Perioperative Medicine, the reference

Each item loses points only for its own problem. Awake, moving every limb and above 92% on air keeps 2 each; dyspnoea drops respiration to 1, so the total is one short of 10 before circulation is counted. Miller bands circulation by percentage of the pre-anaesthetic pressure; Barash prints the same bands in mmHg. A stated fall gives one grade; unstated hypotension leaves two possible totals. Pain and nausea are never scored. Miller requires 9 or more for phase 1 discharge; its sleep-apnoea checklist accepts less only with extra tests.

Any hypotension earns middle circulation point. Any hypotension is a partial deficit earning the middle circulation point, giving a higher total, only when the fall is in the middle band. A fall of half the baseline or more, such as four-fifths, scores zero; with dyspnoea the total is one below the higher total.

Miller's modified Aldrete table grades activity, respiration, circulation, consciousness and oxygenation at 2, 1 or 0. It bands circulation by percentage of the pre-anaesthetic pressure. The maximum is 10, and 9 or more permits phase 1 discharge. Nothing else is scored.

Sources: Miller 10e

Short-acting agents, regional blocks and sedation techniques mean many patients are fit for step-down care when they leave theatre; admitting them to the recovery unit only delays them with repeat observations. The modified Aldrete score checks activity, breathing, circulation, consciousness and oxygenation but ignores pain and nausea, which the first-stage unit usually manages. White and Song therefore added pain and nausea-vomiting categories. A patient still needing intravenous analgesia or antiemetics scores lower and belongs in the first-stage unit. Eligibility depends on the whole score, not one sign.

Train-of-four threshold indicates fast-track recovery readiness. A train-of-four threshold looks like a firm recovery gate. In fact it is a residual-block check made in theatre before extubation, and it is not an item on the Aldrete or fast-track scores. Movement is scored clinically, and no single sign, including oxygen saturation, permits bypass on its own.

Patients already meeting discharge criteria on reaching recovery may bypass to the second stage. The modified Aldrete score omits pain and PONV, so White and Song added both, crediting pain controlled with oral analgesics and penalising nausea needing parenteral treatment.

Sources: Miller 10eBarash 9e

The modified Aldrete score is out of 10. Miller's recovery table deems a patient fit to leave phase 1 at ≥9. Miller's sleep-apnoea box sets a lower bar of ≥8, but only alongside vital signs close to baseline, treated nausea, controlled pain and a passed room-air challenge. The thresholds therefore don't conflict: a score of ≥9 satisfies both, and the apnoea patient also needs the oxygenation test. No fixed minimum stay applies. Going home needs a responsible adult escort and a plan for pain.

Patient must be pain-free before discharge. People often assume a day-case patient must be pain-free before leaving, so this looks like the false statement. It is true. The criteria ask for pain controlled with oral analgesia, not abolished. Barash expects patients to cope with pain at home if they have adequate analgesia.

Miller's: modified Aldrete score ≥9/10 means fit to leave phase 1; the sleep-apnoea pathway accepts ≥8 plus a room-air challenge. No minimum stay is mandated; an adult escort is required. Barash accepts discharge with pain if analgesia is adequate.

Sources: Miller 10eBarash 9e

Patients with OSA have an upper airway that collapses easily. Sedatives, opioids and volatile agents add respiratory depression on top of that, so the danger lasts beyond emergence. A recovery score reached at one moment cannot show obstruction that comes back once stimulation fades, so an added observation period is needed. The guideline treats risk as ongoing until the patient keeps their own baseline saturation while breathing room air. It also advises recovering them lateral, semi-upright or in another nonsupine position.

Full Aldrete score permits immediate OSA discharge. Aldrete is the familiar discharge tool, so leaving on a score feels correct. But a low score is not a passing score. In OSA, even a full pass only starts the clock: patients need an extra hour's observation and baseline saturation on room air before leaving.

An anaesthesia textbook suggests keeping OSA patients in recovery for an extra 60 minutes after modified Aldrete criteria are met. The ASA sleep apnoea guideline regards them as at risk until they keep baseline room-air saturation, and favours nonsupine extubation and recovery.

Sources: FaustOutside our books: Practice Guidelines for the Perioperative Management of Patients with Obstructive Sleep Apnea: An Updated Report (2014)

To go home, a day-case patient must be comfortable, keep down oral fluids and be able to walk. Nausea and pain each block these criteria. They lead to rescue treatment, a longer recovery stay or an unplanned overnight admission. Barash names nausea, with or without vomiting, as the leading factor. Miller's names postoperative pain first. Both books agree these two preventable problems are the main causes. Faster awakening does not shorten the time until a patient is fit for discharge. So prevention should target nausea and pain together.

Residual sedation delays most discharges commonly. Leftover sedation seems like the obvious thing keeping a patient from going home. But neither textbook ranks it as a leading cause of delay. Agents that wake patients faster do not bring forward fitness for discharge. Nausea and pain are the causes the books name.

Barash calls nausea with or without vomiting probably the most important contributor to delayed discharge and unplanned admission; Miller's calls pain the most significant cause. A consensus guideline adds that PONV prolongs recovery-unit stay by about 1 hour and raises nursing workload and cost.

Sources: Barash 9eMiller 10eOutside our books: Fourth Consensus Guidelines for the Management of Postoperative Nausea and Vomiting (2020)

Patients are discharged once they meet safety criteria such as stable vital signs and walking. That is not the same as full recovery from the sedative and anaesthetic drugs. Leftover central nervous system depression therefore carries on at home and is felt as sleepiness by most patients. Sore throat and muscle aches mostly come from airway instrumentation and relaxants, so they affect only intubated patients. Modern antiemetic practice means nausea and vomiting continuing at home are less common than either.

Pain and nausea are main discharge complaints. Pain and nausea and vomiting are the problems taught most often as the main barriers in day surgery, so they seem the obvious answer. The source lists drowsiness well ahead of them. Aches and sore throat are common only in intubated patients, and nausea and vomiting are much less frequent.

Minor adverse effects are frequent after day-case anaesthesia. Drowsiness tops the list at 62%. In intubated patients sore throat affects 49% and aches 47%. Headache affects 25% and dizziness 20%, while nausea (17%) and vomiting (7%) are less common after discharge.

Sources: Miller 10e

Pressurised gas pushes the abdominal contents and diaphragm towards the head and compresses the lung bases. Dependent alveoli collapse, so FRC and vital capacity fall and the respiratory system gets stiffer. Airway pressures climb, and alveoli that are perfused but poorly ventilated increase V/Q mismatch. This happens in every patient and is worst in obesity. Absorbed CO2 separately adds a carbon dioxide load, causing hypercapnia within 15 to 30 minutes. Raising minute ventilation usually corrects it.

Hypercapnia invariably develops during pneumoperitoneum. A big rise in PaCO2 sounds right because insufflated CO2 is absorbed. The sources give no fixed percentage, though, and extra ventilation usually prevents the hypercapnia. The change that ventilator settings cannot avoid is worse V/Q mismatch.

The texts list higher V/Q mismatch and airway pressure with lower FRC, vital capacity and compliance. Obesity adds atelectasis and mismatch. CO2 absorption brings hypercapnia within 15 to 30 minutes, and more ventilation reverses it in most patients.

Sources: Barash 9eMiller 10e

Raised intra-abdominal pressure compresses the aorta, so afterload rises. Stimulating the peritoneal autonomic nerves also causes a sympathetic surge: catecholamines, renin–angiotensin activation and vasopressin all add vasoconstriction. Arterial pressure therefore rises because resistance has risen, not because flow has, and cardiac output falls as the gas goes in. Visceral perfusion falls as well. The insufflated carbon dioxide raises arterial CO2 and causes a respiratory acidosis. The respiratory system becomes stiffer, so compliance falls and peak airway pressure rises.

Elevated pressure reflects increased cardiac output. It is tempting because mean arterial pressure and SVR both rise during insufflation, and people read the higher pressure as higher flow. The rise actually comes from aortic compression and neurohumoral vasoconstriction loading the ventricle. Cardiac output is observed to fall.

One text explains the rise in SVR and MAP as aortic compression increasing afterload, and records a fall in cardiac output on insufflation. Another lists hypercarbia with acidosis, lower compliance with higher peak airway pressure, and reduced renal, portal and splanchnic flow.

Sources: Miller 10eBarash 9e

Tilting head-up lets blood pool in the legs, so venous return, filling pressures and stroke volume drop. Baroreceptors sense the fall and drive sympathetic output: arterioles constrict, raising SVR, and the heart speeds up. Under anaesthesia this tachycardia offset the smaller stroke volume, so cardiac index held steady; a sleeve gastrectomy series found heart rate also climbed with pneumoperitoneum, supine or tilted. As a complication, an abrupt preload loss can instead provoke severe hypotension with reflex bradycardia, managed by flattening the table and desufflating.

Head-up positioning increases preload via gravity. Preload is the variable tilting most obviously changes, so people get its direction wrong or carry over the head-down rule. Head-up does the opposite: blood falls away from the heart and filling pressures drop. The reflex tachycardia, not extra preload, is what preserves output.

A monitored laparoscopic study found head-up tilt lowered cardiac and stroke index and filling pressures while raising SVR; under anaesthesia a compensatory heart rate rise kept cardiac index stable. Faust warns abrupt preload loss may cause profound hypotension and reflex bradycardia.

Sources: Outside our books: The adverse hemodynamic effects of anesthesia, head-up tilt, and carbon dioxide pneumoperitoneum during laparoscopic cholecystectomy (2000)FaustOutside our books: The Impact of Intraoperative Position Changes on Hemodynamics and Cardiac Electrophysiological Balance Index in Patients with Severe Obesity Undergoing Laparoscopic Sleeve Gastrectomy (2026)

Head-down tilt moves leg blood towards the chest and head. Miller finds this autotransfusion raises output about 9% within a minute, but the rise fades and output drifts back towards baseline within about 10 minutes. In a normally filled patient the lasting response, Barash notes, is reflex vasodilation, lowering peripheral resistance; cerebral arterial pressure rises while leg perfusion pressure falls. For steep, sustained tilt, Miller's robotic chapter lists lower output and organ perfusion and higher myocardial oxygen use. So leg resistance falls; the output rise is transient.

Sustained rise in cardiac output. Tempting because head-down tilt plainly improves venous return from the legs, so it looks like a fluid challenge. In a normally filled patient, though, the central shift triggers reflex vasodilation, the early output rise fades within minutes, and steep sustained tilt can lower output and organ perfusion.

Miller lists, for steep head-down tilt in a normovolaemic patient, lower leg perfusion pressure, higher cerebral arterial pressure, more central blood volume, reduced output and organ perfusion, and greater myocardial oxygen use. Barash notes better venous return but also reflex vasodilation.

Sources: Barash 9eMiller 10e

The scale grades fitness for surgery by severity of coexisting illness, not the risk of the proposed operation, which must be judged separately. Classes rise with severity. Class 4 is severe disease constantly threatening life. Class 5 is a moribund patient not expected to survive without the operation. Class 6 is not a risk grade for a living patient; it is a brain-dead person whose organs are being retrieved for donation. Suffix E denotes an emergency and does not change the class number.

ASA 4 and 5 criteria are identical. Both ASA 4 and ASA 5 describe critically ill patients, so the wording is easy to mix up. A constant threat to life defines class 4. Class 5 is reserved for the moribund patient, whose survival depends on having the operation at all.

The ASA table lists six categories. They run from normal health through mild, severe and life-threatening systemic disease to the moribund patient and the brain-dead organ donor. An added E means emergency surgery. The original scheme had five classes and did not weigh procedural risk.

