Critical Care / ICU

01 ICU Physiology

Critical care medicine is grounded in applied cardiopulmonary physiology. The intensivist must understand oxygen delivery, ventilation-perfusion relationships, and acid-base chemistry at a quantitative level to make rational decisions about ventilator settings, fluid resuscitation, and vasoactive therapy.

Oxygen Delivery (DO₂)

The fundamental equation of critical care is the oxygen delivery equation:

Determinants of DO₂ and therapeutic targets: hemoglobin (transfuse if Hb < 7 g/dL in most ICU patients per the TRICC trial; threshold 8–10 in ACS), SaO₂ (target ≥ 92%; dissolved O₂ contributes minimally), cardiac output (optimize preload, contractility, afterload). Mixed venous oxygen saturation (SvO₂) from a PA catheter or central venous oxygen saturation (ScvO₂) from a central line reflects the balance between DO₂ and VO₂; ScvO₂ < 70% suggests inadequate delivery or excessive demand.

V/Q Mismatch, Shunt & Dead Space

Gas exchange depends on matching ventilation (V) to perfusion (Q) across 300 million alveoli. The normal V/Q ratio is approximately 0.8.

Acid-Base Physiology — Stewart Approach

The traditional Henderson-Hasselbalch approach (pH = 6.1 + log[HCO₃⁻]/(0.03 × PaCO₂)) identifies respiratory vs. metabolic derangements, but the Stewart physicochemical approach identifies three independent determinants of pH:

Anion Gap & Delta-Delta

Anion gap (AG) = Na⁺ − (Cl⁻ + HCO₃⁻); normal 12 ± 2 (or 8–10 if albumin-corrected). Always correct for albumin: corrected AG = AG + 2.5 × (4.0 − measured albumin). AGMA causes: MUDPILES — Methanol, Uremia, DKA, Propylene glycol, Isoniazid/Iron, Lactic acidosis, Ethylene glycol, Salicylates.

The delta-delta (Δ/Δ) = ΔAG / ΔHCO₃⁻ = (AG − 12) / (24 − HCO₃⁻). If Δ/Δ < 1, there is a concurrent non-AG metabolic acidosis. If Δ/Δ > 2, there is a concurrent metabolic alkalosis. If 1–2, pure AG metabolic acidosis.

Winter's formula for expected respiratory compensation: expected PaCO₂ = 1.5 × HCO₃⁻ + 8 (± 2). If measured PaCO₂ is higher than expected, a concurrent respiratory acidosis is present; if lower, a concurrent respiratory alkalosis.

02 ICU Monitoring

Arterial Line & Waveform Analysis

An arterial line provides continuous beat-to-beat blood pressure monitoring and facilitates frequent blood gas sampling. Preferred sites: radial (most common), femoral, dorsalis pedis, brachial (avoid — end-artery risk). The Allen test assesses collateral ulnar circulation before radial artery cannulation.

The arterial waveform has a systolic upstroke (correlates with LV contractility), a dicrotic notch (aortic valve closure), and a diastolic runoff. A dampened waveform (loss of dicrotic notch, narrowed pulse pressure) suggests catheter kink, clot, air bubble, or arterial spasm — flush the system and reposition. An overdampened waveform underestimates systolic and overestimates diastolic; an underdampened (resonant) waveform overestimates systolic.

Central Venous Pressure (CVP)

CVP is measured via a central venous catheter with the transducer zeroed at the phlebostatic axis (4th intercostal space, midaxillary line). Normal CVP: 2–8 mmHg. CVP reflects right atrial pressure and is used as a surrogate for RV preload. However, CVP is a poor predictor of fluid responsiveness — systematic reviews show no correlation between CVP and preload or fluid response. Dynamic parameters (pulse pressure variation, stroke volume variation, passive leg raise) are far superior.

Pulmonary Artery (PA) Catheter

The Swan-Ganz catheter is a flow-directed, balloon-tipped catheter inserted via the IJ, subclavian, or femoral vein and advanced through the RA → RV → PA. It measures RA pressure, RV pressure, PA pressure, and pulmonary capillary wedge pressure (PCWP, a surrogate for LA pressure and LV preload).

ScvO₂ vs. SvO₂

SvO₂ (mixed venous) is measured from the PA catheter and represents true mixed venous blood. ScvO₂ (central venous) is measured from the SVC via a central line and is a surrogate; it runs approximately 5% higher than SvO₂ under normal conditions. Both track the balance between DO₂ and VO₂. Target ScvO₂ ≥ 70% in sepsis resuscitation (Rivers early goal-directed therapy, although subsequent trials ProCESS/ARISE/ProMISe did not show benefit of protocolized ScvO₂-guided resuscitation over usual care).

Lactate Trending

Lactate is a marker of tissue hypoperfusion and anaerobic metabolism. Normal < 2 mmol/L. Elevated lactate in sepsis independently predicts mortality. The Surviving Sepsis Campaign recommends lactate-guided resuscitation: remeasure lactate every 2–4 hours and target normalization or a decrease of ≥ 20% within 2 hours. Lactate clearance = (initial − delayed)/initial × 100%. Causes of elevated lactate beyond tissue hypoperfusion: epinephrine/albuterol use, hepatic dysfunction, seizures, mesenteric ischemia, thiamine deficiency, malignancy, and metformin.

Point-of-Care Ultrasound — RUSH Exam

The RUSH (Rapid Ultrasound for Shock and Hypotension) exam is a systematic bedside ultrasound protocol for undifferentiated shock:

Dynamic Assessment of Fluid Responsiveness

Non-Invasive Cardiac Output Monitoring

Modern ICU practice increasingly uses non-invasive or minimally invasive CO monitors instead of PA catheters. Arterial pulse contour analysis (FloTrac/Vigileo, PiCCO, LiDCO): estimates SV from arterial waveform morphology; PiCCO requires transpulmonary thermodilution calibration and also provides extravascular lung water (EVLW) and global end-diastolic volume (GEDV). Echocardiographic CO: CO = VTI × LVOT area × HR (measure LVOT diameter in parasternal long axis, VTI in apical 5-chamber with pulsed-wave Doppler). Bioreactance (NICOM): non-invasive thoracic electrodes measure changes in phase shift of applied current to estimate SV; useful for trending. The PA catheter remains indicated when precise hemodynamic profiling is essential (cardiogenic shock, RV failure, pulmonary hypertension, complex mixed shock states).

03 Terminology & Abbreviations

A working knowledge of ICU abbreviations is essential for reading notes, interpreting orders, and communicating efficiently during rounds. A comprehensive master list is provided in Section 30. Key foundational terms:

04 Sepsis & Septic Shock

Sepsis-3 Definitions (2016)

Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection. Organ dysfunction is identified as an acute change in total SOFA score ≥ 2 points consequent to the infection. Septic shock is a subset of sepsis with circulatory, cellular, and metabolic dysfunction associated with a higher mortality risk — clinically identified by a vasopressor requirement to maintain MAP ≥ 65 mmHg AND serum lactate > 2 mmol/L despite adequate volume resuscitation.

SOFA Score (Sequential Organ Failure Assessment)

qSOFA (Quick SOFA)

A bedside screening tool (not a definition of sepsis) to identify patients outside the ICU who may have sepsis. Score 1 point each for: RR ≥ 22, altered mentation (GCS < 15), SBP ≤ 100 mmHg. A qSOFA ≥ 2 should prompt further evaluation for organ dysfunction (SOFA) and consideration of sepsis.

Surviving Sepsis Campaign 2021 — Hour-1 Bundle

2021 guideline changes vs. 2016: Weak recommendation against using qSOFA as sole screening tool (sensitivity too low); suggestion to use balanced crystalloids (LR, Plasmalyte) over normal saline; recommendation against albumin for initial resuscitation; suggestion of capillary refill time-guided resuscitation as an adjunct to lactate; corticosteroids (IV hydrocortisone 200 mg/day) suggested for patients with ongoing vasopressor requirement after adequate fluid resuscitation.

