one down. Hypothermia inhibits thrombin activity, acidosis inhibits clotting factors, and coagulopathy in turn keeps the patient bleeding, growing colder and more acidotic still. The exam loves to swap in hypotension or arrhythmia in place of coagulopathy to trick you, but neither of those is a member of the triangle — coagulopathy is. The upgraded version is called the lethal diamond, adding hypocalcemia as a fourth corner, because during massive transfusion the citrate in blood products chelates calcium ions, and calcium is an essential player in the coagulation cascade and myocardial contraction — so massive transfusion requires monitoring and replacing ionized calcium, and that is the diamond's fourth corner.

The modern approach to massive traumatic hemorrhage is called damage control resuscitation, and its core spirit is to stop the bleeding first and not push the blood pressure too high. Why — because before a clot has had time to firm up, forcing the blood pressure up to one twenty will only wash away the freshly formed clot, dilute clotting factors, and worsen the bleeding, so the guideline calls for permissive hypotension, with a target systolic pressure of about eighty to ninety, enough to palpate a radial pulse and maintain consciousness. The second counterintuitive move is to give less crystalloid — a large volume of normal saline floods the body with chloride, causing hyperchloremic metabolic acidosis and actually pushing the lethal triad forward, so the first choice is a balanced solution or going straight to ratio-based transfusion. The third is that red cells, fresh frozen plasma, and platelets should be given at a ratio of one to one to one, because replacing red cells alone dilutes the clotting factors, which is feeding the third corner of the lethal triad with your own hands. The fourth is that the antifibrinolytic must be given within three hours to reduce mortality; past that window it loses its effect.

When the blood type is unknown, emergency red cell transfusion uses type O, with O-negative chosen for women of childbearing age to prevent Rh antibody sensitization, because type O red cells carry no A or B antigens on their surface and any recipient can accept them. But the rule for plasma runs exactly opposite — type AB is the universal plasma donor, because type AB plasma contains no anti-A or anti-B antibodies; the two directions are reversed, and the exam loves to trip you up with exactly this. Distinguishing fluids and acid–base disturbances is also an easy point to bank: a large volume of normal saline gives you hyperchloremic metabolic acidosis, vomiting away hydrochloric acid from the upper GI tract gives you metabolic alkalosis, excessive mechanical ventilation blowing off too much carbon dioxide gives you respiratory alkalosis, and sepsis or poor perfusion shifting to anaerobic metabolism gives you lactic acidosis. The whole chapter comes down to one line to hold onto: the body will hold the line for you until the very last moment, but you cannot wait for the blood pressure to fall before you act.

A thirty-two-year-old motorcyclist is wheeled in with a large scalp laceration and a Glasgow Coma Scale of eight; the CT shows a lens-shaped white opacity against the inner table of the temporal bone. Neurosurgery is calling for an emergency operation when the on-call resident suddenly realizes the patient was still talking at the roadside — how is he down to moaning by the time he reached the hospital? The attending turns back and says that is the lucid interval, the signature of arterial bleeding. To understand intracranial pressure, remember just one sentence: the cranial vault is a rigid box of fixed volume, holding brain tissue at about eighty percent, blood at about ten percent, and cerebrospinal fluid at about ten percent; the box is already full, so whenever one component grows, something else must yield space, and once that yielding is exhausted, pressure rises exponentially and forces out herniation.

Normal intracranial pressure is seven to fifteen, and anything above twenty requires treatment. When the pressure is about to burst the box, the body sounds its own alarm automatically, presenting as Cushing's triad — blood pressure rises, heart rate slows, respirations become irregular — because the brain raises blood pressure to defend its own perfusion, and baroreceptors reflexively slow the heart rate in response. Intracranial pressure by itself is only one variable; what truly determines whether the brain becomes ischemic is the pressure arriving through the cerebral arteries minus the counter-pressure inside the skull, and this is the formula for cerebral perfusion pressure — cerebral perfusion pressure equals mean arterial pressure minus intracranial pressure, and mean arterial pressure is systolic pressure plus twice the diastolic pressure, divided by three, because the heart spends longer in diastole than in systole; systolic pressure can never be substituted directly for it, and this is the most common calculation error in the exam hall. Following the formula naturally yields two paths — lowering intracranial pressure or raising mean arterial pressure can both improve cerebral perfusion, and every treatment falls into one of these two paths.