Sources: Miller 10eMayo review

In an anxious patient, a raised reading may be a white-coat surge rather than poor long-term control, particularly if an ACE inhibitor was skipped, as many patients are advised. Check technique and repeat the reading. Barash cites older literature delaying elective surgery when diastolic exceeded 110, but the cited study found no major morbidity in that group. Miller and Stoelting find no outcome link below 180/110 and no proof that delay helps. The threshold therefore stands, but delay above it is a judgement, not automatic.

Single raised reading warrants operative cancellation. Cancelling feels safe because high pressure is linked to cardiac events and older teaching delayed surgery above a diastolic cut-off. But one anxious reading does not confirm hypertension, and delay has no proven benefit. Postponement is only weighed for marked, confirmed hypertension or for end-organ injury that treatment could improve.

Stoelting notes day-of-surgery readings may be white-coat and that ACE inhibitors are often deliberately interrupted; confirm technique and recheck. Miller finds no risk link below 180/110 and no benefit from delay. Barash's own data undercut automatic delay above 110 diastolic. No book names an anxiolytic.

Sources: Stoelting 8eMiller 10eBarash 9e

Hypertension raises perioperative cardiovascular risk overall. That excess risk is not seen below 180 systolic or 110 diastolic, and no evidence shows that postponing surgery to tighten control lowers risk. The UK figure of 160/100 is not a competing cut-off. It covers documented community readings, which let a patient be accepted without further work-up. The 180/110 line marks where risk becomes evident. Above it, delaying surgery is a judgement: the gain from better control is set against the harm of waiting. Below it, you proceed.

Pressure below severe threshold needs escalated care. It feels like a safe middle course because hypertensive patients are known to be haemodynamically labile. Below the severe threshold, though, raised pressure carries no clear added cardiovascular risk. Nothing supports escalating postoperative care for that reason alone, so the books' answer is simply to proceed.

Below 180 systolic and 110 diastolic, hypertension shows no clear link with cardiac complications, and deferral has not been shown to lower risk. UK practice accepts community readings under 160/100. Above the severe threshold, the benefit of optimisation is weighed against the harm of delay.

Sources: Miller 10e

A thoracic curve restricts the lungs, and long-standing hypoxaemia can rarely drive pulmonary hypertension and cor pulmonale, which spirometry cannot reveal. Barash's spine-surgery chapter asks for ECG and echo before correction to define ventricular function and pulmonary pressures. Miller's general guidance discourages echo without a diagnostic question, and routine cardiology referral on curve size alone is low-yield. So for scoliosis surgery echo is the best test to pick, sharpened by symptoms, a murmur, a large curve, low vital capacity or an ECG change; stable patients having other surgery skip it.

Chest CT for the deformed thorax. A deformed thorax invites chest imaging, and CT looks thorough. But spirometry is already normal, so lung anatomy adds little, and CT tells nothing about pulmonary pressures or ventricular function, the hidden cardiac risk the spine chapter wants an echocardiogram to exclude. DLCO likewise adds nothing once spirometry is normal.

Barash advises ECG and echocardiography to assess ventricular function and pulmonary pressures, since scoliosis can rarely cause cor pulmonale. Miller finds echo useful for pulmonary hypertension yet discourages routine echo without a diagnostic question. A scoliosis cohort found Cobb-angle-based referral rarely changed findings or timing.

Sources: Barash 9eMiller 10eOutside our books: Low utility of pre-operative cardiology referrals in adolescent idiopathic scoliosis using a cutoff of Cobb > 70 degrees (2026)

A spinal curve stiffens the chest wall, causing a restrictive lung defect. A small vital capacity means a weak cough and little reserve after long prone surgery. Barash links a vital capacity below 30–35% of predicted with prolonged postoperative ventilation, so spirometry decides the extent of surgery and the need for ventilation. Chronic hypoxia from restriction raises pulmonary vascular resistance, so echo measures right heart strain. Miller advises selective testing, covering routine screening for other surgery; in scoliosis correction, the deformity is the reason to test.

Echocardiography should precede pulmonary function testing. Echo is tempting because scoliosis can strain the right heart, and texts do advise it before correction. But the restrictive lung defect affects every patient, and it decides how far the surgery goes and whether ventilation is needed afterwards. Echo follows lung testing. It does not replace it.

Miller says formal lung function testing guides how extensive the surgery can be and whether ventilation is needed afterwards, and that transthoracic echo helps grade pulmonary hypertension. Barash links a vital capacity below 30% to 35% of predicted with prolonged ventilation.

Sources: Barash 9eMiller 10e

A patient's health can change between clinic and theatre: new illness, changed medication or a worsening chronic condition can make an old assessment misleading. The Joint Commission and CMS standard therefore works in two layers. The full documented history and examination must be no more than 30 days old; anything older is repeated in full, not replaced by a brief look on the day. A later update, documented within 24 hours of admission or registration and before surgery, records any change; Miller quotes it as a 48-hour window.

Full evaluation valid for only one day. People remember that the patient is seen again just before surgery, so a day seems right. That late contact is only the update, which the Joint Commission and CMS place within 24 hours of admission or registration and before surgery. The full evaluation's limit is 30 days.

Miller: the Joint Commission requires documented history and examination no more than 30 days before surgery, plus reassessment in the 48 hours beforehand. Barash gives the same 30-day limit. CMS wants the update within 24 hours of admission, before surgery.

Sources: Miller 10eBarash 9eOutside our books: 42 CFR 482.22(c)(5) and 482.51(b)(1): hospital Conditions of Participation, history and physical (2026)

Time is the main lever. New postoperative stroke risk is highest when surgery falls within 3 months of the index stroke, falls over 3 to 6 and 6 to 11 months, then plateaus near 9 months without returning to baseline. Miller and Barash continue antiplatelets because stopping them adds thrombosis or embolism to surgical hypercoagulability. SNACC finds aspirin continuation does not prevent stroke after noncardiac surgery. Keep antiplatelets running, but credit the delay; for a booked operation, only a 12-month wait among the usual options clears the 9-month plateau.

Six-week delay substantially reduces perioperative stroke. Delaying surgery is the right instinct, and an older consensus proposed only a brief wait. But six weeks still falls inside the highest-risk period, so it gains little. If this is the alternative, continuing existing antiplatelet therapy is the sounder choice.

ACC/AHA data show odds ratios of 14.23 under 3 months, 4.85 at 3 to 6, 3.04 at 6 to 11 and 2.47 beyond 12 months. Miller advises continuing antiplatelets and statins to avoid withdrawal thrombosis; SNACC calls aspirin continuation not indicated for stroke prevention.

Sources: AHA/ACC 2024 · §6.7Miller 10eOutside our books: Perioperative Care of Patients at High Risk for Stroke during or after Non-Cardiac, Non-Neurologic Surgery: Consensus Statement from SNACC (Mashour GA et al.) - 2014 predecessor of the 2020 guideline (2014)

A test is only useful when there is a real chance the patient has the disease it looks for. In a fit patient, screening finds almost no true disease, adds nothing to prognosis and does not improve outcome. Most abnormal results are false positives. These lead to repeat testing, delays, cost and sometimes harmful treatment. So testing is targeted to the patient: for example, a blood count for malignancy, a bleeding disorder or expected major blood loss.

Full blood count benefits healthy elective patients. A full blood count feels like a cheap, harmless baseline, especially before day-case surgery. In a fit patient, though, unsuspected anaemia is rare and the result seldom changes the anaesthetic plan. A spurious abnormal value can delay or cancel the case. Order it only for specific risk factors.

Both major texts advise against routine preoperative tests in healthy patients: poor diagnostic yield, little prognostic information, no improved outcomes, and false positives that do not change management. Choose tests for specific comorbidities or the type of surgery.

Sources: Barash 9eMiller 10e

Anaesthesia and surgery can harm a fetus, so a positive result changes management. The sources have moved over time. An older, simpler line recommends testing all women of reproductive age. The ASA advisory refines this: offer the test when the result would change management, but do not make it mandatory. Ask every patient, record the discussion and test with consent whenever doubt remains. Routine blood counts, clotting tests or chest films in a fit young patient add little and can lead to costly investigation of false positives.

Blood count more important than pregnancy testing. A blood count feels like the default baseline before a general anaesthetic, and a fracture suggests blood loss. But a closed reduction in a healthy 20-year-old involves little bleeding, so a normal count changes nothing. Missing a pregnancy would change management.

Sources recommend testing women of reproductive age. The ASA view is that the test is offered when it would alter management, with consent, rather than required. Routine testing in healthy patients adds cost and leads to follow-up of false positives.

Sources: Barash 9eMiller 10eOutside our books: Pituitary Surgery (2026)

A higher STOP-Bang score means a greater chance of moderate-to-severe OSA. Miller's box labels any positive screen high risk; Barash and Faust split that range into 3–4 intermediate and 5–8 high, Faust tying 5 or more to worse disease and more complications. So 3 makes the screen positive; 5 makes it high risk. Neck item, two versions: official 40 cm (16 in) or more in either sex; the books' table (Barash, Faust, Miller) 17 in (43 cm) men, 16 in (41 cm) women. Once diagnosed, severity and PAP settings matter.

Rescreening with STOP-Bang in known OSA. Screening feels like the safe first step, but STOP-Bang only estimates risk in patients not yet diagnosed. Once OSA is established, the step is to confirm how severe it is and which PAP setting is prescribed before operating. Rescreening adds nothing.

STOP-Bang: 3–4 positive answers indicate intermediate risk of moderate to severe OSA, and 5–8 indicate high risk. A score of 5 or more predicts more severe OSA and more postoperative complications. SASM guidance favours knowing OSA severity and the recommended PAP settings before surgery.

Sources: FaustBarash 9eMiller 10e

A fasting patient who still takes an oral agent can become hypoglycaemic, and sulfonylureas carry the greatest risk. So the tablets are omitted that morning and restarted once the patient eats normally. Apixaban is stopped days ahead so its anticoagulant effect has worn off before needle placement. One textbook sets a minimum of 3 days; ASRA sets 72 hours for high-dose therapy. These are the same interval. When the expected major-bleed risk is below 2%, the guideline allows a shorter break with prompt restart.

Cardiac drugs require day-of-surgery preoperative hold. People fear hypotension or bradycardia under anaesthesia and reach for the cardiac drug. The books do not name it as a day-of-surgery hold. The drug they name is the oral glucose-lowering agent, because fasting while still taking it invites hypoglycaemia.

Omit oral hypoglycaemics on the surgical morning and resume with normal diet; sulfonylureas particularly risk hypoglycaemia. Stop apixaban at least 3 days (72 hours at high dose) before neuraxial block. Shorter interruption suits surgery with major-bleed risk below 2%.

Sources: Barash 9eOutside our books: Regional anesthesia in the patient receiving antithrombotic or thrombolytic therapy: ASRA Pain Medicine evidence-based guidelines (2025)Faust

Warfarin's effect only fades as the INR falls over several days. So the drug is stopped early and the INR is checked again within 24 hours of surgery. Stoelting's valve chapter allows 3 to 5 days and Miller says 5, so five days satisfies both. Stopping leaves the patient unprotected. Low-risk patients can accept that gap. High-risk patients, including those with valve disease on a vitamin K antagonist, get unfractionated or low-molecular-weight heparin once the INR drops below range. The heparin continues up to the eve or morning of surgery.

Warfarin cessation alone protects mitral stenosis. A plain five-day stop with an INR check is the correct default and matches the low-risk pathway, so it looks complete. It fails because rheumatic mitral stenosis with AF carries high thrombotic risk. Stopping alone leaves an unprotected gap, so heparin cover is needed once the INR is subtherapeutic.