Source Control

Source control is essential in sepsis and should be achieved as soon as medically feasible. Examples: drainage of abscess (percutaneous or surgical), debridement of necrotizing fasciitis, removal of infected device (central line, hardware), cholecystectomy for gangrenous cholecystitis, bowel resection for perforated viscus. Delay in source control independently increases mortality.

Corticosteroids in Septic Shock

The ADRENAL trial (2018) found that hydrocortisone did not reduce 90-day mortality but shortened the duration of shock and mechanical ventilation. The APROCCHSS trial (2018) found that hydrocortisone + fludrocortisone reduced 90-day mortality in septic shock. SSC 2021 suggests IV hydrocortisone 200 mg/day (50 mg q6h or continuous infusion) for patients who remain on escalating vasopressors despite adequate fluid resuscitation. Do NOT perform cosyntropin stimulation testing to select patients — relative adrenal insufficiency is common in critical illness and the test does not reliably guide therapy. Taper over 3–5 days once vasopressors are weaned. Monitor for hyperglycemia (insulin requirements increase) and hypernatremia.

Fluid Resuscitation in Sepsis

Crystalloid type: Balanced crystalloids (LR, Plasmalyte) are suggested over normal saline based on the SMART and BaSICS trials (reduced composite of death, new RRT, or persistent renal dysfunction). Avoid starch-based colloids (HES) — associated with increased AKI and mortality (6S, CHEST trials). Albumin: May be considered as a resuscitation adjunct in patients who have received large volumes of crystalloid (SSC weak recommendation); the SAFE trial showed albumin was safe but not superior to saline in general ICU patients; in the sepsis subgroup, there was a non-significant trend toward benefit. Volume assessment: After the initial 30 mL/kg bolus, further fluid administration should be guided by dynamic measures of fluid responsiveness (PPV, SVV, PLR, echocardiographic assessment) rather than static measures (CVP, MAP alone). The concept of "fluid stewardship" emphasizes that excessive fluid is harmful — the FACTT trial showed conservative fluid management improved outcomes in ARDS.

05 Vasopressors & Inotropes

Vasoactive medications are the pharmacologic backbone of shock management. They are classified by receptor activity (α₁ = vasoconstriction, β₁ = inotropy/chronotropy, β₂ = vasodilation/bronchodilation, V₁ = vasoconstriction, D₁ = renal/splanchnic vasodilation) and clinical effect. The MAP target is 65 mmHg in most patients (higher in chronic hypertension, lower in young patients with hemorrhagic shock is not recommended).

Hemodynamic Profiles by Shock Type

06 Cardiogenic Shock

Cardiogenic shock is defined hemodynamically as CI < 2.2 L/min/m² with an elevated PCWP > 15 mmHg (evidence of congestion) in the presence of adequate or elevated intravascular volume. It is the most lethal form of shock, with in-hospital mortality 40–60%.

Etiology

Acute MI (most common, especially large anterior STEMI — > 40% LV mass infarcted), acute-on-chronic cardiomyopathy, fulminant myocarditis, acute valvular catastrophe (mitral regurgitation from papillary muscle rupture, acute aortic regurgitation), ventricular septal rupture, free wall rupture with tamponade, stress cardiomyopathy (Takotsubo), and right ventricular infarction.

SCAI Shock Classification

Management

Emergent revascularization: PCI for AMI-related cardiogenic shock is the most important intervention; the SHOCK trial demonstrated mortality benefit of early revascularization. Pharmacologic support: norepinephrine first-line vasopressor (SOAP II trial showed less arrhythmia than dopamine); dobutamine for inotropic support if adequate MAP; avoid milrinone loading dose (hypotension). Avoid excessive fluids — elevated PCWP means the patient is already volume-overloaded; further fluid worsens pulmonary edema.

Mechanical Circulatory Support (MCS)

07 Distributive, Obstructive & Hypovolemic Shock

Anaphylaxis

Anaphylaxis is a severe, potentially fatal allergic reaction causing distributive shock via massive histamine-mediated vasodilation and capillary leak. Treatment: Epinephrine IM 0.3–0.5 mg (1:1000) into anterolateral thigh — repeat every 5–15 min if no response. IV epinephrine drip (1–10 mcg/min) for refractory hypotension. Adjuncts: aggressive IVF (up to 5–10 L for capillary leak), diphenhydramine 50 mg IV (H1-blocker), ranitidine/famotidine (H2-blocker), methylprednisolone 125 mg IV (prevents biphasic reaction), albuterol nebulized (for bronchospasm). Remove trigger, maintain airway (intubate early if angioedema threatens airway).

Neurogenic Shock

Results from disruption of sympathetic outflow (spinal cord injury at T6 or above). Classic triad: hypotension + bradycardia + warm/flushed skin (loss of sympathetic tone with unopposed vagal activity). Treatment: IV fluids first, then vasopressors — norepinephrine or phenylephrine. Atropine for symptomatic bradycardia. Target MAP ≥ 85–90 mmHg for 5–7 days post-SCI to optimize spinal cord perfusion. Distinguish from spinal shock (loss of all reflexes below injury level, which is a neurological phenomenon, not a hemodynamic one).

Obstructive Shock — Tension Pneumothorax

Tension pneumothorax causes mediastinal shift, compression of great vessels, and obstructive shock. Clinical: tracheal deviation away from affected side, absent breath sounds ipsilaterally, distended neck veins, hypotension, tachycardia. Treatment is immediate needle decompression (14–16G needle, 2nd intercostal space midclavicular line or 4th/5th ICS anterior axillary line) followed by chest tube placement. This is a clinical diagnosis — do not delay treatment for imaging.

Obstructive Shock — Cardiac Tamponade

Beck's triad: hypotension, distended neck veins (JVD), muffled heart sounds. Pulsus paradoxus: > 10 mmHg drop in SBP during inspiration (exaggerated normal physiologic variation). ECG may show low voltage and electrical alternans. Bedside echo is diagnostic (pericardial effusion with RA/RV diastolic collapse). Treatment: pericardiocentesis (subxiphoid approach, ideally echo-guided) or surgical pericardial window. Bridge with IV fluids (increase preload) — avoid diuretics and positive pressure ventilation (both decrease preload and worsen tamponade physiology).

Obstructive Shock — Massive PE

Massive PE (> 50% pulmonary vascular occlusion) causes acute RV failure and obstructive shock. Signs: acute hypotension, tachycardia, elevated JVP, RV heave. Echo: RV dilation (RV:LV > 1:1), septal bowing, McConnell sign (RV free wall akinesis with apical sparing). Treatment: systemic thrombolysis with alteplase 100 mg IV over 2 hours (or 50 mg bolus in cardiac arrest). Alternatives: catheter-directed therapy, surgical embolectomy. Anticoagulate with heparin. Consider ECMO as bridge in refractory cases.

Hemorrhagic Shock Classification (ATLS)

08 Ventilator Modes

Understanding ventilator modes requires mastery of three concepts: the trigger (what initiates the breath — patient effort or time), the target/limit (what controls gas delivery during inspiration — volume or pressure), and the cycle (what terminates inspiration — volume delivered, time elapsed, or flow decay).

Common Ventilator Modes

Initial Ventilator Settings

09 ARDS

Berlin Definition of ARDS (2012)

Low-Tidal-Volume Ventilation (ARDSNet Protocol)

ARDSNet PEEP/FiO₂ Tables

Lower PEEP table:

Higher PEEP table:

Prone Positioning

The PROSEVA trial demonstrated a significant mortality benefit (16% vs. 32.8%) for early prone positioning (≥ 16 hours/day) in severe ARDS (P/F < 150). Mechanism: improves V/Q matching by redistributing perfusion to ventral lung, recruits dorsal atelectasis, decreases ventral overdistension, improves secretion drainage. Contraindications: spinal instability, open abdomen, hemodynamic instability (relative), increased ICP (relative). Continue prone sessions until P/F > 150 on PEEP ≤ 10 and FiO₂ ≤ 0.6 in supine for at least 4 consecutive hours.