The Glasgow Coma Scale adds three subscores together; the motor score is taken from the best-performing limb, so a fractured side is assessed on the opposite side instead, and an intubated patient's verbal score is written as V T and must never be given an arbitrary number. The ladder for lowering intracranial pressure runs from noninvasive to invasive: first elevate the head thirty degrees with the head and neck in the midline, promoting venous return and lowering intracranial pressure — the cheapest step and done first. Next comes sedation and analgesia, avoiding fever and seizure to lower cerebral metabolic demand. Then comes osmotic dehydration — mannitol or hypertonic saline — pulling water out of the tissue and into the vessels to be carried away. Hyperventilation is a double-edged sword: a falling carbon dioxide partial pressure constricts the cerebral vessels, reduces cerebral blood flow, and lowers intracranial pressure, but dropping it to twenty causes over-constriction and cerebral ischemia, and the effect exhausts itself in under a day, with rebound dilation once it is stopped — so it is only an emergency bridge against herniation, with the target held at thirty to thirty-five. Only refractory cases proceed to external ventricular drainage, barbiturate coma, or decompressive craniectomy. The contraindication the exam most loves to reverse is that prophylactic steroids are contraindicated after head injury — the CRASH trial has already shown that steroids actually increase mortality in head injury and must never be used to reduce cerebral edema. Head injury without seizure does not routinely receive long-term antiepileptic drugs; once a seizure truly progresses to status epilepticus, the first line is a benzodiazepine, not propofol, and this is a common exam distractor.

Intracranial hemorrhage's two shapes sit behind two entirely different pieces of vascular anatomy. Epidural hemorrhage comes from an artery, especially the middle meningeal artery, often with a temporal bone fracture, progressing quickly with a lucid interval, because although arterial bleeding is fast, it still takes time to accumulate enough to deform the brain tissue, and during that window consciousness temporarily recovers; once accumulation passes the critical threshold, the patient deteriorates rapidly, with unequal pupils and contralateral hemiplegia, appearing lentiform on CT and not crossing suture lines. Subdural hemorrhage comes from a vein, especially torn bridging veins, progressing slowly and able to drag on into a subacute or even chronic course, favoring elderly patients with brain atrophy and those on anticoagulation, because their shrunken brain tissue leaves more buffer space and blood can accumulate slowly over weeks, often presenting as progressive dementia and unsteady gait and easily mistaken for a diagnosis of dementia, appearing crescent-shaped on CT and able to cross suture lines, spreading along the convexity.

The most classic presentation of subarachnoid hemorrhage is the worst thunderclap headache of one's life plus neck stiffness, with over eighty percent arising from rupture of a berry aneurysm. It is an acquired lesion — the arterial wall has a congenital lack of a tunica media muscle layer, and only combined with long-term hemodynamic shear stress does a thin-walled sac develop at a branch point of the circle of Willis, most commonly the anterior communicating artery. Diagnosis starts with a non-contrast CT; only when it is negative but suspicion remains high does a lumbar puncture look for xanthochromia to confirm whether old blood is present, and once confirmed, CTA or DSA locates the aneurysm — the lumbar puncture is used to confirm whether bleeding occurred, not to find its cause. The most lethal complication afterward is vasospasm, typically occurring on the fourth to the fourteenth day, peaking on day seven, not the first to the fifth day, prevented and treated with nimodipine. Cerebrospinal fluid is produced at a rate of about zero point three five milliliters per minute, about twenty milliliters per hour, and an exam value of forty milliliters is too high and wrong. The whole chapter comes down to one line: once the box is full, something must yield — reason it through in that direction, and every intervention makes sense.

An ambulance from a mountain road brings in a middle-aged man bitten by a snake, his left calf swollen like a loaf of bread, his family fumbling out a length of rope, saying they already tied it tight for him. That same night, the ICU admits a woman pulled from a house fire, her nasal hairs singed black, her voice hoarse, yet her oxygen saturation reads ninety-eight. Both stories are traps of looking fine while actually being lethal.