Miller stops vitamin K antagonists five days ahead for low-risk patients, bridges high-risk patients with heparin and advises an INR check within 24 hours. Stoelting's valve chapter stops them 3 to 5 days ahead and starts heparin as soon as the INR is subtherapeutic.

Sources: Miller 10eStoelting 8e

Each rheumatoid drug raises a perioperative issue, but only steroids call for a specific anaesthetic action. Patients on long-term steroids may need supplementation. Above 10 mg prednisone daily, a stress dose of hydrocortisone is given during surgery on top of the usual morning dose. Methotrexate, by contrast, is simply continued. Other DMARDs raise infection risk but do not change antibiotic prophylaxis. NSAIDs mainly prompt you to consider their effect on platelet function.

Methotrexate needs specific perioperative anaesthetic action. Methotrexate is the anchor DMARD, so it invites worry about immunosuppression, wound healing and a theoretical folate problem with nitrous oxide. Yet current evidence supports continuing it through surgery, and it needs no specific anaesthetic intervention, unlike steroid cover.

Stress-dose hydrocortisone is advised during surgery above 10 mg prednisone daily, and long-term steroid users may need supplementation. Methotrexate is continued perioperatively. DMARDs raise infection risk without any change to antibiotic prophylaxis. NSAIDs warrant consideration of their platelet effects.

Sources: Barash 9eMiller 10eStoelting 8e

Clear liquids leave the stomach quickly. By the 2-hour mark, residual gastric volume is safe in children as well as adults, so a longer wait does not lower aspiration risk. A longer wait does cause harm: children kept fasting well beyond the guideline can become hypoglycaemic, dehydrated and acidotic, and can become haemodynamically unstable at induction. So once the clear-fluid interval is met, keeping the child waiting longer brings only harm.

Juice requires longer fasting than clear fluids. Juice feels like more than water, so a further wait seems a cautious compromise. It may also be confused with the longer fasts for milk or formula. But fruit juice without pulp is a clear liquid, the interval is already met, and extra fasting only adds hunger, dehydration and hypoglycaemia.

American fasting guidelines list pulp-free fruit juice with water as a clear liquid, requiring a minimum 2-hour fast. The update applies this to healthy children aged 2 to 16. A major text notes near-universal agreement on 2 hours, versus 6 for a light meal.

Sources: Barash 9eOutside our books: Practice Guidelines for Preoperative Fasting and the Use of Pharmacologic Agents to Reduce the Risk of Pulmonary Aspiration: Application to Healthy Patients Undergoing Elective Procedures (2017)Miller 10e

Clear fluids pass through the stomach quickly, so one drunk a couple of hours before induction leaves no larger residual volume and no extra aspiration risk. On that evidence the ASA allows clear fluids to 2 hours, and the newer European paediatric consensus, cited in Miller's, has shortened it to one hour; stopping at 2 hours satisfies both. Milk empties more slowly, human milk faster than formula or cow's milk, so breast milk gets a shorter fast. Pneumonitis classically requires roughly 0.4 mL/kg aspirated at pH below 2.5.

One millilitre per kilogramme is safe aspiration. A round 1 mL/kg sounds like a sensible paediatric rule of thumb, but the classic pneumonitis threshold is much lower: about 0.4 mL/kg at a pH below 2.5. So a much smaller aspirate can already injure the lung.

ASA gives clear liquids 2 hours, breast milk 4, formula and non-human milk 6; breast milk empties faster than other milks. Liberal clear fluids do not increase residual gastric volume; European paediatric consensus permits one hour. Pneumonitis needs 0.4 mL/kg at pH below 2.5.

Sources: Miller 10eBarash 9eStoelting 8e

Skin organisms enter the bloodstream and wound at incision, so tissue levels must be adequate then and not yet have worn off. A bolus agent such as cefazolin achieves this if started within the final hour. Vancomycin and fluoroquinolones require slow infusion, so are started earlier, up to 120 minutes. Dosing 120 to 180 minutes beforehand increases infections. Barash's 30 minutes to 1 hour, Stoelting's 1 hour and the guideline's 60 minutes describe the same window, not competing time points.

Antibiotic timing from theatre call is adequate. Sending the dose when the patient is called feels safely early and is a familiar habit. But transfer and waiting times vary, so the drug may go in well over an hour before incision. By then tissue levels are falling and the infection risk is higher.

The guideline says to start the antibiotic within 60 minutes before incision, or 120 minutes for vancomycin or fluoroquinolones. Barash's ideal of 30 minutes to 1 hour sits inside this. Dosing within the hour lowers surgical site infection compared with 120 to 180 minutes beforehand.

Sources: Outside our books: Clinical Practice Guidelines for Antimicrobial Prophylaxis in Surgery (2013)Barash 9eFaustStoelting 8e

Wound protection depends on the tissue drug level at incision and throughout contamination. An older prospective cohort, grouping timing at 2 hours, found far more infections when dosing was 2–24 hours before incision. Later work showed dosing within 60 minutes beat 120 to 180 minutes, so guidelines narrowed the target to 60 minutes. A dose given hours before a delayed start is too old under both standards. Redosing intervals run from the original dose, so waiting until after incision leaves the cut uncovered.

Intraoperative redosing covers antibiotic need at incision. This looks like routine intraoperative redosing, so it feels safe. But it leaves the incision itself covered only by a dose given hours earlier, in the early window with higher infection rates. Redosing keeps cover going during surgery; it cannot restore cover at the moment of cutting.

Guidelines advise intravenous prophylaxis within 60 minutes before skin incision. Dosing within 2 hours before incision gave 0.59% infection, against 3.8% for dosing 2–24 hours early. Dosing within 60 minutes beat 120 to 180 minutes. Redosing intervals count from the preoperative dose.

Sources: FaustOutside our books: Clinical Practice Guidelines for Antimicrobial Prophylaxis in Surgery (2013)

Sedation is a continuum. Each level is graded by the stimulus needed to get a purposeful response. In minimal sedation the patient answers voice normally, and in moderate sedation voice or light touch still works. Airway and breathing are intact at both levels. As the drug effect deepens, voice and light touch stop working and only repeated or painful stimuli get through. The airway may now need support, although the circulation usually holds. Once even pain cannot rouse the patient, they are under general anaesthesia.

Sedated appearance indicates deep sedation level. This is tempting because the patient has clearly been given sedative drugs and looks drowsy, so any sedated state can seem 'deep'. But a normal answer to voice is the lightest level, minimal sedation. A deeply sedated patient no longer reliably responds to voice and needs repeated or painful stimulation.

Minimal sedation: patient responds normally to voice. Moderate: responds to voice or light touch. Deep: not easily roused, but purposefully responds to repeated or painful stimuli; airway may need assistance. Unarousable to pain indicates general anaesthesia.

Sources: Barash 9eMiller 10e

Moderate sedation is defined by a purposeful response to voice or light touch. Sedatives may slightly blunt breathing, but respiratory drive and airway patency remain adequate without airway manoeuvres or assisted breathing. The ASA continuum grades ventilation by depth: minimal sedation leaves it untouched, moderate may blunt it mildly but it stays adequate, and deep may impair it. Barash calls ventilation preserved and Faust allows mild impairment, so both agree it stays adequate. Miller says "usually" because patients can drift deeper, hence drugs are titrated and the patient watched.

Ventilation remains unaffected in moderate sedation. 'Unaffected' sounds the same as 'adequate', and one textbook calls ventilation 'preserved' at this level. But 'unaffected' is the standard wording for minimal sedation. In moderate sedation breathing may be mildly depressed and still be adequate, so 'unaffected' claims too much.

The ASA definition, repeated in the major texts, says that during moderate sedation no airway intervention is needed and spontaneous ventilation is adequate. One table allows at most mild impairment. Ventilation is unaffected under minimal sedation and may be impaired under deep sedation.

Sources: Outside our books: Statement on Continuum of Depth of Sedation: Definition of General Anesthesia and Levels of Sedation/Analgesia (2024)FaustBarash 9e

Sedatives depress the laryngeal and pharyngeal reflexes that guard the lower airway, and this can outlast wakefulness: after midazolam, swallowing stays impaired for up to 2 hours. Moderate sedation means, by definition, an airway that needs no support and adequate breathing. Obstruction is therefore not its defining hazard, but sedation can drift deeper. When titration overshoots, the patient slides into deeper sedation with an unsecured airway. Closed-claims data found that respiratory compromise from too much sedation was the leading cause of death or brain injury.

Obstruction is chief moderate sedation complication. Obstruction feels like the obvious sedation hazard, and it can happen if sedation deepens. But moderate sedation is defined by an airway that needs no intervention, so obstruction is not its typical risk. The leading killer is oversedation with respiratory compromise, and obstruction is only one part of that.

Sedation depresses protective laryngeal and pharyngeal reflexes, and midazolam impairs swallowing for up to 2 hours despite wakefulness. Closed-claims analysis named excessive sedation with respiratory compromise as the commonest cause of death. The ASA continuum defines moderate sedation as needing no airway intervention.

Sources: Barash 9eOutside our books: Statement on Continuum of Depth of Sedation: Definition of General Anesthesia and Levels of Sedation/Analgesia (2024)

Light sedation works only when the patient can tolerate discomfort and cooperate on modest doses. The awake-surgery literature puts a history of alcohol or drug abuse alongside anxiety, psychiatric illness, chronic pain and low pain tolerance as predictors that sedation will fail. Plan a deeper technique, controlled by an anaesthetist, from the start. In intracranial disease the hazard is raised ICP, not the tumour itself: sedatives depress breathing, carbon dioxide rises and intracranial hypertension gets worse.

Brain tumour itself contraindicates procedural sedation. Brain tumour sounds dangerous because sedatives can raise ICP. But tumour patients are sedated routinely, even for awake craniotomy. Only raised ICP means sedative premedication must be carefully titrated or avoided, and this option names a tumour, not raised ICP.

Risk factors for failed sedation include psychiatric disorders, anxiety, chronic pain, alcohol or drug abuse and poor pain tolerance. Refusal, language barrier, altered mentation or non-cooperation rule out awake craniotomy. With raised ICP, premedication is titrated or omitted because respiratory depression raises PaCO2.

Sources: FaustBarash 9e

Restlessness can be an early sign that the brain is short of oxygen or substrate. Hypoxaemia, low glucose, low blood pressure and acidaemia all disturb mental clarity and can cause agitation. If you give a sedative or analgesic at that point, you hide the warning sign and leave the real problem untreated. Once the physiology is normal, it then matters whether the distress is anxiety or pain. Benzodiazepines do not relieve pain, and opioids are poor anxiolytic sedatives.

Assume anxiety without first checking hypoxia. Fear is tempting because people are often anxious before a procedure, and the sedative in your hand looks like the cure. But anxiety is a diagnosis you make only after excluding physiological causes. Sedating a patient who is hypoxic treats the symptom and masks a dangerous cause.

Barash advises against giving sedatives or analgesics when altered mental status could reflect hypoxaemia, hypoglycaemia, hypotension or acidaemia. It notes that metabolic upsets such as hypoglycaemia cause agitation. Miller lists hypoxia among the precipitating causes to check promptly in a confused patient.

Sources: Barash 9eMiller 10e

Pulmonary embolism, rather than intra-operative events, accounts for most deaths following office-based surgery. In the accreditation-registry review by Keyes and colleagues, embolism lay behind 13 of 23 deaths, while just one death occurred during the operation. That review and the one by Young and colleagues both attribute these fatal emboli chiefly to abdominoplasty. The textbook quotes a separate mortality figure for liposuction, 19.1 per 100,000, with embolism diagnosed in 23.1% of those deaths. It does not rank the procedures by risk, so that ranking derives solely from the external sources.