Neuromuscular Blockade in ARDS

The ACURASYS trial (2010) showed mortality benefit with 48 hours of cisatracurium in early severe ARDS. However, the ROSE trial (2019) using lighter sedation found no benefit of early NMB over a strategy of light sedation with PRN NMB. Current practice: use NMB (cisatracurium 1–3 mcg/kg/min) for severe ventilator dyssynchrony, refractory hypoxemia, or high plateau pressures despite optimization, not routinely. Ensure adequate sedation before and during paralysis (patient has no ability to signal distress). Monitor with train-of-four (target 1–2/4 twitches).

Recruitment Maneuvers

Sustained inflation (30–40 cmH₂O for 30–40 seconds) or stepwise PEEP increases to recruit collapsed alveoli. The ART trial (2017) showed that aggressive recruitment + high PEEP titration actually increased mortality. Use cautiously, if at all; stepwise increases with monitoring for hemodynamic compromise preferred. A brief recruitment maneuver after a disconnection or suction event is more widely accepted.

Adjunctive Therapies in Severe ARDS

10 Ventilator Management & Troubleshooting

Auto-PEEP (Intrinsic PEEP)

Auto-PEEP occurs when expiratory flow does not reach zero before the next breath, causing gas trapping and increased end-expiratory pressure above the set PEEP. Common in obstructive disease (COPD, asthma) and when RR is too high or I:E ratio is too short. Detection: perform an expiratory hold maneuver — total PEEP (measured) minus set PEEP = auto-PEEP. On the flow-time waveform, expiratory flow does not return to zero baseline before the next inspiration.

Management: decrease RR (reduce minute ventilation), decrease I:E ratio (shorten inspiratory time, increase expiratory time), decrease TV (if minute ventilation is fixed), tolerate permissive hypercapnia, bronchodilators (albuterol, ipratropium), and in severe cases add external PEEP (set PEEP at ~80% of auto-PEEP to stent open airways and reduce triggering effort without further increasing total PEEP).

High Peak vs. High Plateau Pressure

Ventilator Alarms

High-pressure alarm: Investigate high peak/plateau as above. Low-pressure alarm: Circuit disconnection, cuff leak, inadequate seal around tracheostomy. Low tidal volume alarm: Cuff leak, circuit leak, patient not triggering (over-sedation, neuromuscular weakness). High RR/minute ventilation alarm: Pain, anxiety, metabolic acidosis (compensatory tachypnea), PE, sepsis. Apnea alarm: Patient not triggering — check sedation depth, neuromuscular blockade, brainstem function.

Patient-Ventilator Asynchrony

Permissive Hypercapnia

In ARDS, maintaining protective TV (6 mL/kg IBW) may result in elevated PaCO₂. Permissive hypercapnia allows PaCO₂ to rise to 60–80 mmHg (or even higher) as long as pH remains ≥ 7.20 (some protocols tolerate ≥ 7.15). Compensatory metabolic alkalosis develops over 24–48 hours. Contraindications to permissive hypercapnia: elevated ICP (CO₂ causes cerebral vasodilation and increases ICP), severe pulmonary hypertension (hypercapnia increases PVR), significant right heart failure, and severe metabolic acidosis already present. Sodium bicarbonate infusion may be considered to buffer pH if severely acidotic, but evidence is limited and it generates CO₂ that must be ventilated off.

Ventilator-Associated Pneumonia (VAP) Diagnosis

Suspect VAP when a ventilated patient (≥ 48 h on ventilator) develops new or progressive infiltrate on CXR plus at least 2 of: fever > 38°C, leukocytosis (> 12,000) or leukopenia (< 4,000), purulent secretions. Obtain lower respiratory tract cultures before starting antibiotics: endotracheal aspirate (quantitative: ≥ 10⁶ CFU/mL) or BAL (≥ 10⁴ CFU/mL). The Clinical Pulmonary Infection Score (CPIS) ≥ 6 suggests VAP but has limited sensitivity/specificity. Start empiric antibiotics targeting MRSA and Pseudomonas (see Section 22); narrow at 48–72 h based on culture results; total duration 7 days (procalcitonin-guided discontinuation may shorten further).

11 Liberation from Mechanical Ventilation

Readiness for SBT (Spontaneous Breathing Trial)

Assess daily once the patient meets the following criteria: underlying cause of respiratory failure improving/resolved, adequate oxygenation (PaO₂/FiO₂ ≥ 150 or SpO₂ ≥ 90% on FiO₂ ≤ 0.4 and PEEP ≤ 8), hemodynamically stable (no or low-dose vasopressors), able to initiate inspiratory effort, no significant acidosis (pH ≥ 7.25).

Rapid Shallow Breathing Index (RSBI)

RSBI = f / V_T (respiratory rate in breaths/min divided by tidal volume in liters, measured during 1 minute of unassisted breathing or minimal support). RSBI < 105 predicts successful extubation (sensitivity ~97%, specificity ~64%). RSBI ≥ 105 predicts weaning failure. The RSBI is measured before or at the start of the SBT.

SBT Techniques

SBT Pass/Fail Criteria

Fail criteria (terminate SBT): RR > 35, SpO₂ < 90%, HR > 140 or change > 20%, SBP > 180 or < 90 mmHg, significant diaphoresis, increased anxiety/agitation, accessory muscle use, paradoxical breathing pattern. Pass criteria: Tolerated 30–120 min SBT without meeting fail criteria; RSBI < 105 throughout; adequate gas exchange maintained.

Daily Awakening + Breathing Trial (ABC Trial)

The Girard et al. (2008) ABC trial demonstrated that pairing daily sedation interruption (spontaneous awakening trial, SAT) with daily SBT reduced ventilator days, ICU LOS, and 1-year mortality. Protocol: (1) Hold sedation each morning (SAT), (2) Once patient awakens and passes safety screen, perform SBT, (3) If SBT passed, evaluate for extubation.

Post-Extubation Stridor & Cuff Leak Test

The cuff leak test assesses risk of post-extubation laryngeal edema/stridor. Method: deflate the ETT cuff and measure the difference between inspiratory and expiratory tidal volumes (the "leak"). A cuff leak volume < 110 mL (or < 24% of delivered TV) suggests significant laryngeal edema and higher risk of post-extubation stridor. For patients at risk (prolonged intubation > 7 days, traumatic intubation, prior failed extubation), administer methylprednisolone 20–40 mg IV q4–6h starting 12–24 hours before planned extubation. Post-extubation stridor treatment: nebulized racemic epinephrine 2.25% (0.5 mL in 3 mL NS), IV dexamethasone, heliox; reintubate if severe respiratory distress.

12 Non-Invasive Ventilation & High-Flow

CPAP vs. BiPAP

High-Flow Nasal Cannula (HFNC)

Delivers heated, humidified oxygen at flows up to 60 L/min through large-bore nasal prongs. Mechanisms: washout of nasopharyngeal dead space (improves CO₂ clearance), provides low-level PEEP (~1 cmH₂O per 10 L/min flow, approximately 3–5 cmH₂O at 50–60 L/min with mouth closed), delivers precise FiO₂ (0.21–1.0), improved mucociliary clearance and comfort. The FLORALI trial demonstrated reduced intubation rates with HFNC vs. standard O₂ or NIV in acute hypoxemic respiratory failure (P/F ≤ 300). Settings: start at 40–60 L/min flow, FiO₂ titrated to SpO₂ target.

Contraindications to NIV

Absolute: respiratory arrest, inability to protect airway, uncooperative/agitated patient, facial surgery/trauma precluding mask fit, undrained pneumothorax. Relative: hemodynamic instability, excessive secretions with poor cough, severely impaired consciousness (GCS < 10), high aspiration risk, inability to clear mask for emesis. Monitor closely for NIV failure: if no improvement in RR, SpO₂, pH, or work of breathing within 1–2 hours, proceed to intubation. Delayed intubation after NIV failure is associated with worse outcomes.