Taiwan's six major venomous snakes are first divided by toxin type into three classes: the hemotoxic type includes the Taiwan habu, the brown spotted pit viper, and the hundred-pace pit viper; the neurotoxic type includes the banded krait and the cobra; the mixed type is Russell's viper alone. The more exaggerated the local swelling, the more likely it is hemotoxic; when the local wound looks unremarkable yet the patient gasps, the eyelids droop, and swallowing becomes difficult, it is more likely neurotoxic. The banded krait is the most classic neurotoxic trap — the fang marks are tiny, the pain mild, and the patient looks fine at first, only to develop sudden respiratory muscle paralysis hours later — never let such a patient leave the hospital early. Taiwan currently stocks four types of corresponding antivenom; the bivalent hemotoxic antivenom covers only the Taiwan habu and the brown spotted pit viper — the hundred-pace pit viper, although hemotoxic, is not covered by this bivalent antivenom and needs its own dedicated antivenom, and Russell's viper likewise needs its own dedicated antivenom, while the bivalent neurotoxic antivenom covers the banded krait and the cobra — this is the most frequently tested point in the matching trap.

Antivenom dosing is determined by clinical severity, not body weight, and the dose is not reduced for children. Why? Because the antivenom needs to neutralize "the amount of venom the snake injected," not the patient's body weight. A small child has a smaller body, so the same amount of venom reaches a higher concentration in him and may actually be more severe — the dose must never be reduced just because the patient weighs less. Reassess after administration, and if there is progression, give more, titrating to response. Antivenom is a heterologous protein that can cause allergic reactions or even anaphylactic shock, so adrenaline must be ready before administration.

The fundamental skill for burns is first getting the depth and area right. First-degree is only epidermal redness and is not counted into the total area; second-degree, with blisters or pallor, is counted; third-degree, waxy white or charred across the full thickness, with nerve destruction that makes it painless instead, must also be counted. The adult Rule of Nines gives the head nine, each upper limb nine, the anterior and posterior trunk eighteen each, and each lower limb eighteen, with children having a proportionally larger head. Scattered burns are estimated using the patient's entire palm as equal to one percent. The Parkland formula is four milliliters times kilograms times percentage, as a twenty-four-hour volume, using lactated Ringer's solution with no dextrose, giving half in the first eight hours and the other half over the next sixteen, counted from the time of injury rather than the time of arrival, with a target urine output for adults of zero point five per kilogram per hour. With electrical burns or concurrent rhabdomyolysis, when the urine runs dark red, the target urine output should be raised to one to one point five, to flush the renal tubules and prevent myoglobin-induced acute kidney injury. Albumin should wait until eight to twenty-four hours have passed and crystalloid alone proves insufficient, because early capillary permeability is high and any colloid given would simply leak into the tissue and be wasted.

Chemical burns are generally irrigated with copious water, but there are two exceptions worth remembering. Lime and cement are strong alkalis, not acids, so irrigating directly generates heat with the water and actually worsens the burn — the dry powder must be brushed off first, then irrigated. Hydrofluoric acid, though an acid, has unique toxicity: its fluoride ions chelate calcium and can cause fatal hypocalcemia, so after irrigation a calcium gluconate gel must be applied to neutralize it, with serum calcium and ECG monitored. Alkali causes liquefactive necrosis, dissolving the tissue and boring deeper as it goes, while acid causes coagulative necrosis, whose eschar actually limits deeper penetration, so alkali burns run deeper and are harder to manage than acid burns. Escharotomy incises only the skin without entering the fascia, indicated when a circumferential full-thickness burn constricts a limb or the chest wall and impairs perfusion or ventilation; fasciotomy must incise down to the fascia, indicated for compartment syndrome — the two differ both in depth of incision and in indication.