Liposuction has highest office-surgery mortality rate. Liposuction is tempting because it is the most common cosmetic procedure, and the textbook discusses its deaths and gives a liposuction mortality rate. That rate does not compare procedures. The outside reviews that do compare them put abdominoplasty first, through fatal embolism.

Outside the textbooks, Keyes and colleagues found embolism caused 13 of 23 outpatient deaths, and, like Young and colleagues, named abdominoplasty as the procedure most tied to fatal embolism. The textbook gives liposuction mortality of 19.1 per 100,000, embolism in 23.1%, without comparing procedures.

Sources: Outside our books: Mortality in outpatient surgery (2008)Outside our books: Office-based surgery and patient outcomes (2018)Barash 9e

No machine can judge the whole patient. The anaesthetist keeps checking how the patient responds to voice. This lets them adjust sedation depth and spot early neurological, respiratory or cardiovascular trouble before it becomes serious. They also keep watching, touching and listening to the patient. Standard monitors make this clinical watchfulness work better, so every room should have them. But a monitor does not replace the clinician. A device measures one variable, while the clinician pulls everything together.

Oximetry is most critical safety monitor. Pulse oximetry is tempting because it is the device most linked to safe sedation and it picks up falling oxygen saturation. But it is one device measuring one variable. It only supports a clinician who is present and paying attention, and the textbook ranks that clinician above every machine.

The textbook names a careful, well-trained anaesthetist as the most vital monitor, expected to stay in the room throughout. Monitoring devices enhance their performance, and checking the patient's response to voice guides sedation and aids early detection of problems.

Sources: Barash 9e

Each of the four predictors raises risk on its own, so the score is a simple count. The points convert to risk in steps: none about 10%, one 20%, two 40%, three 60% and four 80%. Smoking protects, so a smoker loses the non-smoker point. Opioids score only when you are told they will be given. The type of operation and the anaesthetic technique are not on the list. Local anaesthesia therefore neither adds nor removes a point: female sex plus prior PONV is still two points, which is 40%.

Opioid automatically increases PONV risk points. Adding an opioid point because laparoscopic cholecystectomy usually needs opioids gives three factors and 60%. Only stated predictors count; smoking removes the non-smoker point, leaving female sex and prior PONV: two factors, 40%. Local anaesthesia changes nothing either way, so carpal tunnel release under local is also 40%.

The guideline and the textbooks list the same four predictors: female sex, prior PONV or motion sickness, not smoking, and postoperative opioids. They predict an incidence of about 10, 20, 40, 60 and 80% for none, one, two, three and all four factors.

Sources: Outside our books: Fourth Consensus Guidelines for the Management of Postoperative Nausea and Vomiting (2020)

Prospective adult studies found a small set of traits that each add PONV risk independently. Risk rises with every extra factor; the simplified Apfel score counts four of them: female sex, non-smoking, prior PONV or motion sickness, and postoperative opioids. Of these traits, being a woman adds the most. Smoking protects, so a smoker scores lower, not higher. Postoperative opioids add risk. That is why opioid-sparing multimodal analgesia, for example NSAIDs, lowers baseline risk. A male patient lacks the main factor, so male sex is not a risk factor.

Prior PONV is strongest risk factor. A past episode of PONV feels like the most direct proof that a patient is susceptible, and it is a real independent risk factor. But prospective adult studies rank female gender above it as the strongest single predictor.

A review source names female gender as the strongest independent predictor, both of PONV and of needing rescue antiemetics. The consensus guideline lists female sex, previous PONV or motion sickness, non-smoking and young age, and advises cutting opioids through multimodal analgesia.

Sources: Mayo reviewOutside our books: Fourth Consensus Guidelines for the Management of Postoperative Nausea and Vomiting (2020)Miller 10e

Volatile agents are a major cause of postoperative vomiting, and switching to a different inhaled agent does not avoid it. Running anaesthesia on a propofol infusion removes that trigger, and propofol itself has a low emetic profile, so vomiting falls substantially. A previous episode of PONV marks the patient as high risk. That calls for prophylaxis combining several antiemetic classes, not a single rescue drug. Opioids cause more nausea the more is given, so regional blocks and non-opioid analgesia reduce nausea by cutting opioid need.

Volatile agent appropriate after PONV history. Volatile anaesthesia is the familiar default and can seem the neutral choice for a transplant patient. But it keeps the strongest avoidable emetic trigger, whichever inhaled agent is used. In someone who has already had PONV, propofol TIVA removes that trigger.

Propofol TIVA greatly lowers PONV compared with pure volatile anaesthesia, and inhaled agents differ little from one another. Every patient's risk should be assessed. Those with risk factors get prophylaxis, high-risk patients get combined regimens, and regional analgesia reduces opioid-related adverse effects.

Sources: Barash 9eMiller 10eFaust

Insufflated CO2 under pressure can reach the pleural space. It may pass through a breach in the diaphragm or falciform ligament, or through natural openings such as the oesophageal hiatus, aortic hiatus and caval opening. Dissecting close to the diaphragm raises this risk. Gas collecting in the chest can shift the cardiac electrical axis and reduce the QRS amplitude. A simple capnothorax may leave the circulation intact. High airway pressure, hypoxia and severe hypotension come with tension physiology. Stable vital signs therefore do not exclude capnothorax.

Pericardial effusion causes sudden voltage drop. Pericardial effusion is the classic cause of low QRS voltage, but nothing explains one appearing mid-surgery in a previously healthy patient. A sudden voltage drop during diaphragmatic dissection under CO2 insufflation instead suggests gas tracking into the chest.

Capnothorax is more likely when surgery is close to the diaphragm. Its early signs include a shift in the ECG axis and reduced amplitude, together with surgical emphysema of the upper torso and marked hypercarbia. Tension capnothorax adds high airway pressure, hypoxia and severe hypotension.

Sources: Barash 9e

End-tidal carbon dioxide comes only from alveoli that have blood flowing past them. An embolus blocks pulmonary blood flow, so the alveoli beyond it are ventilated but not perfused. Their gas has no carbon dioxide in it and dilutes the exhaled sample. End-tidal falls while arterial carbon dioxide holds steady, so the gap widens. With shunt, blood bypasses ventilated lung. Carbon dioxide crosses the membrane so easily that end-tidal still tracks arterial values. Oxygen falls, but the gap hardly moves.

Endobronchial tube widens arterio-alveolar CO2 gradient. In laparoscopy the pneumoperitoneum and head-down tilt push the carina upwards. The tube can then slip into a bronchus and cause hypoxaemia, so this choice feels likely. But one-lung intubation is a shunt. It lowers oxygen but leaves the carbon dioxide gap almost unchanged. It cannot explain a widened gap.

Anything that cuts pulmonary blood flow, such as embolism or cardiac arrest, adds alveolar dead space and widens the gap between end-tidal and arterial carbon dioxide. Shunt hardly changes that gap, because carbon dioxide diffuses rapidly across the alveolar membrane.

Sources: Mayo review

Insufflated CO2 can escape the peritoneum and reach the pleura, either by dissecting through the mediastinum or by passing through congenital diaphragmatic channels. Pleural gas causes one-sided chest signs and hypoxia. Because the gas is CO2, end-tidal CO2 also rises. The two books name this differently: Barash calls it capnothorax and labels it a CO2 pneumothorax, while Miller calls it pneumothorax from gas tracking through the diaphragm. Capnothorax is the specific type of pneumothorax, so it is the more precise label.

Pneumothorax is correct diagnosis term. Rising end-tidal CO2 with one-sided hyperresonance and reduced air entry suggests pneumothorax (Miller's term), but the specific diagnosis is capnothorax (Barash): pleural insufflated CO2. Choose the specific term when offered.

Barash defines capnothorax as CO2 collecting in the pleura after it leaves the peritoneum. It lists capnothorax as a cause of both hypercarbia and hypoxia in laparoscopy, with severe hypercarbia as an early sign. Miller reports the same event as pneumothorax through diaphragmatic channels.

Sources: Barash 9eMiller 10e

Positive-pressure breaths repeatedly change intrathoracic blood volume and venous return. When stroke volume depends on preload, the amplitude of the pulse-oximeter waveform rises and falls with each breath. The index expresses that swing, so a large value predicts that giving volume will raise output. The signal is cleanest under controlled ventilation with full paralysis. Arrhythmia, tidal volume, temperature and ambient light make it less reliable, and so does spontaneous effort in a patient who is not paralysed. There is also no agreed cut-off.

More paralysis explains high pleth variability. The waveform can swing because the patient is breathing against the ventilator, and the index is unreliable without paralysis, so more relaxant looks logical. But the twitch monitor already shows profound block. The swing is not breathing effort; it reflects preload dependence.

Breath-to-breath swings in the plethysmographic waveform predict fluid responsiveness in ventilated patients. Fluid therapy guided by the index improved outcomes in major abdominal surgery. Several physical confounders limit it, thresholds lack consensus, and it performs poorly in ventilated patients who are not paralysed.

Sources: Miller 10e

PVI measures how much the pleth amplitude swings with each ventilated breath. Big swings mean the ventricle is on the steep Frank–Starling limb and fluid lifts stroke volume; small swings mean it is near the plateau, so fluid adds little. Miller warns thresholds are not agreed and the signal is easily confounded. Outside the textbooks, Cannesson's study and Liu's meta-analysis show cut-offs vary between studies, but a very low reading falls beneath all of them, so the disagreement does not change the decision.

Fluid bolus raises pressure with low PVI. Tachycardia and hypotension during an acute abdomen look like hypovolaemia, and a crystalloid bolus is the reflex first step. But a low PVI says the ventricle will not convert extra preload into stroke volume, so the bolus fails to lift pressure and only adds fluid load.

Outside the textbooks, Cannesson found a pre-bolus PVI above 14% separated fluid responders from non-responders (81% sensitive, 100% specific). Miller notes little agreement on thresholds. Stoelting turns to vasopressors once intravascular volume is judged adequate.

Sources: Outside our books: Pleth variability index to monitor the respiratory variations in the pulse oximeter plethysmographic waveform amplitude and predict fluid responsiveness in the operating theatre (2008)Stoelting 8eMiller 10e

The common peroneal nerve, a sciatic branch also called the common fibular or external popliteal nerve, runs superficially around the fibular head, so a stirrup bar or leg-holder frame can crush it against bone, giving foot drop. Hip hyperflexion with knee extension stretches the parent sciatic nerve too. The two counts fit: Miller's large retrospective review ranked motor neuropathies, where peroneal made up 78%; Barash cites a prospective series counting every lower-limb neuropathy, where obturator was most frequent and sciatic plus peroneal fewer.

Obturator is commonest motor nerve injury. Obturator is tempting because a prospective series found it the most frequent lower-limb neuropathy after lithotomy. That ranking counts every neuropathy. For the commonest motor injury, which is the classic foot-drop palsy from stirrup compression at the fibular head, the answer is common peroneal.

Miller: in a large lithotomy review, common peroneal injury was the commonest lower-limb motor neuropathy at 78%, caused by compression between fibular head and leg-support frame. Barash: a prospective series found obturator 50 per 10,000, sciatic and common peroneal 30 per 10,000.

Sources: Miller 10eMayo reviewBarash 9e

The radial nerve spirals around the shaft of the humerus in the musculospiral groove. On the lateral side of the mid-arm it lies directly against bone, so an encircling cuff crushes it against the humerus with every inflation. Frequent automated cycling, or hours of use in one spot, adds these compressions together until a neurapraxia results. Changing the cuff site during very long operations spreads the pressure and reduces the risk.