13 Acute Kidney Injury & RRT

KDIGO Staging of AKI

ICU AKI causes: prerenal (hypovolemia, low CO, hepatorenal syndrome — 40–60%), intrinsic (ATN from ischemia or nephrotoxins — 30–40%), and postrenal (obstruction — < 5%). FE_Na < 1% suggests prerenal; FE_Na > 2% suggests intrinsic (ATN). In patients on diuretics, use FE_Urea (< 35% = prerenal). Nephrotoxin avoidance (aminoglycosides, contrast, NSAIDs), hemodynamic optimization, and avoidance of fluid overload are the cornerstones of AKI prevention.

Renal Replacement Therapy (RRT)

CRRT Modes

CVVH (Continuous Venovenous Hemofiltration): convective clearance only (replacement fluid pre- or post-filter). CVVHD (Continuous Venovenous Hemodialysis): diffusive clearance only (dialysate runs countercurrent to blood). CVVHDF (Continuous Venovenous Hemodiafiltration): both convective + diffusive clearance. The RENAL and ATN trials showed that higher-intensity CRRT dosing (> 25 mL/kg/h) provides no benefit over standard dosing (20–25 mL/kg/h).

Anticoagulation for CRRT

Regional citrate anticoagulation is preferred (KDIGO recommendation): citrate chelates calcium in the circuit, preventing clotting; calcium is then infused systemically to restore normal ionized calcium. Advantages: longer filter life, no systemic anticoagulation, lower bleeding risk. Contraindications: severe liver failure (citrate metabolized to bicarbonate by liver — accumulation causes citrate toxicity: low ionized calcium with elevated total calcium, metabolic acidosis). Heparin anticoagulation: systemic unfractionated heparin via circuit; simpler but increases bleeding risk; monitor aPTT.

Timing of RRT Initiation

The STARRT-AKI trial (2020) demonstrated no benefit of accelerated (early) RRT initiation compared with a standard strategy (wait for conventional indications) in critically ill patients with severe AKI. Conventional (urgent) indications for RRT: refractory hyperkalemia (K⁺ > 6.5 with ECG changes), severe metabolic acidosis (pH < 7.1), refractory fluid overload (pulmonary edema unresponsive to diuretics), uremic complications (pericarditis, encephalopathy, bleeding), and certain toxic ingestions (methanol, ethylene glycol, lithium, salicylates).

14 Hepatic Failure & Encephalopathy

Acute Liver Failure (ALF)

Defined as coagulopathy (INR ≥ 1.5) and hepatic encephalopathy in a patient without preexisting liver disease, with illness duration < 26 weeks. Causes: acetaminophen (most common in US/UK), viral hepatitis (A, B, E), drug-induced liver injury (idiosyncratic), autoimmune hepatitis, Wilson disease, Budd-Chiari, ischemic hepatitis ("shock liver"), HELLP syndrome, and acute fatty liver of pregnancy.

King's College Criteria (Acetaminophen)

West Haven Classification of Hepatic Encephalopathy

Treatment of hepatic encephalopathy: Lactulose (first-line): 30–45 mL PO/NG q1–2h until bowel movement, then titrate to 2–3 soft stools/day; lactulose enemas (300 mL in 700 mL water) for patients who cannot take PO. Rifaximin (550 mg PO BID): non-absorbable antibiotic added to lactulose for secondary prophylaxis or refractory encephalopathy. Identify and treat precipitants: GI bleeding, infection/sepsis, constipation, hypokalemia, dehydration, sedating medications, dietary protein excess.

Acute-on-Chronic Liver Failure (ACLF)

ACLF is an acute deterioration of liver function in a patient with chronic liver disease, associated with organ failures and high 28-day mortality. Defined by the EASL-CLIF consortium (CLIF-SOFA score): Grade 1 = single organ failure (renal: Cr ≥ 2.0; or non-renal organ failure + renal dysfunction Cr 1.5–1.9 or HE grade 1–2); Grade 2 = two organ failures; Grade 3 = three or more organ failures. 28-day mortality: Grade 1 ~22%, Grade 2 ~32%, Grade 3 > 75%. Precipitants: infection (most common), alcoholic hepatitis, GI bleeding, drug-induced injury. Management: treat the precipitant, organ support as in any ICU patient, early transplant evaluation. Liver transplantation is the only definitive treatment for ACLF Grade 3 with failing medical therapy.

Cerebral Edema in ALF

Cerebral edema occurs in 75–80% of Grade IV encephalopathy and is the leading cause of death in ALF. Monitor: serial neurological exams, pupil reactivity; consider ICP monitoring (EVD or bolt) if Grade III–IV. Target ICP < 20 mmHg, CPP > 60 mmHg. Treatment: head of bed 30°, minimize stimulation, hypertonic saline (23.4% bolus 30 mL via CVC for acute herniation) or mannitol (0.5–1 g/kg IV — ensure serum osmolality < 320), avoid hyperthermia (target 35–36°C), hyperventilation (only as temporizing measure for acute herniation), correct hyperammonemia.

MELD-Na Score

Used for liver transplant allocation. MELD-Na = MELD + 1.32 × (137 − Na) − [0.033 × MELD × (137 − Na)], where MELD = 3.78 × ln(bilirubin) + 11.2 × ln(INR) + 9.57 × ln(creatinine) + 6.43. Na is bounded at 125–137 mEq/L. Higher MELD-Na = higher mortality = higher transplant priority.

15 GI Bleeding in the ICU

Upper GI Bleeding

Risk stratification: The Glasgow-Blatchford Score (GBS) identifies patients at very low risk (GBS = 0) who may not need endoscopy; components include BUN, hemoglobin, SBP, pulse, melena, syncope, hepatic disease, heart failure. The Rockall Score (pre-endoscopy and full) predicts rebleeding and mortality using age, shock status, comorbidities, diagnosis, and stigmata of recent hemorrhage.

Management: Resuscitate (2 large-bore IVs, type and crossmatch, transfuse pRBC if Hb < 7 g/dL, or < 9 in ACS). PPI drip: pantoprazole 80 mg IV bolus then 8 mg/h continuous infusion (initiate pre-endoscopy; continue 72 h post-endoscopy if high-risk stigmata). IV erythromycin 250 mg (prokinetic) 30–60 min before endoscopy to improve visualization. Endoscopy within 24 hours for all patients with significant UGIB; within 12 hours for high-risk (hemodynamic instability, bloody NG aspirate). Hold anticoagulants; reverse with appropriate agents as indicated by severity.

Variceal Hemorrhage

Variceal bleeding in cirrhotic patients is a medical emergency with 20% 6-week mortality. Triple therapy: (1) Octreotide 50 mcg IV bolus then 50 mcg/h infusion (reduces portal pressure via splanchnic vasoconstriction), (2) IV ceftriaxone 1 g daily (reduces bacterial translocation and rebleeding — give for 7 days), (3) Endoscopic variceal band ligation (preferred over sclerotherapy). If endoscopy fails, consider balloon tamponade (Blakemore tube — temporizing only, max 24 h) or TIPS (transjugular intrahepatic portosystemic shunt).

Lower GI Bleeding

Causes: diverticular bleeding (most common), angiodysplasia, colonic malignancy, ischemic colitis, hemorrhoids. Workup: CTA (identifies active extravasation if bleeding rate > 0.3–0.5 mL/min), colonoscopy after bowel prep (can diagnose and treat), angiography with embolization (for hemodynamically significant bleeding identified on CTA). Transfusion targets as for UGIB; restrictive transfusion strategy (Hb < 7) in hemodynamically stable patients.

16 Coagulopathy & DIC

ISTH DIC Score

Score ≥ 5 = compatible with overt DIC. Score < 5 = suggestive, not affirmative; repeat in 1–2 days. Treat the underlying cause (sepsis, malignancy, trauma, obstetric emergency). Transfuse: FFP (15–20 mL/kg) for INR > 1.5 with active bleeding, cryoprecipitate (10 units) for fibrinogen < 150 mg/dL, platelets (1 apheresis unit) for count < 50 × 10⁹/L with bleeding or < 20 × 10⁹/L prophylactically. PCC (Prothrombin Complex Concentrate): 4-factor PCC for warfarin reversal or when FFP volume is prohibitive.