Inhalation injury is the invisible killer. Fire in an enclosed space, a singed face or nasal hairs, a hoarse voice, carbonaceous sputum, stridor, and soot in the oropharynx are all warning signs. The management principle is one line: since the airway will progressively swell, intubate early while it can still be done, because waiting for complete obstruction is too late. The diagnostic gold standard is fiberoptic bronchoscopy, directly visualizing the mucosal injury; chest X-ray is often normal early on, has extremely low sensitivity, and is not the first choice — this is a very easy option to pick wrongly. The most fatal trap in carbon monoxide poisoning is that the pulse oximeter reading can be falsely normal, because it compares only two wavelengths and cannot distinguish carboxyhemoglobin from oxyhemoglobin, and since carboxyhemoglobin also absorbs red light the reading skews high, appearing normally oxygenated while the tissue is in fact severely hypoxic — carboxyhemoglobin must be measured directly to reveal the truth, and treatment is one hundred percent oxygen, with hyperbaric oxygen when needed. Burn patients exist in a hypermetabolic state: a net body weight loss of ten percent impairs healing, twenty percent raises the risk of infection, thirty percent markedly increases major complications, and beyond forty percent survival itself is threatened, so early high-protein, high-calorie nutritional support is essential. The whole chapter comes down to one line: behind every counterintuitive intervention there is a mechanism explaining exactly why it works that way.

At four in the morning, paramedics carry in a seventy-year-old man — pulseless out-of-hospital cardiac arrest for four minutes, CPR in progress. The moment the resident takes over compressions, the monitor shows coarse ventricular fibrillation, and she shouts for the defibrillator to be charged. In the adjoining ICU, a middle-aged woman with septic shock is wheeled in at the same time, oxygen saturation ninety-one, blood pressure eighty-five, lactate five point two. Two timelines, one measured in minutes, the other in hours, yet both held up by the same underlying logic.

CPR changed from A B C to C A B, compressions first, and the logic is that coronary perfusion pressure matters more urgently than ventilation — every interruption zeroes out coronary perfusion pressure, so interruptions must be minimized. The compression rate is one hundred to one hundred twenty, not just at least one hundred — there is an upper limit, because going too fast prevents full recoil and actually harms outcome; depth is five to six centimeters in adults, about a third of chest depth in children; the ratio is thirty to two when not intubated, and once intubated it becomes continuous compressions plus one breath every six seconds — never over-ventilate, because too much air raises intrathoracic pressure, lowers venous return, and lowers cardiac output. Compression quality is monitored using the quantitative carbon dioxide waveform once intubated — below ten millimeters of mercury means poor compressions, and a sudden spike often signals the return of spontaneous circulation. Rhythm triage must be kept straight: ventricular fibrillation and pulseless ventricular tachycardia are shockable rhythms, requiring immediate defibrillation plus adrenaline plus amiodarone; pulseless electrical activity and asystole are non-shockable rhythms — no shock, only adrenaline plus a search for reversible causes. Immediately resume compressions for two minutes after a shock before re-checking the rhythm; do not stop to look at the ECG, which wastes precious perfusion time.

Pediatric defibrillation dosing is an exam favorite: the first shock two joules per kilogram, the second four, and every shock after that no less than four, up to ten, never exceeding the adult dose — remembered as start at two, follow with four. Adrenaline is given as one milligram every three to five minutes; refractory ventricular fibrillation uses amiodarone at an initial dose of three hundred milligrams, then one hundred fifty, and the first-line antiarrhythmic is not lidocaine. Reversible causes must be searched for using five H's and five T's — hypoxia, hypovolemia, acidosis, potassium abnormality, hypothermia, and tension pneumothorax, cardiac tamponade, toxins, thrombosis — and without searching for a reversible cause, no amount of compression time will help. On the legal side, Taiwan's do-not-resuscitate order applies only to terminal patients, requiring confirmation by two relevant specialists plus either the patient's own advance directive or a closest relative's consent form, with the directive carrying more authority than the consent form; if the patient has not yet been determined terminal, such as in acute choking asphyxiation, the do-not-resuscitate order does not apply, and the physician should resuscitate first — a verbal statement on the spot by someone without legal authority, or a document not personally signed at this hospital, is never sufficient to withhold resuscitation.