Ulnar nerve injury. Ulnar is the reflex answer for arm neuropathy, and Mayo lists ulnar paraesthesia among automated-cuff complications. But the cuff sits on the mid-arm, where the radial nerve winds around the humeral shaft, is pinned against bone with each inflation, and is the nerve Barash and most published cuff injuries name.

Barash lists excessive automated cuff cycling among causes of radial compression on the lateral arm, where the nerve can be pressed against bone; its spiralling course makes it prone to compression neurapraxia, and relocating the cuff every few hours in very prolonged cases is advised.

Sources: Barash 9eMayo review

Sitting up raises the brain above the heart, so the blood column between them leaves the circle of Willis at a lower pressure than the arm; the posture lowers cerebral perfusion and may drop systemic pressure. A heart-level transducer or arm cuff overstates brain pressure; a leg cuff overstates it even more. The external auditory meatus marks brain level. Miller's text notes the gradient's relevance has been questioned, but as that is unsettled, the safe course is measuring at, or correcting to, brain level.

Calf cuff accurately measures cerebral perfusion. A calf cuff looks convenient because it keeps the cuff away from the operative shoulder. But the leg sits far below the brain, so its reading exaggerates cerebral pressure even more than an arm cuff and can hide cerebral hypoperfusion.

Positions with the head above the heart reduce cerebral perfusion pressure. An arterial transducer should be zeroed at circle of Willis level, and the external auditory meatus is the gold-standard reference. Lower-limb cuffs widen the gradient further. Some have questioned whether the gradient matters.

Sources: Miller 10eFaust

Padding belongs under the chest, below the axillae, not inside them. An older Barash figure runs the rolls towards the groin. Miller's later warns that going past the iliac crests compresses the femoral vessels and nerve, so the safe span runs from clavicle to crest. A poorly supported head that presses on the globe lowers ocular perfusion pressure, and hypotension lowers it further, risking retinal ischaemia. Even so, most visual loss after prone spine surgery is ischaemic optic neuropathy, not pressure-related retinal occlusion.

Pressure-related retinal occlusion. Retinal artery occlusion feels like the classic prone eye injury because headrest pressure on the globe can cause it. People therefore assume it is the commonest cause. Registry data show ischaemic optic neuropathy is far more frequent.

Barash advises padding under the chest, below the axillae. Miller's places supports from clavicles to iliac crests, warning that going further compresses femoral structures. Globe pressure reduces ocular perfusion; hypotension worsens it. Ischaemic optic neuropathy caused 83 of 93 registry visual-loss cases.

Sources: Barash 9eMiller 10e

Most visual loss after spine surgery is posterior ischaemic optic neuropathy. The part of the optic nerve behind the globe has a watershed blood supply, so it is vulnerable during long, bloody prone cases. It usually appears within 24 to 48 hours, often on waking, so a complaint later the same day fits. Field loss, reduced acuity, no light perception or a pupil change needs urgent specialist examination to tell optic neuropathy apart from retinal artery occlusion or cortical blindness.

Urgent brain imaging before eye examination. Cortical blindness is a recognised cause, so urgent brain imaging feels decisive. But most cases are optic nerve ischaemia, which even orbital MRI often misses. Both textbooks and the ASA advisory put immediate ophthalmological examination first; imaging follows that call, to exclude intracranial causes such as cortical blindness.

Both textbooks say that eye pain, no light perception, field loss, reduced acuity or a lost pupil response needs immediate ophthalmological examination. Most spine cases are posterior ischaemic optic neuropathy, which usually appears within 24 to 48 hours of surgery, often on waking.

Sources: FaustMiller 10e

Textbooks rank posterior ION first after prone spinal fusion. Long, bloody operations in obese men on a Wilson frame starve the optic nerve behind the globe. Because the cause is systemic, both eyes are usually hit, but one eye can be affected alone, so one-sided loss does not settle it. Retinal artery occlusion needs external pressure on the globe, so the case must show that pressure. Cortical blindness dominated one paediatric scoliosis database, but it darkens both eyes, so a single blind eye argues against it.

Retinal occlusion from eye compression. One eye blind after prone positioning invites retinal artery occlusion, which is a genuine prone injury. But it requires external pressure on the globe, and nothing here suggests it. Ischaemic optic neuropathy can also strike a single eye and is far commoner after spine surgery.

Barash ranks posterior ischaemic optic neuropathy as the main cause of post-spine visual loss, ahead of retinal occlusion and cortical blindness. Miller says ION may be one- or two-sided, mostly bilateral, while retinal occlusion usually follows eye compression. A paediatric database found only cortical blindness.

Sources: Barash 9eMiller 10eOutside our books: Visual loss after corrective surgery for pediatric scoliosis: incidence and risk factors from a nationwide database (2016)

Barash gives a normal range of 12–20 mmHg, with a daily swing of 2 to 3 mmHg. Miller gives a mean of 16 ± 5 mmHg and regards readings above 25 mmHg as abnormal. Together, values in the high teens are normal and only readings above 25 mmHg suggest disease. Posture normally influences pressure: Miller lists supine, prone and head-down positions as raising it, and Barash says a change of position raises it by about 6 mmHg at most.

Normal absolute pressure range. This range looks familiar because it is almost exactly the normal resting pressure. Someone who confuses the absolute value with the change caused by posture will pick it. Lying flat adds only a few mmHg, and at most 6.

Barash gives normal pressure as 12–20 mmHg with a small diurnal swing; position change can raise it by up to 6 mmHg. Miller gives 16 ± 5 mmHg, calls readings above 25 mmHg pathological, and names lying supine as a physiological cause of a rise.

Sources: Barash 9eMiller 10e

Subcutaneous emphysema is insufflated carbon dioxide that has escaped the peritoneum into tissue; absorbed from there, it raises arterial carbon dioxide, which extra ventilation can clear, so hyperventilation comes first and desufflation only if it fails. The same gas can track into the pleural space. Once it pressurises the chest, hyperventilation alone no longer suffices: airway pressure rises, oxygenation falls and blood pressure drops. Removing the source by immediate desufflation is primary; fluids or vasopressors bridge the circulation, and needle or chest drain decompression is reserved for severe cases.

Suction tube obstruction. High peak pressure and desaturation suggest a blocked or kinked tube, so suctioning feels like the reflex airway check. But tube obstruction does not explain upper-torso emphysema, severe hypercarbia and hypotension together; that cluster points to pressurised carbon dioxide in the chest, and suction delays stopping the gas source.

Tension capnothorax presents acutely with high airway pressure, hypoxia and severe hypotension; its primary treatment is immediate desufflation, with haemodynamic support and, in severe cases, needle or chest drain decompression. Simple subcutaneous emphysema is treated with hyperventilation, desufflating if that fails.

Sources: Barash 9e

Raised abdominal pressure squeezes the vena cava, cutting venous return, and raises systemic vascular resistance; in patients with cardiac disease it lowers ejection fraction and cardiac index. A poorly contracting ventricle cannot overcome that extra afterload, so fluid simply congests it and a pure vasoconstrictor adds further afterload. Letting the gas out does the opposite: resistance falls and ejection fraction and cardiac index rise above baseline. Hypercarbia adds vasodilatation and myocardial depression, so ventilation should also be optimised, but it does not remove the mechanical cause.

Raising ventilation to wash out CO2. The raised carbon dioxide and low saturation make ventilation look like the culprit, and extra minute ventilation does correct hypercarbia. But the hypotension has already failed fluid and vasopressor. The books escalate refractory hypotension to abdominal decompression, which is also the step for hypoxia that will not resolve.

Transient hypotension at laparoscopy responds to less volatile, fluid or lower abdominal pressure. Refractory hypotension needs immediate decompression and neutral positioning. In patients with cardiac risk, pneumoperitoneum raises resistance and lowers ejection fraction, and desufflation reverses this. More minute ventilation usually controls carbon dioxide.

Sources: Barash 9e

When the patient is tipped head-down, the abdominal contents and the gas-filled abdomen push the diaphragm up. This moves the mediastinum and carina towards the head. The tube is fixed at the mouth, so its tip is now relatively deeper and can enter a main bronchus. One-lung ventilation then causes shunt and hypoxaemia. The same tilt also reduces lung volume and compliance and raises airway pressures, which encourages atelectasis. A tube placed with its cuff only just below the cords leaves more room before the tip reaches the carina.

Gas embolism. Gas embolism is a known danger of laparoscopy, so it sounds dramatic and plausible. But it usually happens while the abdomen is being insufflated and presents as cardiovascular collapse. Desaturation that starts when the patient is tilted points to the tube moving relative to the carina.

Head-down positioning reduces functional residual capacity and compliance and raises peak pressures, leading to atelectasis and hypoxia. The carina shifts towards the head, so the tube tip can enter a main bronchus and need repositioning. Cuff placement just beyond the cords is advised.

Sources: FaustBarash 9e

Sudden distension of the peritoneum sets off a reflex rise in vagal tone. This slows the heart and can stop it altogether. Because it is a reflex, it starts at the moment of stretch and eases when the stretch is reduced. That is why slower gas entry, lower pressure, an anticholinergic beforehand and opening the port all help. Caval compression is a separate effect. It reduces venous return, which lowers cardiac output and blood pressure but does not produce an abrupt vagal arrest.

Gas embolism. Gas embolism is the dramatic diagnosis people think of when the heart stops during laparoscopy, so it feels like the obvious answer. But the sources explain bradycardia and asystole at the moment of distension by a vagal reflex from the stretched peritoneum, not by gas getting into the circulation.

Barash says peritoneal distension during insufflation can cause vagal bradyarrhythmias, even asystole. Slower insufflation, lower pressure and glycopyrrolate premedication reduce it, and the port is opened if it happens. Miller links reflex bradycardia with pneumoperitoneum to peritoneal stretch and vagal stimulation.

Sources: Barash 9eMiller 10e

PEEP keeps intrathoracic pressure high through expiration. This narrows the gradient that returns systemic venous blood to the right atrium. Right ventricular filling falls, left ventricular filling falls a beat later, and stroke volume and arterial pressure drop. The loss of preload is greater under pneumoperitoneum and in a hypovolaemic patient. Positive pressure also raises right ventricular afterload, but the texts put the pressure fall down to lost preload. Reduced cardiac output is the result, not the underlying mechanism.

Cardiac output does fall, so this feels right, but it describes the result, not the mechanism. The question asks why output and pressure fell. Raised intrathoracic pressure cuts venous return and cardiac filling first, and the drop in output follows from that.

The texts agree that PEEP and positive pressure ventilation raise intrathoracic pressure, which reduces venous return, cardiac filling and therefore cardiac output. With pneumoperitoneum this lowers preload and cardiac index, most of all in hypovolaemia. Right ventricular afterload also rises.

Sources: Stoelting 8eMiller 10eBarash 9e

Insufflated gas raises intra-abdominal and perirenal pressure. This squeezes the renal parenchyma and renal vein and increases renal vascular resistance. Renal perfusion falls, which releases renin, aldosterone and antidiuretic hormone, also called vasopressin. Free water is retained and urine flow drops even in a well-filled patient. Miller and Barash both describe this. Barash's teaching that oliguria in recovery usually reflects hypovolaemia still applies in recovery. During pneumoperitoneum, however, the pressure–hormone effect explains a low output despite adequate hydration. The effect is temporary.

Hypovolaemia. Hypovolaemia is the reflex explanation, and Barash does link oliguria in recovery mainly to low volume. During pneumoperitoneum, though, urine output falls even with adequate filling, because pressure compresses the kidneys and drives antidiuretic hormone release. A low output here does not by itself mean an underfilled circulation.