Heparin-Induced Thrombocytopenia (HIT)

Immune-mediated platelet activation via anti-PF4/heparin antibodies, causing thrombocytopenia AND paradoxical thrombosis (50% develop DVT/PE/arterial thrombosis). The 4T Score predicts pretest probability:

Score 0–3 = low probability; 4–5 = intermediate; 6–8 = high. If intermediate/high probability: stop all heparin immediately (including flushes), send PF4/heparin ELISA and SRA, start non-heparin anticoagulant (argatroban IV for hepatic metabolism, bivalirudin for renal/hepatic). Do NOT give warfarin until platelets recover > 150 (risk of venous limb gangrene).

Massive Transfusion Protocol & TEG/ROTEM

MTP: Activated for expected need of ≥ 10 units pRBC/24h. Target 1:1:1 ratio (pRBC:FFP:platelets). Give TXA 1 g IV within 3 hours of injury (then 1 g over 8 hours). TEG (Thromboelastography) / ROTEM (Rotational Thromboelastometry): viscoelastic assays that measure clot formation, strength, and lysis in real time. Guide targeted component therapy: prolonged R-time/CT = factor deficiency (give FFP); low MA/MCF = low fibrinogen or platelets (give cryo/platelets); increased LY30/ML = hyperfibrinolysis (give TXA). Goal-directed TEG/ROTEM-guided transfusion reduces blood product usage and improves outcomes in trauma and cardiac surgery.

17 Nutrition in the ICU

Enteral Nutrition (EN) — Preferred Route

Early enteral nutrition (within 24–48 hours of ICU admission) is recommended for all hemodynamically stable critically ill patients who cannot maintain voluntary intake. Enteral feeding maintains gut mucosal integrity, reduces bacterial translocation, modulates the immune response, and is associated with lower infection rates and shorter ICU stay compared with parenteral nutrition. Start via gastric feeding (NG/OG tube); advance to post-pyloric (NJ tube) if recurrent aspiration or gastroparesis.

Trophic vs. Full Feeds

The EDEN trial (ARDS Network) showed that trophic feeds (10–20 mL/h, ~400 kcal/day) for the first 6 days were non-inferior to full caloric feeding in early ARDS, with fewer GI complications. Current practice: trophic feeds in early critical illness (first 48–72 h), then advance to goal rate. The TARGET trial showed no benefit of energy-dense (1.5 kcal/mL) vs. standard (1 kcal/mL) formula for achieving higher caloric targets.

Caloric & Protein Targets

Caloric targets: 25–30 kcal/kg/day (use IBW for obese patients; some guidelines suggest 11–14 kcal/kg actual weight for BMI 30–50, or 22–25 kcal/kg IBW for BMI > 50). Protein targets: 1.2–2.0 g/kg/day (higher end for burns, multitrauma, surgical patients; lower for hepatic encephalopathy). Protein delivery is likely more important than meeting full caloric goals in the ICU.

Gastric Residual Volume (GRV) Debate

Routine GRV monitoring (checking residuals q4–6h, holding feeds for GRV > 250–500 mL) was standard practice but is now debated. The NUTRIREA2 sub-study and other data suggest that routine GRV monitoring does not reduce aspiration pneumonia and leads to unnecessary feed interruptions. SCCM/ASPEN 2016 guidelines suggest NOT routinely monitoring GRV. If monitored, threshold for holding feeds is GRV > 500 mL (not 200–250 mL). Prokinetics (metoclopramide 10 mg IV q6h or erythromycin 250 mg IV q12h) for gastroparesis/high residuals.

Parenteral Nutrition (PN)

Reserve PN for patients in whom EN is contraindicated (bowel obstruction, mesenteric ischemia, high-output fistula, short bowel syndrome) or when EN alone is insufficient after 7–10 days. Early PN (< 48 h) is associated with increased infections (EPaNIC trial). When PN is used, start at ~80% of target and advance; monitor glucose closely (hyperglycemia is a major complication); add supplemental PN to EN only after day 7 if EN alone is insufficient.

Refeeding Syndrome

18 Status Epilepticus

Status epilepticus (SE) is defined as a seizure lasting > 5 minutes or two or more seizures without full recovery of consciousness between them. Convulsive SE has a mortality of 10–20%; the longer the seizure, the harder it is to terminate and the greater the risk of neuronal injury.

Management Algorithm

Continuous EEG monitoring is essential for all patients with refractory SE (clinical seizures may stop while electrographic seizures persist) and should be applied within 1 hour of onset. Non-convulsive SE accounts for up to 20% of ICU patients with unexplained altered mental status — a high index of suspicion is required.

19 Traumatic Brain Injury

Glasgow Coma Scale (GCS)

GCS 13–15 = mild TBI; 9–12 = moderate TBI; 3–8 = severe TBI (intubate for airway protection). The motor component is the most prognostically valuable individual component.

ICP Monitoring

Indications: severe TBI (GCS 3–8) with abnormal CT, or GCS 3–8 with normal CT plus 2 of: age > 40, unilateral or bilateral motor posturing, SBP < 90. Devices: External Ventricular Drain (EVD) — gold standard, allows both monitoring and therapeutic CSF drainage; placed into the lateral ventricle. Intraparenchymal bolt — fiber-optic or strain-gauge sensor; easier to place, cannot drain CSF, prone to drift. Target: ICP < 22 mmHg (BTF 2016 guidelines, revised from < 20). CPP targets: 60–70 mmHg (CPP = MAP − ICP; avoid CPP > 70 to reduce risk of ARDS from aggressive fluid/vasopressor therapy).

ICP Management Ladder

Mannitol vs. Hypertonic Saline

Multimodal Neuromonitoring

Beyond ICP monitoring, advanced neuromonitoring in severe TBI includes: brain tissue oxygen (PbtO₂) monitoring via Licox probe (target PbtO₂ > 20 mmHg; BOOST-3 trial evaluating PbtO₂-guided therapy); cerebral microdialysis (lactate/pyruvate ratio > 40 suggests cerebral metabolic crisis); continuous EEG (detect non-convulsive seizures, which occur in 20–30% of severe TBI); transcranial Doppler (evaluate for vasospasm, assess autoregulation via pulsatility index and mean flow velocities). Integrated multimodal monitoring provides a more comprehensive picture of brain physiology than ICP alone.

20 Stroke in the ICU

Acute Ischemic Stroke (AIS)

IV alteplase (tPA): 0.9 mg/kg (max 90 mg), 10% as bolus then remainder over 60 min. Eligible within 4.5 hours of symptom onset (or last known well). Exclusions for 3–4.5 hour window: age > 80, NIHSS > 25, oral anticoagulant use, history of both DM and prior stroke. Tenecteplase (0.25 mg/kg, max 25 mg single IV bolus) is emerging as an alternative with simpler administration and non-inferior efficacy (AcT trial 2024). BP must be < 185/110 before and < 180/105 for 24 hours after tPA.

Mechanical thrombectomy: for large vessel occlusion (ICA, M1, proximal M2) with NIHSS ≥ 6, ASPECTS ≥ 6, up to 24 hours from onset (DAWN and DEFUSE 3 trials using perfusion imaging to select patients with salvageable penumbra). Combined with IV tPA if within tPA window.

Blood Pressure Management

Malignant MCA Infarction

Large MCA territory infarcts may develop massive edema peaking at 48–96 hours with midline shift, herniation, and death (80% mortality with medical management alone). Decompressive hemicraniectomy within 48 hours reduces mortality from ~80% to ~20% in patients age < 60 (DECIMAL, DESTINY, HAMLET trials pooled analysis). However, many survivors have significant disability (mRS 4). Discuss goals of care and acceptable functional outcomes with family before proceeding.