The four major types of shock are sorted along three axes, asking whether the patient is dry, weak, blocked, or has a vasculature gone slack. Only distributive shock presents with warm extremities, vasodilation, and a fall in peripheral resistance; the other three all show compensatory vasoconstriction with cold extremities. Mixed venous oxygen saturation falls in low-cardiac-output shock, because tissue is desperately extracting oxygen, but runs paradoxically high in sepsis, because impaired cellular oxygen utilization plus arteriovenous shunting mean oxygen cannot get into the cells, leaving plenty of oxygen left over on the venous side. The order for septic shock is to find the source of infection, give early broad-spectrum antibiotics within one hour, give crystalloid at thirty milliliters per kilogram, and only start a pressor if still hypotensive — the first choice is norepinephrine, predominantly alpha with a touch of beta, raising peripheral resistance without excessively increasing heart rate; the second line adds vasopressin, and dobutamine is added only when cardiac output is inadequate, with mistakenly choosing dopamine or epinephrine as the first choice both common errors. The mean arterial pressure target is sixty-five, and an initial lactate of four or greater, or the need for a pressor, already qualifies as critical illness.

The most frequently tested nutrition trap is that early in shock, hemodynamics are unstable and gut perfusion is inadequate, so aggressive enteral nutrition at this point can worsen gut ischemia or aspiration — so earlier and more aggressive is always better is wrong, and hemodynamics must be stabilized first before nutrition is started. Ventilator weaning is judged by the rapid shallow breathing index, equal to respiratory rate divided by tidal volume — below one hundred five is suitable for extubation, above one hundred five is likely to fail, which is simply how winded quantified. The cost of positive-pressure ventilation is a rise in intrathoracic pressure, a fall in venous return, and a fall in cardiac output; the real pressure threshold for barotrauma is a plateau pressure greater than thirty, not ten — ten falls far short, and that is the exam's favorite wrong answer. Discontinuing either total parenteral nutrition or enteral nutrition can cause rebound hypoglycemia and requires gradual tapering; the glucose target of one hundred forty to one hundred eighty should not be made too strict. One last detail to close with: if asymptomatic carotid stenosis is severe, sixty to seventy percent or greater, and surgical risk is low, endarterectomy or stenting can still be considered — it is not always managed with medical therapy alone. The whole chapter comes down to one line: what stands between life and death is simply getting the right order of what must be done.

In the operating room, an anesthesiology resident is about to infiltrate a wound with local anesthetic in a 60 kg woman. He calculates that 7 mg per kg times 60 equals 420 mg, so lidocaine with epinephrine should be plenty — but before he finishes the thought, the patient suddenly reports perioral numbness, tinnitus, and blurred vision, and then her muscles begin to twitch. The attending shouts systemic toxicity. Down the same hallway, in another operating room, a young woman receiving succinylcholine with inhalational anesthesia has an end-tidal CO₂ that suddenly jumps from 35 to 65, her masseter rigid, and the on-call nurse reaches for dantrolene. In this entire chapter on anesthesia, not a single intervention appears out of nowhere — every drug, every monitor, every moment of extubation follows directly from a mechanism.

The core mechanism of local anesthetics is plugging the voltage-gated sodium channel from inside the cell — the action potential cannot rise, and pain sensation is cut off; it is not calcium, not potassium. Nerve fibers that are thin in diameter and myelinated are blocked first, so autonomic sympathetic fibers are lost first, then pain and temperature sensation, then touch and pressure, and motor function last; recovery runs in reverse, with motor function returning first. Inflamed tissue is acidic, so a higher fraction of the drug becomes ionized and crosses the membrane poorly — which is why local anesthetic injected into an abscess so often fails, and there is a real reason for that, not simply an inadequate dose. The naming clue is that amides carry an extra i before "-caine," as in lidocaine, bupivacaine, ropivacaine, and mepivacaine, while esters usually carry only one i, as in procaine, tetracaine, and cocaine — though there are exceptions, since mepivacaine looks like it has only one i yet is an amide. So the most reliable approach is to remember the metabolic pathway: amides go through the liver, esters go through pseudocholinesterase, and reversing the two is wrong. For lidocaine, the limit is 4.5 mg/kg without epinephrine and 7 mg/kg with it; adding epinephrine constricts vessels, delays absorption, and lowers toxicity — it does not dilate them — and it must never be added at end arteries such as the fingers, toes, penis, nose tip, or ear pinna. In systemic toxicity, the central nervous system is poisoned first and the heart second: perioral numbness, a metallic taste, tinnitus, blurred vision, muscle twitching, and seizures come first, followed by arrhythmia and cardiovascular collapse — hallucination is not a typical feature. Bupivacaine is the most cardiotoxic, and the rescue is 20% lipid emulsion acting as a sink that pulls the lipophilic local anesthetic back out — lidocaine must never be given again to stack more toxicity on top. An epidural needle passes from superficial to deep — skin, subcutaneous tissue, supraspinous ligament, interspinous ligament, ligamentum flavum — and then a sudden loss of resistance marks entry into the epidural space; advancing further to puncture the dura reaches the site of spinal anesthesia, and the ligamentum flavum is that single "pop." The height of a spinal block is affected by baricity, dose, position, height, and obesity — obesity really does matter, because a rise in intra-abdominal pressure engorges the vertebral venous plexus, narrows the subarachnoid space, and lets the drug spread more widely.