Miller says intraoperative oliguria can occur despite adequate hydration. Perirenal gas pressure raises renal vascular resistance and triggers aldosterone and antidiuretic hormone release, with diuresis afterwards. Barash links vasopressin release to raised abdominal pressure. It gives hypovolaemia as the usual cause of oliguria in recovery.

Sources: Miller 10eBarash 9e

Raising the inspired concentration of desflurane quickly, especially beyond 5–6%, sets off a burst of sympathetic outflow. Adrenaline, noradrenaline and vasopressin are released, so heart rate and blood pressure rise sharply. The circulatory surge is brief, about 4 to 5 minutes, while the hormonal rise lasts longer. Absorbed CO2 from a pneumoperitoneum can also cause tachycardia and dysrhythmias, but hypercarbia takes 15 to 30 minutes to build. It therefore does not explain a sudden rise in heart rate minutes after the vaporiser was turned up.

Hypercarbia from CO2. Laparoscopy is starting, and CO2 is known to cause tachycardia and arrhythmias, which is worrying with a long QT. But hypercarbia builds slowly, over many minutes, and the timing here follows the desflurane increase rather than any gas absorption.

Sudden increases in desflurane, particularly above 5–6% inspired, cause hypertension and tachycardia through sympathetic activation, lasting about 4 to 5 minutes. Insufflated CO2 causes hypercarbia, with tachycardia and dysrhythmias, only after 15 to 30 minutes.

Sources: Barash 9eMayo reviewMiller 10e

Convection is heat carried off by a moving fluid as it passes over a warmer surface. In a pneumoperitoneum, gas enters cold from the supply and flows across the large, moist peritoneal surface. It picks up heat, then escapes through ports and leaks, and fresh cold gas replaces it. Colder, drier gas is thought to increase this loss, and long operations add to it. That is why inline warmers and humidifiers exist, although warmed, humidified gas has not been shown to beat cold, dry gas.

Evaporation heat loss. Evaporation seems the obvious route, since dry insufflated gas meets moist peritoneum. However, the textbook classifies heat loss during insufflation as convective: cold gas flows over warm viscera and carries heat away, rather than water turning to vapour.

Hypothermia in laparoscopic surgery comes partly from long operating times and from insufflating cold CO2. Convective loss is thought to be worse when abdominal contents meet cold, dry gas. Warmed, humidified insufflation has not been shown to be superior.

Sources: Barash 9eMiller 10e

Anaesthetic agents vasodilate, so core heat first spreads to the cooler periphery. After that, the skin keeps giving heat to a colder theatre. Radiation is infrared emission to surfaces the patient does not touch. Convection is heat carried away by air moving over exposed skin. Both act across the whole body surface. Conduction needs direct contact, and only a small area touches the table. Evaporation plays a minor part. Net loss exceeds metabolic production for several hours, so a longer operation means more heat lost.

Conduction from table. The patient lies on a cold table, so contact seems an obvious route, and radiation is correct, which makes the pair look right. But conduction acts only through the small contact area. Air flowing over large exposed surfaces (convection) removes far more heat. Contact loss is conduction, not convection.

Anaesthetic vasodilation redistributes heat, dropping core temperature 1.0–1.5°C in the first hour. Skin loss, mostly radiant and convective, continues for 2–3 hours. Conduction means contact; convection means moving air. In infants these two routes cause about 75% of loss.

Sources: FaustBarash 9eMayo review

Nearly every anaesthetic dilates vessels, so warm core blood moves into the cooler limbs. Core temperature therefore drops quickly, even though the heat is still inside the body. After that the fall slows, because the body is now really losing heat through the skin, mostly by radiation and convection, faster than metabolism makes it. In the end loss and metabolic production balance out. Core temperature then levels off in a thermal steady state, an equilibrium rather than continued cooling.

Conduction heat loss. Heat does pass to the table by conduction, so this option sounds plausible. But conduction adds little compared with radiation and convection, and skin losses only take over after the first hour. A fall inside the first hour comes from heat moving within the body, not from contact surfaces.

Vasodilating anaesthetics drop core temperature 1.0°C to 1.5°C in the first hour as heat redistributes. The decline is gentler over the next 2 to 3 hours as skin losses exceed metabolism. After 3 to 5 hours it plateaus, with loss matching production.

Sources: Faust

Active warming through the skin is the central measure. Faust ranks forced-air convective warming, with resistive carbon-fibre blankets, as the most effective way to conserve heat. Stoelting says it gives the best mix of efficacy, safety and cost. Miller lists room warming and fluid warmers alongside forced air as good prevention. That fits once they are treated as add-ons, because Stoelting reserves fluid warming mainly for large-volume resuscitation. For laparoscopy, Barash adds that forced-air warming also helps prevent surgical site infection.

Intravenous fluid warming. Fluid warmers appear in every warming list, so they can look like the main measure. The books point them mainly at large-volume resuscitation. A routine laparoscopic sleeve in a healthy adult does not need that, but it does need active skin warming, which is ranked most effective.

Faust names forced-air convective or resistive blanket warming as the most effective heat-conservation techniques. Stoelting calls skin warming key and forced air the best balance of efficacy, safety and price. Barash says it prevents hypothermia in laparoscopy. Miller lists room and fluid warming alongside it.

Sources: FaustStoelting 8eBarash 9eMiller 10e

Cooling sets off a sympathetic stress response. Catecholamines raise vascular resistance, blood pressure and heart rate, so the heart works harder and becomes prone to ischaemia. Cold also slows platelets, clotting enzymes and drug metabolism. That causes more bleeding and transfusion, longer neuromuscular block and slower waking. Infections also become more common. Each of these problems delays recovery, so patients stay in hospital longer. Mild hypothermia gives no protection during routine surgery.

Sympathetic shutdown. Cold seems to slow everything down, so a sympathetic shutdown sounds plausible. In fact cooling triggers stress-hormone and catecholamine release. This raises vascular resistance, pressure and heart rate, and it is part of why myocardial ischaemia occurs.

Randomised trials show that even mild hypothermia increases cardiac complications, wound infection, coagulopathy, transfusion, recovery time and hospital stay. Another text reports up to a threefold rise in infection, along with stress-hormone and catecholamine release. Platelet, clotting-factor and drug-metabolism function are all depressed.

Sources: Stoelting 8eFaustMiller 10e

Shivering mostly answers a fall in core temperature, yet it also follows anaesthesia in normothermic patients, where kappa-opioid, NMDA and serotonin receptor actions are implicated; opioids, clonidine, ondansetron and ketamine suppress it, and pethidine is the usual choice. Miller reserves forced-air warming for the hypothermic patient, while Mayo pairs rewarming with pethidine; these fit together: warm the cold patient, give the drug to the warm one. Axillary probes read below core, so a normal axillary value signals normothermia.

Midazolam as a sedative for shivering. A sedative seems the way to calm a trembling patient. No reference text lists a benzodiazepine against shivering; Barash notes midazolam, unlike propofol, leaves thermoregulatory thresholds unchanged, and one trial found it alone barely altered post-spinal shivering, 50% against 60%. Network meta-analyses rank nefopam, tramadol, pethidine and clonidine top.

Opioids, clonidine, ondansetron and ketamine each prevent or treat shivering, with pethidine 12.5–25 mg IV the commonest adult choice. Forced-air warming is for hypothermia, and axillary readings underestimate core. Morphine rests only on the general opioid statement; lornoxicam appears in no source.

Sources: Miller 10eMayo review

Each prophylactic drug has already occupied the receptor it targets. If nausea breaks through, a bigger dose of the same class has nothing new to block, so evidence shows a same-class redose within 6 hours is no better than placebo. A drug from an untouched class, such as a nonselective antihistamine like promethazine, adds a new mechanism. Once a short-acting drug has worn off after 6 hours, a second 5-HT3 antagonist or butyrophenone dose may be considered, but only if nothing else is available.

Repeat ondansetron. Ondansetron is the familiar first-line antiemetic, so topping it up feels safe and logical. In recovery, the dose was given only hours earlier. A same-class repeat within 6 hours works no better than placebo. Repeating metoclopramide fails for the same reason. Promethazine is the only option from a new class.

Guidance says failed prophylaxis should be treated with a different class. A same-class redose within 6 hours performs like placebo. A second 5-HT3 dose is kept for beyond 6 hours, when no alternative exists. Promethazine rescues effectively but sedates.

Sources: Outside our books: Fourth Consensus Guidelines for the Management of Postoperative Nausea and Vomiting (2020)Barash 9e

If vomiting continues after rescue therapy, it may not be simple PONV. Miller says to rule out serious causes before treating the symptoms. After that, Miller supports two measures that work together. An isotonic fluid bolus restores circulating volume and reduces nausea and dizziness. Intramuscular ephedrine works about as well as droperidol and is less sedating than placebo. Giving the same antiemetic again within the stated interval adds no benefit. If first-line management fails, choose a different second-line drug.

Insufficient volume bolus. It names the fluid type Miller recommends after ambulatory surgery, so it looks right at a glance. But the volume is only half of the studied bolus. The reduction in nausea and dizziness was shown with 20 mL/kg.

Miller advises excluding serious causes before symptomatic relief. A 20 mL/kg isotonic electrolyte bolus reduces nausea and dizziness after day-case surgery. Intramuscular ephedrine 0.5 mg/kg prevents and treats symptoms about as well as droperidol. Repeating a used antiemetic within 6 hours has no value.

Sources: Miller 10e

Parkinson's disease is a loss of dopamine in the extrapyramidal motor pathways. Any antiemetic that blocks central dopamine receptors lowers dopaminergic tone further. That worsens rigidity and slowness and can cause extrapyramidal reactions. This covers haloperidol, droperidol, the phenothiazines including promethazine, and metoclopramide. Ondansetron blocks serotonin receptors instead and spares the dopamine pathways. The UK tutorial names domperidone first because it stays outside the blood–brain barrier. Domperidone is given only by mouth or rectum, so ondansetron is the usual intravenous choice and is the preferred drug in the American review.

Promethazine as antihistamine. Antihistamine antiemetics such as cyclizine are safe in Parkinson's disease, so promethazine looks like a harmless antihistamine. It is actually a phenothiazine that also blocks dopamine receptors. That means it can cause extrapyramidal effects and is listed as contraindicated.

The Mayo review lists droperidol, promethazine and metoclopramide as contraindicated and names ondansetron the preferred antiemetic. Faust rules out all antidopaminergic drugs. The ATOTW tutorial chooses domperidone because it does not enter the brain, and also allows serotonin antagonists.

Sources: Mayo reviewOutside our books: Parkinson's Disease (2026)Faust

Delirium appears when a vulnerable brain meets a perioperative insult. Predisposing traits belong to the patient: older age, pre-existing cognitive impairment or dementia, impaired hearing or sight, frailty, and a heavy illness burden, which shows up as a poor ASA grade or a low serum albumin. Precipitating insults come from the episode: surgery and anaesthesia, pain, lost sleep, enforced immobility, hypoxia, dehydration, metabolic derangement, opioids and anticholinergics. Keeping patients mobile, rested, hydrated and able to see and hear removes precipitants; mobility and a normal albumin sit at the safe end.

ASA three excluded. ASA 3 sounds like a general fitness grade, not something specific to delirium, so people pick it as the exception. But ASA III or worse is one of the main listed risk factors. Raised albumin is the false item, because low albumin is what predisposes to delirium.