Hemorrhagic Stroke — Reversal Agents

21 Sedation & Analgesia

PADIS Guidelines 2018 (Pain, Agitation/sedation, Delirium, Immobility, Sleep)

Key principles: Analgesia-first approach (treat pain before treating agitation), target light sedation (RASS 0 to −2) unless deeper sedation required (ARDS prone positioning, NMB, ICP management), conduct daily sedation interruption paired with SBT, use non-benzodiazepine sedatives (propofol or dexmedetomidine preferred over midazolam/lorazepam to reduce delirium), and implement early mobility.

Richmond Agitation-Sedation Scale (RASS)

Sedation & Analgesia Agents

ICU Delirium — CAM-ICU

The Confusion Assessment Method for the ICU (CAM-ICU) is the validated screening tool. Positive CAM-ICU requires: (1) acute onset or fluctuating mental status AND (2) inattention (squeeze hand on letter "A" in "SAVEAHAART" — ≥ 2 errors) PLUS EITHER (3) altered level of consciousness (RASS ≠ 0) OR (4) disorganized thinking (ask standardized questions + command: "hold up this many fingers"). ICU delirium affects 60–80% of mechanically ventilated patients and is independently associated with increased mortality, longer ICU stay, and long-term cognitive impairment. Prevention: ABCDEF bundle (Assess/prevent/manage pain; Both SAT + SBT; Choice of sedation — avoid benzodiazepines; Delirium assessment; Early mobility; Family engagement). Pharmacologic treatment of hyperactive delirium: haloperidol 2–5 mg IV q4–6h PRN (monitor QTc); quetiapine 50 mg PO BID as scheduled (MIND-USA trial showed no benefit of antipsychotics for delirium duration).

22 Antibiotics in the ICU

Empiric Antibiotic Regimens by Suspected Source

Extended / Prolonged Infusions

Time-dependent antibiotics (β-lactams) are more effective when free drug concentration remains above the MIC for a greater percentage of the dosing interval (%fT > MIC). Extended infusions (3–4 hour infusion): piperacillin-tazobactam 4.5 g q8h infused over 4 hours; meropenem 1–2 g q8h infused over 3 hours; cefepime 2 g q8h infused over 4 hours. Continuous infusion: some centers use 24-hour continuous infusion of meropenem or pip-tazo. Evidence suggests improved clinical cure rates and reduced mortality, especially in critically ill patients with augmented renal clearance.

MRSA Coverage — Vancomycin Dosing

Traditional trough-based monitoring (target trough 15–20 mcg/mL) has been replaced by AUC/MIC-guided dosing (target AUC/MIC 400–600) per 2020 consensus guidelines. AUC-guided dosing reduces nephrotoxicity while maintaining efficacy. Loading dose: 25–30 mg/kg IV (based on actual body weight). Maintenance: 15–20 mg/kg IV q8–12h, adjusted by AUC calculation (Bayesian software or two-level strategy).

Antifungal Coverage

Consider empiric antifungal in: prolonged ICU stay (> 7 days) on broad-spectrum antibiotics, TPN, recent abdominal surgery, immunocompromised, multifocal Candida colonization. Micafungin 100 mg IV daily (echinocandin — first-line for invasive candidiasis per IDSA guidelines). Fluconazole 400–800 mg IV daily (non-critically ill, susceptible Candida species — not for C. glabrata or C. krusei). De-escalate from echinocandin to fluconazole once species identified and susceptible, patient clinically improved.

De-Escalation & Procalcitonin

De-escalation: Narrow antibiotics at 48–72 hours based on culture results and clinical improvement. Broad empiric coverage is a bridge to directed therapy, not a destination. Procalcitonin (PCT): PCT-guided antibiotic discontinuation is supported by multiple RCTs (ProACT, PRORATA, SAPS) and reduces antibiotic exposure without increasing mortality. Algorithm: PCT < 0.25 ng/mL or ≥ 80% decrease from peak suggests bacterial infection resolved — consider stopping antibiotics. Do not use PCT to decide whether to START antibiotics in a critically ill patient (empiric antibiotics should not be withheld while awaiting PCT results).

23 Electrolyte & Acid-Base Emergencies

Hyperkalemia

Severe Hyponatremia

Symptomatic hyponatremia (seizures, obtundation, respiratory arrest) requires urgent correction with 3% hypertonic saline: 100–150 mL IV bolus over 10–20 min; may repeat up to 3 times if symptoms persist. Target: increase Na by 4–6 mEq/L in first 1–2 hours (enough to stop symptoms). Total correction must not exceed 10 mEq/L in 24 hours (8 mEq/L in patients at high risk for osmotic demyelination syndrome: hyponatremia > 48 h, chronic alcoholism, malnutrition, hypokalemia, liver disease). If over-corrected: administer DDAVP 2 mcg IV q8h + D5W infusion to re-lower sodium ("rescue therapy").

Metabolic Acidosis — Systematic Approach

Non-AG metabolic acidosis causes (HARDUPS): Hyperalimentation/Hyperchloremia, Acetazolamide/Addison disease, RTA (types 1, 2, 4), Diarrhea, Uretero-enterostomy, Pancreatic fistula/drainage, Saline (large-volume NS resuscitation).

Severe Hypocalcemia

Ionized calcium < 0.8 mmol/L (symptomatic: perioral tingling, carpopedal spasm, Chvostek sign, Trousseau sign, QT prolongation, seizures, laryngospasm, cardiac arrest). Treatment: calcium gluconate 1–2 g IV over 10–20 min (via CVC: calcium chloride 1 g), followed by calcium infusion. Common ICU causes: massive transfusion (citrate chelation), pancreatitis, renal failure, post-parathyroidectomy (hungry bone syndrome), sepsis. Always check and correct hypomagnesemia concurrently (magnesium is required for PTH secretion and action).

Severe Hypomagnesemia

Mg < 1.0 mg/dL: refractory hypokalemia (Mg is required for renal K+ retention via Na-K-ATPase), refractory hypocalcemia, arrhythmias (torsades de pointes), seizures. Treatment: magnesium sulfate 2–4 g IV over 15–60 min (for torsades: 2 g IV push); maintenance 1–2 g IV q4–6h. Oral repletion for asymptomatic mild deficiency: magnesium oxide 400–800 mg PO daily. In ICU patients, target Mg ≥ 2.0 mg/dL.

Severe Hypernatremia

Na > 160 mEq/L: lethargy, irritability, seizures, intracranial hemorrhage (brain shrinkage). Correct slowly: lower Na by no more than 10 mEq/L per 24 hours (risk of cerebral edema with rapid correction). Use free water (D5W IV or free water via NG tube). Calculate free water deficit: 0.6 × weight (kg) × [(Na/140) − 1]. Address underlying cause: diabetes insipidus (central — DDAVP; nephrogenic — remove offending drug, thiazide + amiloride), insensible losses, osmotic diuresis.

Severe Metabolic Alkalosis

pH > 7.55 with HCO₃⁻ > 40: lethargy, arrhythmias, seizures, impaired oxygen unloading (left-shifted oxyhemoglobin curve). Classify as chloride-responsive (urine Cl < 25 mEq/L: vomiting, NG suction, diuretics — treat with NS and KCl) vs. chloride-resistant (urine Cl > 40 mEq/L: hyperaldosteronism, Cushing, severe hypokalemia — treat underlying cause). For severe refractory alkalosis: acetazolamide 250–500 mg IV (promotes bicarbonaturia); very rarely, HCl infusion via CVC (0.1–0.2 N, calculate deficit and infuse slowly); dialysis with low-bicarbonate bath.

24 Central Lines & Arterial Access

Central Venous Catheter (CVC) Placement

Site selection:

Seldinger Technique

(1) Identify vein with ultrasound (real-time guidance is standard of care — reduces complications and improves first-pass success); (2) puncture vein with introducer needle under US visualization; (3) advance guidewire through needle (J-tip first, monitor for arrhythmia); (4) remove needle, keeping wire in place; (5) make skin nick with scalpel; (6) advance dilator over wire, then remove; (7) thread catheter over wire to appropriate depth; (8) remove wire; (9) aspirate and flush all ports; (10) secure catheter and apply sterile dressing.