Choosing the wrong analgesic often comes from not sorting out the mechanism first: pain from tissue inflammation responds to NSAIDs and opioids, while neuropathic pain needs a gabapentinoid, a tricyclic antidepressant, or an SNRI, and responds poorly to opioids and NSAIDs. Pregabalin binds the α2δ subunit of the voltage-gated calcium channel and lowers glutamate and substance P — it is not the sodium channel; the local anesthetic is the one that acts on the sodium channel, and the two are often mixed up. Opioid respiratory depression is mediated by the μ receptor, not κ; it acts on the medullary respiratory center and lowers sensitivity to carbon dioxide, and tolerance never develops to miosis or constipation — long-term use still produces both. Naloxone reverses opioids but has a short half-life and may be followed by re-narcotization, so it needs repeat dosing; flumazenil reverses benzodiazepines, taking effect in two minutes with a short duration of thirty to sixty minutes, so watch for re-sedation; neither ketamine nor dexmedetomidine can reverse opioid respiratory depression. Choosing a block by surgical site is never a single trick for every case: the gold standard for thoracotomy is an epidural, because the pain spans multiple intercostal levels, and the key to the epidural's systemic benefit is that it suppresses, rather than increases, sympathetic outflow — so bowel motility recovers faster, pulmonary complications fall, and opioid use drops. Pain after total knee arthroplasty is mainly anterior, so a femoral nerve or adductor canal block works, while a sciatic nerve block works worst because it covers only the posterior knee and calf. Cervical epidural injection is indeed used clinically, though its proximity to the spinal cord makes it high-risk and requiring image guidance. Enhanced recovery after surgery actively uses peripheral nerve blocks rather than avoiding them.

The preoperative physical status classification judges the severity of systemic disease, not the size of the surgery, and appending E marks an emergency. The mnemonic for difficult mask ventilation is mask seal, obesity or obstruction, age, no teeth, and snoring or stiff lungs; rheumatic heart disease affects the valves, not the upper airway, so it does not raise airway risk, and it is a favorite distractor on the licensing exam. Difficult intubation is judged by the soft palate and uvula: class I shows the most, and class IV, where only the hard palate remains visible, is hardest to intubate. The fasting rule is clears at two hours, milk at four to six, a light meal at six, and a fatty meal at eight. Applying a magnet to a cardiac pacemaker mostly switches it to asynchronous mode, which only reduces interference and cannot fully prevent it, so bipolar cautery is still required, kept away from the device, with external pacing on standby. The nerve injury most often reversed on exams is that lithotomy positioning stretches and injures the sciatic and common peroneal nerves, not the femoral nerve; femoral nerve injury comes from excessive hip flexion, and the most common anesthesia-related nerve injury is actually the ulnar nerve, from compression of the medial arm. The four high-risk factors for postoperative nausea and vomiting are female sex, non-smoking, a history of motion sickness or previous PONV, and postoperative opioid use — smoking is instead a protective factor. The surgical consent form protects patient autonomy; decision-making capacity is judged by whether the patient can understand, reason, and express a choice, not by a psychiatric diagnosis itself; the Declaration of Helsinki is not satisfied merely by regulatory approval, and it must weigh both the value of the treatment and the avoidance of harm; an inappropriate physician-patient relationship requires immediately ending the treatment relationship and referring the patient elsewhere.