The textbooks list patient factors: older age, cognitive impairment or dementia, hearing or visual loss, severe comorbidity, ASA III or higher, hypoalbuminaemia, dehydration and frailty. They also list perioperative triggers: pain, sleep deprivation, immobilisation, hypoxia, opioids and anticholinergic drugs.

Sources: Stoelting 8eFaustMiller 10eBarash 9e

EEG-guided dosing trials give inconsistent results. The large ENGAGES trial found no fall in delirium, partly because anaesthetic dose barely differed between groups. Even so, Miller notes that severe EEG suppression, especially burst suppression, is consistently linked with delirium, and the ESAIC guideline still suggests EEG-guided depth. These views fit if EEG guards against overdosing rather than proving benefit. Barash advises avoiding deep anaesthesia in high-risk patients; Miller advises smaller induction and maintenance doses. No technique is superior: benefit comes from avoiding harm, not the agent chosen.

Total intravenous anaesthesia as prevention. Total intravenous anaesthesia sounds gentler on the brain and avoids volatile agents, so it looks like a delirium-prevention strategy. But no anaesthetic technique has been shown to beat another at preventing postoperative delirium. What matters is avoiding excessive depth, whichever drugs produce it.

Barash recommends avoiding deep anaesthesia in high-risk patients but says the evidence is inconsistent, and that no technique is proven superior. Miller links burst suppression consistently with delirium. He suggests monitoring depth with EEG and giving smaller induction and maintenance doses.

Sources: Miller 10eBarash 9e

In a regular drinker, stopping alcohol can bring on withdrawal within a day. Withdrawal shows as agitation, a surge of sympathetic activity and a risk of seizures. Benzodiazepines mainly prevent withdrawal seizures and also calm the sympathetic overdrive. Choose the drug by cause. The European delirium guideline advises against benzodiazepines for ordinary postoperative delirium, but it makes an exception for alcohol-withdrawal delirium, where they are first-line. Miller agrees that benzodiazepines suit only the withdrawal form. So once the alcohol history points to withdrawal, the drug choice is settled.

Haloperidol for withdrawal. Haloperidol is the familiar drug for agitated postoperative delirium, so it looks like the natural choice for a combative patient. Here, though, the cause is alcohol withdrawal. Haloperidol does nothing against the seizure risk or the sympathetic surge that benzodiazepines treat, and the guidelines make benzodiazepines first-line for this form.

Withdrawal can start 24 hours after the last drink. Benzodiazepines prevent withdrawal seizures and reduce sympathetic overactivity. They are first-line, in small titrated doses, for delirium caused by alcohol withdrawal, but they are not indicated for other forms of delirium.

Sources: Outside our books: Update of the European Society of Anaesthesiology and Intensive Care Medicine evidence-based and consensus-based guideline on postoperative delirium in adult patients (2024)Miller 10e

When the diaphragm is spared, tidal breathing may be adequate. Forced expiration, however, depends on intercostal and abdominal muscles, and these are denervated below a cervical lesion. Without them, expiratory reserve falls, the chest wall moves paradoxically, vital capacity drops sharply and there is no forceful cough. Secretions pool in the small airways and the alveoli beyond them collapse. The resulting shunt causes hypoxaemia within hours. If the secretions are not cleared, they later seed pneumonia, which can then tip the patient into respiratory failure.

Thromboembolism. Paralysed, immobile patients carry a real thromboembolic risk, so embolism feels like the textbook answer. But on the first day the dominant problem is mechanical. The cough is gone, secretions are retained and the dependent lung collapses. That shunt, not a clot, explains the early desaturation.

Barash states that loss of abdominal and intercostal support makes breathing paradoxical, cuts vital capacity by about 60%, and that inability to cough and clear secretions causes atelectasis. Stoelting attributes impaired cough to reduced expiratory reserve, and later pneumonia to retained secretions.

Sources: Barash 9eStoelting 8e

Miller's oliguria chapter names intravascular depletion as the usual postoperative cause, so fluid comes first. Its obesity text and Barash add a condition: muscle breakdown is common after bariatric surgery (22.7% in one laparoscopic series), with risk rising with BMI and operating time, and higher with laparoscopy and long cases. Myoglobin released forms haem casts that block the tubules, so urine can stay low after volume replacement. Here, oliguria unresponsive to an adequate fluid challenge points to rhabdomyolysis, and hydration remains the treatment.

Hypovolaemia. Hypovolaemia is the commonest textbook cause of early postoperative oliguria, and long bariatric cases need generous fluid. But this patient received several litres plus a further challenge, with little blood loss, and still makes little urine. Non-response in a heavy patient after long surgery signals muscle breakdown.

Miller's reports volume depletion as the commonest cause of early postoperative oliguria, and rhabdomyolysis in 22.7% of laparoscopic bariatric patients, linked to BMI and operative duration. Barash agrees it is commoner with laparoscopy and long cases. Haem casts block tubules; aggressive hydration is the mainstay.

Sources: Miller 10eBarash 9e

Emptying the bladder needs two things: a detrusor that contracts and the feeling of a full bladder. A spinal anaesthetic blocks the S2-S4 segments. This weakens the detrusor and removes the urge to void until sensation returns, while fluids keep filling the bladder. Long operations and large fluid volumes add to the load. Neuraxial opioids and anticholinergic or adrenergic drugs make emptying harder still. Older patients are vulnerable even to modest spinal bupivacaine doses above 7 mg.

Female sex. Being female is a well-established risk for PONV (postoperative nausea and vomiting), so it appears in nearly every list of risk factors. Urinary retention runs the other way: the sources name male sex as the risk.

Miller lists male sex, length of surgery, fluid volume, neurological disease, perineal surgery, anticholinergic, beta-blocking or adrenergic drugs, and neuraxial drugs. It adds older patients given more than 7 mg spinal bupivacaine. A review adds age over 50 and joint replacement.

Sources: Miller 10eMayo review

Barash and Stoelting put these three operations in one high-risk group and do not rank them. Any ranking comes from outside the textbooks. Geerts, reporting the ACCP guideline and the only source for this, gives untreated venographic DVT of 41% to 85% after knee arthroplasty against 42% to 57% after hip, so knee edges ahead for DVT. Falck-Ytter's ACCP chapter picks out hip replacement and hip fracture surgery for overall VTE risk. That is a leaning, not a ranking. Confidence in a single winner is low.

Hip arthroplasty. Hip surgery is the classic high-risk orthopaedic operation, and one outside source picks it out for overall VTE risk, so it feels like the winner. For DVT itself, the only source that ranks gives knee arthroplasty the wider and higher untreated range.

Barash gives untreated DVT of 40% to 80% in orthopaedic patients; Stoelting lists knee or hip replacement and hip fracture as high-risk procedures. Outside sources: Geerts reports 41% to 85% after knee versus 42% to 57% after hip; Falck-Ytter highlights hip replacement and hip fracture.

Sources: Outside our books: The ACCP Guidelines for Thromboprophylaxis in Total Hip and Knee Arthroplasty (2012)Barash 9eOutside our books: Prevention of VTE in orthopedic surgery patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: ACCP Evidence-Based Clinical Practice Guidelines (2012)Stoelting 8e

Two views fit together. Miller's arthroplasty section names hip replacement specifically, though it calls the finding disputed. Hip surgery carries a particular clot stimulus: dislocation and femoral reaming kink the femoral vein, causing stasis. Barash's large database study found fewer thromboembolic events with neuraxial techniques after both hip and knee replacement, without ranking them. Miller's neuraxial chapter calls the overall effect uncertain. So hip is correct under either reading, while knee is correct only under the broader one.

Knee replacement. Knee replacement carries a heavy clot burden, and the same database evidence covers it, so it looks equally strong. But no source shows a larger regional benefit at the knee, and only the hip is singled out. Knee is right under one reading only.

Miller links spinal or epidural anaesthesia for hip replacement to fewer venous thromboses, while noting that this is disputed. Barash reports fewer thromboembolic events with neuraxial techniques in both hip and knee replacement. Elsewhere, Miller calls the overall effect uncertain.

Sources: Miller 10eBarash 9e

Several drugs prevent clot after major orthopaedic surgery, which is why reference lists name many. The ACCP guideline ranks LMWH first: it prevents clot as well as or better than rivals, bleeds little and has the longest track record, so newer drugs are tested against it; fondaparinux, warfarin and aspirin rank below it as alternatives. Stoelting accepts a DOAC alongside LMWH and adds the risk branch: aspirin alone is inadequate for joint replacement or hip fracture, acceptable only for low-risk surgery such as below the knee.

Aspirin monotherapy. Aspirin. It is a valid choice: the ACCP guideline lists it at Grade 1B after hip or knee replacement and hip fracture surgery, and NICE uses it after a course of LMWH. But it is not the preferred first choice, and Stoelting still limits aspirin alone to low-risk surgery.

Stoelting allows LMWH or a DOAC for knee or hip replacement, trauma and hip fracture, and restricts aspirin monotherapy to low-risk cases. The ACCP guideline prefers LMWH for its efficacy, low bleeding and long experience, ranking other agents as alternatives.

Sources: Outside our books: Prevention of VTE in orthopedic surgery patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: ACCP Evidence-Based Clinical Practice Guidelines (2012)Outside our books: Thromboembolism prophylaxis in orthopaedics: an update (2018)Stoelting 8e

Forcing cement into the femoral canal raises intramedullary pressure to as much as 680 mm Hg, versus under 100 mm Hg without cement. This pushes marrow and debris into the venous system and on to the lungs, raising pulmonary vascular resistance and straining the right ventricle, which can cause hypotension, hypoxia or arrest. A long stem and a large volume of cement are both recognised hazards. A previously cemented canal is smoother and less permeable, lowering the risk somewhat but not removing these hazards.

Primary hip replacement. It is tempting because a primary hip has a femoral canal that has never been instrumented, and that is a listed risk factor. But a short stem uses less cement. A long cemented stem for a fracture carries more hazards: the long stem, the fracture and the cement load.

The risk factors for cement implantation syndrome are metastatic disease, a never-instrumented canal, a long stem, hip arthroplasty for pathological fracture, pulmonary hypertension with right ventricular failure, and a large amount of cement. High cementing pressure is thought to drive the embolisation.

Sources: Miller 10e

Pushing cement under pressure into the medullary canal forces marrow fat and debris into the pulmonary circulation. The emboli, and the inflammatory cascade they set off, raise pulmonary vascular resistance. Oxygenation then fails, and the strained right ventricle cannot maintain output, causing hypotension, arrhythmia and, at worst, arrest. Toxicity from methyl methacrylate monomer was once the favoured explanation, and Miller still lists it among proposed mechanisms. Barash, however, notes that measured monomer levels are too low to explain the severity, so embolisation is now the leading explanation.

Hypercapnia. Hypercapnia seems like the natural partner of hypoxia in any respiratory crisis. But the books define this syndrome by low oxygen and low blood pressure. The problem is embolic blockage of pulmonary vessels, not failed ventilation, and the books do not list carbon dioxide retention as a feature.

Barash defines the syndrome as hypoxaemia, hypotension and/or altered consciousness during cemented fracture repair or arthroplasty. Miller adds bronchospasm, arrhythmias, raised pulmonary vascular resistance, right ventricular failure and cardiac arrest.

Sources: Barash 9eMiller 10e

Cement and the femoral stem push marrow contents into the femoral veins as they fill the canal. Lavage washes that marrow out first, so less of it can be driven into the circulation. A vent hole gives the displaced material somewhere else to go. The load is larger when the canal has never been instrumented, because its untouched lining is more permeable, and it also rises with a long stem and a large volume of cement. Previously cemented bone becomes smooth and sclerotic and lets less through.