Catheter depth: Right IJ/subclavian: 15–17 cm (tip at cavoatrial junction); Left IJ/subclavian: 17–20 cm; Femoral: 20–25 cm. Confirmation: CXR (tip should be at or just above the cavoatrial junction, approximately at the carina level; rule out PTX). Venous blood gas from the distal port (confirms venous placement if PaO₂ is venous range).

Arterial Line Placement

Preferred site: radial artery (non-dominant hand, perform Allen test). Alternative: femoral artery (larger, easier in shock/vasoconstriction, but higher infection risk). Technique: US-guided or palpation-guided, 20G catheter, Seldinger (wire-through-needle) or over-the-needle. After placement, connect to pressure transducer zeroed at phlebostatic axis, confirm pulsatile waveform, perform a fast-flush (square wave) test to assess system damping.

Complications

CVC: pneumothorax (subclavian > IJ), arterial puncture, air embolism, arrhythmia (wire in RV), CLABSI, DVT/thrombosis, catheter malposition. Arterial line: thrombosis, infection, hemorrhage (accidental disconnection), pseudoaneurysm, distal ischemia (rare with Allen test), nerve injury.

Ultrasound-Guided Vascular Access

Real-time ultrasound guidance for CVC insertion is the standard of care and is strongly recommended by multiple professional societies (ASA, AIUM, SIR). Benefits: higher first-pass success rate, fewer needle passes, reduced arterial puncture rate, reduced pneumothorax, faster insertion time. Technique: use a high-frequency linear transducer (5–12 MHz) with sterile sheath and gel. Identify the vein (compressible, non-pulsatile, thin-walled) vs. artery (non-compressible, pulsatile, thick-walled). Short-axis (out-of-plane): transverse view, needle appears as a bright dot — easier to center on vessel but cannot see needle tip depth. Long-axis (in-plane): longitudinal view, entire needle shaft and tip visible — better depth control but harder to maintain alignment. Many operators use a combined approach: short-axis for initial puncture, confirm wire in vessel with long-axis or contrast flush. Always confirm venous placement (blood color, pressure waveform, ultrasound visualization of wire in vein) before dilation.

Intraosseous (IO) Access

IO access provides rapid vascular access when IV access cannot be established within 90 seconds or after 2 failed attempts in critical situations. Sites: proximal tibia (most common — 2 cm below tibial tuberosity, flat medial surface), humeral head (faster flow rates, preferred in adults), distal tibia (medial malleolus), distal femur (pediatric). Devices: EZ-IO power driver (most widely used), manual IO needles. All IV medications can be administered IO, including vasopressors, blood products, and anesthetics. Flush with 10 mL NS bolus to prime the marrow space (may cause brief pain in conscious patients — lidocaine 40 mg IO can reduce this). Remove within 24 hours; complications: extravasation, compartment syndrome (rare), osteomyelitis (rare), fracture, fat embolism.

25 Chest Tubes & Thoracic Procedures

Indications

Pneumothorax: symptomatic, large (> 2 cm apex-to-cupola), tension, or requiring positive pressure ventilation. Hemothorax: initial output > 1500 mL (consider thoracotomy) or ongoing output > 200 mL/h for 2–4 h. Empyema: infected pleural fluid (pH < 7.2, glucose < 60, LDH > 1000, positive Gram stain/culture). Malignant effusion: symptomatic, recurrent. Post-operative: cardiac/thoracic surgery.

Tube Size & Insertion

Insertion technique (safe triangle method): Patient upright or lateral decubitus; identify the safe triangle (bordered by lateral edge of pectoralis major, lateral edge of latissimus dorsi, 5th ICS). Prep and drape, local anesthetic to skin, periosteum, and pleura. Incision at the intercostal space, blunt dissection with a clamp over the TOP of the rib (avoid the intercostal neurovascular bundle on the inferior rib margin), puncture the parietal pleura, finger sweep to confirm pleural space, insert tube and direct posteriorly (effusion) or anteriorly/apically (pneumothorax). Secure with suture, connect to drainage system, obtain CXR.

Management

Water seal: Chamber with 2 cm of water acts as one-way valve. Air bubbling = air leak (persistent > 5–7 days may indicate bronchopleural fistula). Suction: −20 cmH₂O is standard; used for hemothorax, empyema, or persistent air leak. Removal criteria: lung re-expanded on CXR, no air leak for 24 h (clamp trial optional), output < 200 mL/24 h (for effusions/hemothorax), clinical stability. Remove at end-expiration or during Valsalva to prevent air entry; apply occlusive dressing immediately.

Needle Decompression

For tension pneumothorax: 14–16G angiocatheter, 2nd intercostal space midclavicular line (traditional) or 4th–5th ICS anterior axillary line (preferred in obese/muscular patients — shorter chest wall distance). A rush of air confirms tension. Follow immediately with chest tube placement. Note: commercial decompression needles (3.25–5 inch) are more reliable than standard 1.5-inch angiocatheters, which may be too short in obese patients.

26 ICU Bundles & Prevention

VAP Prevention Bundle

(1) Elevate head of bed 30–45°; (2) Daily sedation interruption + assessment of readiness to extubate; (3) DVT prophylaxis; (4) Stress ulcer prophylaxis; (5) Oral care with chlorhexidine 0.12% q12h (some guidelines have moved away from CHG oral care due to concerns about increased mortality in certain studies — institution-specific); (6) Subglottic secretion drainage (specialized ETT); (7) Minimize sedation depth; (8) Early mobilization.

CLABSI Prevention Bundle

(1) Hand hygiene; (2) Full barrier precautions during insertion (cap, mask, sterile gown, sterile gloves, large sterile drape); (3) Chlorhexidine skin antisepsis; (4) Optimal site selection (subclavian preferred for lowest infection rate); (5) Daily review of line necessity with prompt removal when no longer needed. Additional: chlorhexidine-impregnated dressings, antimicrobial-coated catheters if institutional CLABSI rate remains elevated despite bundle compliance.

CAUTI Prevention Bundle

(1) Avoid unnecessary urinary catheters (nurse-driven insertion criteria); (2) Insert using aseptic technique; (3) Maintain closed drainage system; (4) Daily assessment of continued need with prompt removal; (5) Use alternatives when possible (condom catheter, intermittent catheterization). Catheter-associated UTI is the most common HAI; most are preventable with bundle adherence.

Stress Ulcer Prophylaxis (SUP)

Indicated in ICU patients with risk factors: mechanical ventilation > 48 hours, coagulopathy (INR > 1.5, platelets < 50,000), history of GI bleed within 1 year, TBI, burns > 35% TBSA, high-dose corticosteroids, 2+ minor risk factors (sepsis, ICU stay > 1 week, occult GI bleed). Agent: pantoprazole 40 mg IV daily (PPIs preferred over H2RAs based on lower bleeding rates; however, the SUP-ICU trial showed no benefit of pantoprazole vs. placebo in overall ICU mortality). PPI risks: C. difficile infection, pneumonia (gastric acid suppression and bacterial overgrowth). Discontinue SUP when risk factors resolve (e.g., extubation).

DVT Prophylaxis

All ICU patients should receive DVT prophylaxis unless contraindicated: Low-molecular-weight heparin (enoxaparin 40 mg SQ daily) or unfractionated heparin (5000 units SQ q8–12h). LMWH is preferred (slightly more effective). Mechanical prophylaxis (intermittent pneumatic compression devices) for patients with contraindications to pharmacologic prophylaxis (active bleeding, severe thrombocytopenia, recent intracranial surgery). Combine pharmacologic + mechanical in highest-risk patients (major trauma, spinal cord injury).

