P2Y12 inhibitor: ticagrelor or prasugrel preferred over clopidogrel for PCI.
Anticoagulation: UFH or enoxaparin; bivalirudin alternative.
High-intensity statin (atorva 80) early.
Beta-blocker within 24 h if no shock/HF/heart block.
Oβ only if SaOβ <90% (DETO2X-AMI). Nitrates for ongoing pain β avoid in RV infarct or recent PDE5i.
Don't Miss
Posterior MI: ST-depression V1βV3 with tall R waves β get posterior leads (V7βV9).
RV infarct: inferior STEMI + hypotension on nitrates β V4R, give fluids, avoid nitrates/morphine.
Aortic dissection mimics ACS β check BP both arms, widened mediastinum.
Clinical Pearls
Time is muscle: every 30 min of delay in STEMI reperfusion increases 1-year mortality ~7.5%.
Don't withhold ASA for "possible dissection" without strong evidence β ACS is far more common.
Type 2 MI (demand ischemia from sepsis, anemia, tachyarrhythmia) is treated by fixing the underlying problem, not by routine cath lab activation.
After discharge: DAPT 12 months (longer if high ischemic risk per DAPT score), high-intensity statin, ACEi/ARB if EF <40% or DM/HTN, MRA if EF <40% + DM or HF, cardiac rehab (Class I).
Secondary prevention LDL target <55 mg/dL (ESC) or β₯50% reduction with statin Β± ezetimibe Β± PCSK9i for high-risk.
MONA-BASHDoor-to-balloon β€ 90
Key Trials
PLATONEJM 2009Ticagrelor vs clopidogrel in 18,624 ACS patients β 16% reduction in CV death/MI/stroke (9.8% vs 11.7%) without β major bleeding. Established ticagrelor as preferred P2Y12 in ACS.
DETO2X-AMINEJM 2017Routine supplemental Oβ vs ambient air in 6,629 normoxic MI patients β no difference in 1-year mortality. Basis for "Oβ only if SaOβ <90%."
COMPLETENEJM 2019Complete revascularization vs culprit-only in STEMI with multivessel disease β 26% reduction in CV death/MI (HR 0.74). Drove staged non-culprit PCI recommendation.
ISAR-REACT 5NEJM 2019Prasugrel vs ticagrelor in invasively-managed ACS β prasugrel reduced death/MI/stroke (6.9% vs 9.3%) without more bleeding. Reopened the prasugrel-vs-ticagrelor question.
ISCHEMIANEJM 2020Invasive vs conservative strategy in stable IHD with moderate-severe ischemia β no reduction in death/MI; PCI improved angina. Reinforces medical therapy first in stable CAD (ACS still gets cath).
2
Acute Decompensated Heart Failure
Based on 2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure
Learning Objective: Classify the patient by perfusion and volume status, then initiate guideline-directed medical therapy (GDMT) for HFrEF.
IV loop diuretic: dose β₯ 2Γ home oral dose IV (DOSE trial). Reassess UOP and weight q6β8 h.
Oxygen / NIPPV for respiratory distress; HOB up.
Vasodilators (NTG) for hypertensive ADHF, severe MR/AR.
Restrict Na <2β3 g/d, fluid <1.5β2 L/d if congested or hyponatremic.
HFrEF GDMT β The 4 Pillars
ARNI (sacubitril-valsartan) > ACEi/ARB β Class I.
Evidence-based Ξ²-blocker: carvedilol, metoprolol succinate, or bisoprolol.
MRA (spironolactone/eplerenone) if K <5.0, eGFR >30.
SGLT2 inhibitor (dapagliflozin/empagliflozin) β benefit across all EFs.
Clinical Pearls
Start all 4 pillars early and simultaneously when feasible β do not wait to "max out" one before adding another (STRONG-HF).
Discharge bundle: euvolemic, on oral diuretics, GDMT initiated, follow-up in 7β14 days; reduces 30-day readmission.