Femoral canal plug. A plug looks like it should block emboli, and it is a real cementing step. However, it neither clears marrow from the canal nor relieves the pressure that drives debris into the veins. The sources name lavage and venting, not plugging, as the protective measures.

Miller says that vigorous pulsatile washing of the canal and drilling distal vent holes before the femoral stem goes in can lessen the haemodynamic effects of marrow embolism. Risk factors include an uninstrumented canal, a long stem, pre-existing pulmonary hypertension and a lot of cement.

Sources: Miller 10e

Deaths after donor nephrectomy are rare: about 3.1 per 10,000 donors within 90 days (0.03%), a rate that has not changed for 15 years. The question is therefore which complication causes those few deaths. Serious bleeding is uncommon, with transfusion needed in only 0.4%. Donors carry a moderate risk of venous thromboembolism, and pulmonary embolism, seen in 0.1%, is the main cause of death. Prevention therefore means compression devices plus prophylactic heparin, continued until the donor is mobile.

Haemorrhage. It is tempting because nephrectomy means dividing large renal vessels, so haemorrhage feels like the obvious killer. Registry data show transfusion is needed in only a small fraction of donors. Thromboembolism, not bleeding, causes most of the deaths.

In more than 80,000 living kidney donors, 90-day mortality was 3.1 per 10,000 (0.03%) and stable over 15 years. Pulmonary embolism occurred in 0.1% and caused most deaths. Pneumatic compression and prophylactic heparin are advised until the donor is walking.

Sources: Miller 10e

These figures come from outside the textbooks. The Hu meta-analysis lists advanced age among strong-evidence predictors, alongside cardiac disease, dementia, high ASA grade and multimorbidity; age is the one that predicts death across every cohort, so cardiac disease is a real but lesser predictor. Mortality rises with the time window measured: Lezak's US data found 3.9% in-hospital, the ANZHFR registry report 4.5% for Australia, and Johansen's UK database analysis 6.6%; one-year deaths are several-fold higher. Within the admission, risk climbs steeply with ASA grade.

Thirty percent mortality. Thirty per cent is the mortality figure most often quoted for hip fracture, so it feels familiar. It is the cumulative one-year figure. Deaths during the admission itself are a small fraction of it, in the low single digits in modern registries.

Evidence supports 12 mortality predictors, including advanced age and cardiac disease. Current inpatient mortality after hip fracture is 3.9%; in-hospital mortality is 4.5% in Australia and 6.6% in the 2015 UK database, ranging from 1.8% (ASA 1-2) to 16.5% (ASA 4).

Sources: Outside our books: Preoperative predictors for mortality following hip fracture surgery: a systematic review and meta-analysis (2012)Outside our books: Inpatient mortality following hip fracture in the United States: an updated analysis of over one million cases (2026)Outside our books: Variation in mortality following hip fracture across the Asia Pacific region: Country/region-specific report (2024)Outside our books: Understanding mortality rates after hip fracture repair using ASA physical status in the National Hip Fracture Database (2017)Miller 10e

Stoelting describes smoke causing mucus hypersecretion and poor ciliary clearance; quitting helps partly by restoring macrophage activity, so smoking weakens lung defence. Carbon monoxide cuts oxygen-carrying capacity and nicotine stimulates the heart, so ischaemia on the intraoperative ECG tracks exhaled carbon monoxide: ST depression is the expected change. Miller lists cardiac, pulmonary, stroke and wound complications in smokers. The one exception (Mayo review): smokers vomit less after surgery, so being a nonsmoker is a PONV predictor.

Smoking strengthens defences. It sounds plausible that an inhaled irritant would rally lung defences. In fact Stoelting credits part of the benefit of stopping smoking to recovery of macrophage activity. Smoking therefore depresses macrophage function and weakens defence against postoperative chest infection.

Miller reports higher death, cardiac, pulmonary, stroke and wound-infection risk in current smokers. Stoelting ties carbon monoxide to reduced oxygen-carrying capacity and credits quitting with restored ciliary and macrophage function. The Mayo review reports less PONV in smokers. No source addresses epithelial permeability.

Sources: Mayo reviewStoelting 8eMiller 10eFaust

Risk sits in patient and procedure domains. Prediction models most often include surgical site, then chronic respiratory disease, then age. Miller's current checklist treats cigarette use as present or absent, so any current smoker qualifies. Stoelting's older text adds a dose effect: above 60 pack-years, overall risk doubles and pneumonia triples. The views agree: smoking always counts, heavier exposure more. No source sets a numeric age or low pack-year cut-off; an option built on such a figure is invented, but a listed operation cannot be argued with.

Specified pack-year count. This option is tempting because smoking is a real risk factor, and this history does count as cigarette use. It loses only because its pack-year figure is not a threshold any source sets. The only quantified cut-off is far higher, whereas abdominal surgery appears on the risk list outright.

Miller's risk-factor box lists advanced age, ASA class 2 or more, heart failure, functional dependence, COPD, weight loss, impaired sensorium and cigarette use. It also lists abdominal and thoracic surgery. Stoelting reports that smokers above 60 pack-years carry double the complication risk.

Sources: Miller 10eStoelting 8e

Protection comes from analgesia continuing after surgery, not from the intraoperative block. A thoracic catheter blunts the surgical stress response and hypercoagulability, controls pain and shifts coronary flow favourably; lumbar catheters do not. Older pooled data in Barash found neuraxial anaesthesia reduced pneumonias but not infarctions. Miller later linked postoperative thoracic epidural analgesia to fewer infarctions. The guideline's registry data favour epidural over intravenous analgesia in open abdominal surgery. Keep the catheter running, ideally with local anaesthetic, which dampens the stress response most.

Early catheter removal. Early catheter removal seems safe because it avoids epidural hypotension, and an older meta-analysis found neuraxial anaesthesia did not reduce infarction. However, that study assessed anaesthetic technique. Cardiac benefit depends on postoperative thoracic analgesia, so removing the catheter at the end of surgery forfeits the protective period.

Miller: postoperative thoracic, not lumbar, epidural analgesia may reduce myocardial infarction, and local anaesthetic solutions dampen the stress response most. Guideline: epidural beats intravenous analgesia for cardiopulmonary complications in open abdominal surgery (OR 0.58). Barash's older neuraxial meta-analysis: no reduction in infarction (RR 1.17).

Sources: Miller 10eAHA/ACC 2024 · §8.2Barash 9e

Every laparoscopic insufflation gas risks entering a vessel. Carbon dioxide dissolves well in blood, and the lungs excrete it quickly, so an intravascular bubble shrinks fast and does less harm. Helium and nitrous oxide have a weaker safety profile than carbon dioxide. In the operative field, carbon dioxide also pushes out or dilutes oxidising gases and itself puts out flame, so diathermy inside the abdomen rarely starts a fire. Insufflating oxygen by mistake removes that protection, and fires have resulted.

Helium gas. Helium is tempting because it is inert and non-flammable, and it avoids the carbon dioxide absorption that loads ventilation. But it lacks carbon dioxide's high solubility in blood. If an embolus occurs, the bubble is not quickly dissolved and breathed out, so the safety profile is worse.

Barash states carbon dioxide is safer than helium or nitrous oxide because its high blood solubility and rapid pulmonary elimination limit harm from an accidental embolus. It also displaces oxidising gases and extinguishes flame. Accidental oxygen use has caused fires.

Sources: Barash 9e

Pressurised CO2 escapes the peritoneum by any available path, and three routes lead to the chest. First, a breach in the diaphragm or falciform ligament, for example from a misdirected insufflation needle. Second, the diaphragm's natural openings: the aortic, oesophageal and caval hiatuses. Third, extraperitoneal gas rises into the mediastinum and spreads along the parietal pleura. Surgery at the hiatus, such as fundoplication, works directly on these openings, so the risk is greater than in operations far from the diaphragm.

Anterior wall dissection. Anterior wall dissection seems riskier because it feels more invasive. But gas that escapes outside the peritoneum, including into the retroperitoneum, is the gas that rises into the mediastinum and pleura. No source gives the anterior abdomen any special route to the chest.

The pleura can be reached through trauma to the diaphragm or falciform ligament, or through natural diaphragmatic openings. Risk rises during operations near the diaphragm, such as Nissen fundoplication. CO2 that leaves the peritoneum passes into the mediastinum and then spreads along the parietal pleura.

Sources: Barash 9e

Insufflated CO2 is absorbed into the blood and has to be breathed off. In COPD there are fewer intact alveolar units and less effective alveolar ventilation, so arterial CO2 can climb quickly and severely, and capnography tracks it poorly. The resulting hypercarbia and acidosis weaken contractility, provoke arrhythmias and dilate arteries. Pneumoperitoneum does push the abdomen upward and add atelectasis and V/Q mismatch, most in obese patients. But measured dead space and shunt barely change, so mismatch is not what drives the CO2 rise.

Worsened ventilation-perfusion mismatch. V/Q mismatch sounds like the natural lung problem in COPD, and pneumoperitoneum does add some, especially in obese patients. But measured dead space and shunt change little. What overwhelms the COPD patient is absorbed CO2 that poorly ventilated lungs cannot excrete.

Barash: in advanced COPD, fewer functioning lung units let arterial CO2 rise fast and severely during pneumoperitoneum, and capnography correlates poorly with it. The acidosis depresses the heart and causes arrhythmias. Pneumoperitoneum adds V/Q mismatch, mostly in obesity, yet dead space and shunt barely change.

Sources: Barash 9e

Absorbed carbon dioxide dilates vessels, provokes arrhythmias, depresses the myocardium and raises pulmonary vascular resistance. A failing right ventricle tolerates this poorly, and patients with little reserve cannot absorb these insults. The guideline therefore drops a fixed routine pressure for the minimum that still shows the field. When reserve is limited it even offers wall-lifting with low-pressure gas. Head-down tilt is not harmful in itself: begin level and tilt later if surgery needs it. Correct acute hypertension first by adjusting anaesthetic depth; keep vasoactive drugs for severe cases.

Higher pressure. Higher pressures are common and give the surgeon a roomier field, so they feel standard. With poor cardiac or pulmonary reserve they add to the gas and haemodynamic burden. The guideline asks for the lowest pressure that achieves exposure, not a routine high one.

The European endoscopic surgery guideline advises the lowest abdominal pressure that gives adequate exposure rather than a routine setting. For limited cardiac or pulmonary function it suggests wall-lifting with low-pressure gas. Barash treats acute hypertension by adjusting depth, reserving short-acting vasoactive agents for severe cases.

Sources: Barash 9eOutside our books: The European Association for Endoscopic Surgery clinical practice guideline on the pneumoperitoneum for laparoscopic surgery (2002)

Gut injury can happen at abdominal entry or later in the case. Many intestinal injuries are missed, so the risk of sepsis is high when they do occur. A gas-filled stomach is larger and lies nearer the upper-abdominal entry path, so the insufflation needle or a trocar can hit it more easily. Emptying it before surgery starts makes the target smaller. Barash therefore advises doing this routinely, and Miller says it lowers the chance of a trocar entering the stomach.

Preventing vomiting. This is tempting because the textbook does say that emptying the stomach may help prevent vomiting. That statement comes from tonsillectomy, where stomach contents set off the vomiting centre. For laparoscopic entry the stated reason is avoiding gastric puncture, so reducing organ injury is the answer.

Barash recommends routinely emptying the stomach with an orogastric tube before surgery begins to reduce gastric injury. Miller agrees it may lower the chance of a trocar entering the stomach. The anti-vomiting role of gastric decompression is described for tonsillectomy.

Sources: Barash 9eMiller 10e