Glycemic Control

Target blood glucose 140–180 mg/dL in most critically ill patients. The NICE-SUGAR trial demonstrated that intensive glucose control (81–108 mg/dL) increased mortality compared with conventional control (140–180) due to severe hypoglycemia. Use continuous IV insulin infusion with standardized protocol and frequent glucose monitoring (q1–2h). Avoid hypoglycemia (< 70 mg/dL) aggressively — associated with increased mortality, arrhythmias, and neurological injury.

Early Mobility

ICU-acquired weakness (ICUAW) affects up to 50% of patients ventilated ≥ 7 days. Early mobility (physical and occupational therapy starting within 48–72 hours of ICU admission) reduces delirium, ventilator days, ICU and hospital length of stay, and improves functional outcomes at discharge. Safety criteria for mobilization: hemodynamically stable (MAP > 60, no active vasopressor titration), adequate oxygenation (SpO₂ > 88%, FiO₂ ≤ 0.6), no active procedures planned. Progressive mobility: passive ROM → active ROM in bed → seated edge of bed → standing → ambulation.

Temperature Management

Targeted temperature management (TTM) after cardiac arrest: the TTM2 trial (2021) showed no benefit of hypothermia (33°C) vs. normothermia (37.5°C target with active fever prevention) in out-of-hospital cardiac arrest. Current AHA 2023 guidelines recommend actively preventing fever (target ≤ 37.5°C) for at least 72 hours in comatose patients after ROSC. Avoid hyperthermia (> 37.7°C) aggressively in all neurologically injured ICU patients (TBI, stroke, post-cardiac arrest). Fever in the ICU: Evaluate for infection (cultures, imaging) but also consider non-infectious causes: drug fever, transfusion reaction, DVT/PE, adrenal insufficiency, post-operative/post-procedural, CNS injury, malignant hyperthermia, neuroleptic malignant syndrome. Antipyretics: acetaminophen 1 g IV/PO q6h; external cooling (cooling blankets, intravascular cooling devices) for refractory fever or when targeting specific temperature ranges.

Family Engagement & Open Visitation

The ABCDEF bundle includes Family engagement (F) as a core element. Open visitation policies (flexible hours, family presence during rounds, family participation in care) improve patient and family satisfaction, reduce delirium, and do not increase infection rates. Family members should be invited to participate in rounding, goal-setting, and shared decision-making. Structured communication at each shift change (nurse-to-family update) reduces family anxiety. Provide written materials about the ICU environment, common equipment, and what to expect.

27 Withdrawal of Care & Brain Death

Brain Death Determination

Brain death is the irreversible cessation of all functions of the entire brain, including the brainstem. It is a clinical diagnosis. Legal death is declared at the time of the brain death examination completion. Requirements vary by state/institution but generally follow the AAN 2010 guidelines (updated 2023):

Ancillary tests (used when clinical exam or apnea test cannot be completed): cerebral angiography (gold standard — absent intracranial filling), radionuclide cerebral perfusion scan (absent uptake — "hollow skull sign"), EEG (electrocerebral silence), transcranial Doppler (reverberating/oscillating flow or absent diastolic flow). Most institutions require 1 examination by an attending physician (some require 2 exams separated by a defined observation period, especially in pediatrics).

Comfort Care / Terminal Extubation

When a decision is made to withdraw life-sustaining treatment: (1) Discontinue vasopressors, (2) Reduce FiO₂ and PEEP, (3) Administer morphine 2–10 mg IV bolus then infusion (titrate to comfort, not respiratory rate — the goal is symptom management, not hastening death), (4) Glycopyrrolate 0.2 mg IV for secretions ("death rattle"), (5) Remove the ETT or transition to T-piece, (6) Ensure family at bedside if desired, (7) Pronounce death when cardiac activity ceases. Time from withdrawal to death is variable (minutes to days); prepare family for this uncertainty.

Organ Donation

Contact the organ procurement organization (OPO) for every imminent death per federal law (CMS Conditions of Participation). The OPO determines medical suitability and approaches the family. Donation after brain death (DBD) and donation after circulatory death (DCD) are both established pathways. The treating physician should NOT be the one to discuss donation with the family — trained OPO coordinators are most effective. Maintain hemodynamic and hormonal support (vasopressin, thyroid hormone, methylprednisolone) in brain-dead donors to preserve organ function.

28 Prognostication & Family Communication

APACHE II Score

The Acute Physiology and Chronic Health Evaluation II (APACHE II) score is calculated from 12 physiologic variables (worst values in first 24 hours of ICU admission: temperature, MAP, HR, RR, oxygenation, arterial pH, sodium, potassium, creatinine, hematocrit, WBC, GCS) + age points + chronic health points. Score range: 0–71. Higher score = higher predicted hospital mortality. APACHE II is used for ICU benchmarking and research stratification; it should NOT be used to make individual patient care decisions. APACHE IV is the current version with better calibration.

SOFA Trending for Prognostication

Serial SOFA scores are more prognostically useful than a single admission score. A rising SOFA score over the first 48–96 hours predicts mortality > 50%. A declining SOFA score suggests improving organ function and better prognosis. Delta SOFA (change from admission to 48 hours) is an independent predictor of ICU mortality. Each 1-point increase in SOFA is associated with ~10% increase in mortality.

Family Meetings — Structured Approach

Shared Decision-Making

Shared decision-making is the preferred model for ICU decision-making. The physician provides medical information (prognosis, treatment options, burdens/benefits), and the surrogate provides information about the patient's values, preferences, and goals. Together they reach a decision that is consistent with the patient's wishes. This is distinct from paternalistic decision-making (physician decides) and pure autonomy (family decides without guidance). Most families want a physician recommendation — withholding one when asked is not patient-centered.

Palliative Care Integration

Palliative care should be integrated early in the ICU stay, not reserved for end-of-life. Triggers for palliative care consultation: ICU stay > 7–10 days, multisystem organ failure, metastatic malignancy, recurrent ICU admissions, disagreement about goals of care, patient/family request, predicted mortality > 50%. Early palliative care consultation is associated with shorter ICU length of stay, reduced non-beneficial treatments, and improved family satisfaction without increasing mortality.

29 Scoring Systems Master Table

SOFA Score

See Section 4 for the fully enumerated SOFA score table with all six organ systems and scoring criteria (0–4 points each, total 0–24).

qSOFA

Score ≥ 2 → evaluate for sepsis (SOFA); associated with in-hospital mortality ≥ 10% in infection.

APACHE II

12 acute physiology variables (worst in first 24 h) + age points (0–6) + chronic health points (0–5). Total 0–71. Score ≥ 25 predicts > 50% hospital mortality. See Section 28 for full description.

GCS (Glasgow Coma Scale)

See Section 19 for fully enumerated table. Range 3–15 (Eye 1–4 + Verbal 1–5 + Motor 1–6).

RASS (Richmond Agitation-Sedation Scale)

See Section 21 for fully enumerated table. Range +4 (combative) to −5 (unarousable). Target RASS 0 to −2 for most ICU patients.

CAM-ICU (Confusion Assessment Method for the ICU)

CAM-ICU positive = Feature 1 + Feature 2 + (Feature 3 OR Feature 4) = Delirium present.

KDIGO AKI Staging

See Section 13. Stages 1–3 based on creatinine rise and/or urine output criteria.

Berlin ARDS Classification

See Section 9. Mild (P/F 201–300), Moderate (P/F 101–200), Severe (P/F ≤ 100), all on PEEP ≥ 5.

SCAI Cardiogenic Shock Classification

See Section 6. Stages A (At Risk) through E (Extremis).

West Haven Hepatic Encephalopathy

See Section 14. Grades I–IV from mild cognitive changes to coma.

4T Score (HIT)

See Section 16. Categories: Thrombocytopenia, Timing, Thrombosis, oTher causes. Score 0–8; ≤ 3 = low, 4–5 = intermediate, 6–8 = high probability.

ISTH DIC Score

See Section 16. Components: platelet count, D-dimer, PT prolongation, fibrinogen. Score ≥ 5 = overt DIC.

30 Abbreviations Master List