In HFpEF: SGLT2i (Class 2a, EMPEROR-Preserved/DELIVER), diuretics for congestion, treat HTN, AF, obesity, OSA. Finerenone (FINEARTS-HF) Class 2a for HFmrEF/HFpEF.
Iron deficiency (ferritin <100 or 100β300 with TSAT <20%) β IV iron (FAIR-HF, AFFIRM-AHF) improves functional capacity and reduces HF hospitalization.
ICD for primary prevention if EF β€35% on optimal GDMT for β₯3 months and life expectancy >1 year (Class I for ischemic; 2a for non-ischemic).
Cardiogenic shock (cold and wet): inotropes (dobutamine/milrinone) for low CO; consider mechanical support (IABP, Impella, ECMO). Refer to advanced HF center.
Wet/Dry Γ Warm/ColdARNI Β· BB Β· MRA Β· SGLT2i
Key Trials
PARADIGM-HFNEJM 2014Sacubitril-valsartan vs enalapril in HFrEF (n=8,442) β 20% reduction in CV death/HF hospitalization (HR 0.80); 16% all-cause mortality reduction. Established ARNI as preferred RAAS blocker.
DAPA-HFNEJM 2019Dapagliflozin vs placebo in HFrEF (n=4,744, with and without DM) β 26% reduction in worsening HF or CV death. SGLT2i benefit independent of diabetes.
EMPEROR-ReducedNEJM 2020Empagliflozin in HFrEF β 25% reduction CV death/HF hospitalization. Class effect of SGLT2i in HFrEF confirmed.
EMPEROR-Preserved & DELIVERNEJM 2021/2022Empagliflozin and dapagliflozin reduced HF hospitalization in HFpEF and HFmrEF β first proven therapies in preserved EF. Drove 2023 AHA focused update.
STRONG-HFLancet 2022Rapid up-titration of GDMT post-discharge (within 2 weeks) vs usual care β 34% reduction in 180-day HF readmission/death. Early high-intensity titration is safe and effective.
DOSENEJM 2011High-dose vs low-dose IV furosemide in ADHF β high-dose led to greater symptomatic relief and weight loss with transient β Cr but no harm. Basis for β₯2Γ home oral dose IV.
3
Diabetic Ketoacidosis
Based on 2024 ADA/EASD Consensus on Hyperglycemic Crises in Adults & ADA Standards of Care 2025
Learning Objective: Diagnose DKA, fix the three problems (volume, acid-base/ketones, electrolytes β especially K) in the right order, and avoid the classic transition errors.
Diagnostic Triad
Hyperglycemia >200 mg/dL (may be lower in euglycemic DKA, e.g., on SGLT2i, pregnancy, fasting).
Ketosis: Ξ²-hydroxybutyrate β₯3 mmol/L (preferred) or moderate/large urine ketones.
Metabolic acidosis: pH <7.3 or HCOβ <18, anion gap >10β12.
Find the Trigger β "5 I's"
Infection (UTI, pneumonia)
Infarction (MI, stroke, mesenteric)
Insulin non-compliance / pump failure
Iatrogenic (steroids, SGLT2i, antipsychotics)
Initial presentation of T1DM / pregnancy
Step-by-Step Management
Step
Action
Key Number
1. Fluids
0.9% NS 500β1000 mL/h Γ 2β4 h, then switch to 0.45% NS once euvolemic. Add D5 when glucose <250.
Replace ~50% deficit in first 12 h
2. Potassium
Check K before insulin. K <3.3 β hold insulin, give K. K 3.3β5.2 β add 20β30 mEq/L. K >5.2 β no K, recheck q2 h.
Target K 4β5
3. Insulin
Regular insulin 0.1 U/kg/h IV (skip bolus per 2024 consensus). Goal: glucose β 50β75 mg/dL/h.