Pulmonary edema in the extubation period is negative-pressure, not positive-pressure; the mechanism is that under light anesthesia the glottis reflexively closes when stimulated, the patient inhales forcefully against that closed glottis, intense negative intrathoracic pressure develops, and fluid leaks from the pulmonary capillaries. Neuraxial anesthesia also causes hypothermia, because it blocks sympathetic tone, dilates vessels, suppresses shivering, and lets the lower body lose heat, with consequences including coagulopathy, higher wound infection, more cardiac events, and slower drug metabolism. Malignant hyperthermia is triggered by an inhalational anesthetic plus succinylcholine: a mutation in the skeletal muscle ryanodine receptor gene lets calcium pour out of the sarcoplasmic reticulum, and the muscle stays contracted in a state of hypermetabolism; the earliest sign is a sharp rise in end-tidal CO₂, not a rise in temperature, which appears only late. Treatment is to stop the inhalational agent immediately, switch to non-triggering drugs, and give dantrolene starting at 2.5 mg/kg and repeatable up to a total of about 10 mg/kg, while cooling the patient and treating hyperkalemia, acidosis, and rhabdomyolysis. The first line for anaphylactic shock is adrenaline, and the common culprits during anesthesia are neuromuscular blockers, latex, and antibiotics; the rescue for local anesthetic systemic toxicity is 20% lipid emulsion. A laparoscopic CO₂ gas embolism calls for left lateral decubitus with head-down, feet-up positioning, letting the bubble float away from the right ventricular outlet and back toward the apex of the right atrium — right lateral decubitus is the wrong move and only pushes the bubble further toward the outflow tract. A patient who is awake, has capacity, has signed a do-not-resuscitate order, and refuses intubation should have that wish respected, with symptomatic relief provided and another specialty consulted to complete the terminal determination — intubating anyway violates autonomy.

Minimum alveolar concentration is the concentration at which 50% of patients show no movement in response to a skin incision; it is a measure of potency, and the lower it is, the more potent the agent — age, hypothermia, pregnancy, opioids, and acute alcohol all lower it. Desflurane has the fastest onset and emergence but is irritating, and rapidly raising its concentration activates the sympathetic system so that heart rate rises rather than falls, a classic exam distractor. Benzodiazepines lower cerebral oxygen consumption rather than raising it, and ketamine is the only agent that raises cerebral blood flow and intracranial pressure, so it should be used cautiously in patients with raised intracranial pressure. Pulse oximetry calculates the oxygenation ratio from two wavelengths, 660 and 940 nm, and carboxyhemoglobin makes the reading falsely elevated, so a saturation of 90% does not mean there is no hypoxia. Mixed venous oxygen saturation is normally 60 to 80%; a fall in cardiac output, hemoglobin, or arterial oxygen all lower it, and fever, pain, and shivering raise oxygen consumption and lower it too, but in sepsis it is instead elevated because of impaired cellular oxygen utilization and arteriovenous shunting, as covered in the previous chapter. A capnography waveform that suddenly drops to zero suggests disconnection, esophageal intubation, or cardiac arrest; a sharp rise suggests hypoventilation, malignant hyperthermia, or CO₂ absorption during laparoscopy; a sudden rise during CPR is usually the return of spontaneous circulation. Neuromuscular blockers divide into the depolarizing succinylcholine, which has no specific antagonist and is cleared by pseudocholinesterase, can cause hyperkalemia, and is contraindicated in burns, crush injury, or denervation, and the non-depolarizing agents such as rocuronium, which can be reversed with neostigmine plus an anticholinergic or chelated directly with sugammadex. A train-of-four ratio above 0.9 is the safety threshold for extubation; it assesses recovery from non-depolarizing blockade, and a normal tidal volume does not mean muscle strength has recovered enough — that ratio is the gold standard. Hold on to one sentence for the whole chapter: every intervention is saying the same thing — think through the "why" first, and the action grows out on its own.