Continue until AG closed
4. Bicarbonate
Only if pH <7.0 with hemodynamic instability. Otherwise no benefit.
Rarely needed
5. Transition
Overlap SC long-acting insulin with IV drip for β₯2 h before stopping drip. Patient must be eating/AG closed/HCOβ β₯18.
Don't drop the drip too early
Clinical Pearls
Resolution criteria: glucose <200 AND two of: HCOβ β₯15, venous pH >7.3, AG β€12. Ketones lag β track AG, not urine ketones.
Euglycemic DKA on SGLT2i: hold the SGLT2i, treat with insulin + dextrose ("two-bag" approach with D10 + NS to allow continued insulin while maintaining glucose).
HHS (mostly T2DM): glucose >600, osm >320, minimal ketosis, pH >7.3 β same fluid/K/insulin principles but slower, prolonged correction; mortality higher.
Cerebral edema is the major peds concern β avoid overly rapid fluid/glucose correction in children.
Pseudohyponatremia: corrected Na = measured Na + 1.6 Γ (glucoseβ100)/100. Watch for hypophosphatemia and hypomagnesemia during treatment.
Mild DKA can often be treated with subcutaneous rapid-acting insulin (lispro/aspart) on the floor β avoids ICU.
Fluids β K β Insulin β Bicarb? β TransitionClose the gap, then taper
Key Trials & Evidence
Umpierrez et al.Diabetes Care 2004 / Am J Med 2004Subcutaneous lispro/aspart q1β2 h on the floor was equivalent to IV regular insulin in the ICU for mild-to-moderate DKA β reduced cost and ICU utilization.
Goyal et al.J Emerg Med 2010Skipping the IV insulin bolus produced equivalent glucose decline and time to AG closure β basis for current ADA recommendation against routine bolus.
ADA/EASD ConsensusDiabetes Care 2024Updated diagnostic thresholds (Ξ²-OHB β₯3.0, pH <7.3, HCOβ <18) and severity grading. Reaffirmed insulin 0.1 U/kg/h without bolus and structured 2-bag protocols.
Chua et al.Crit Care 2011Bicarbonate therapy in DKA with pH 6.9β7.1 did not improve outcomes; routine use abandoned outside extreme acidosis.
SGLT2i euDKA registriesFDA, BMJ Open Diabetes 2020Euglycemic DKA reported across SGLT2i agents, particularly with surgery, illness, low carb intake β informed perioperative SGLT2i hold (β₯3 days) recommendation.
4
Community-Acquired Pneumonia
Based on 2019 ATS/IDSA Guideline for Community-Acquired Pneumonia in Adults
Learning Objective: Decide site of care, choose empiric antibiotics, and know when (and when not) to cover for MRSA/Pseudomonas.
Diagnose
Cough, fever, sputum, dyspnea, pleuritic pain + new infiltrate on imaging.
Sputum & blood cultures: in severe CAP, prior MRSA/Pseudomonas, or empirically treating for them.
Procalcitonin should not be used to decide whether to start antibiotics in CAP.
No routine "HCAP" category β replaced by risk-factor-based MRSA/Pseudomonas coverage.
Use clinical judgment + PSI/SMART-COP for ICU triage.
Empiric Antibiotics (ATS/IDSA 2019)
Setting
Regimen
Outpatient, healthy
Amoxicillin 1 g TID or doxycycline or macrolide (only if local pneumococcal resistance <25%)
Outpatient w/ comorbidities
Ξ²-lactam (amox/clav, cefpodoxime) + macrolide/doxy or respiratory fluoroquinolone
Inpatient, non-severe
Ξ²-lactam (ceftriaxone, ampicillin-sulbactam) + macrolide or respiratory FQ monotherapy
Inpatient, severe
Ξ²-lactam + macrolide (preferred) or Ξ²-lactam + respiratory FQ
Add MRSA cover
Vancomycin or linezolid β only if prior MRSA, recent hospitalization + IV antibiotics, or severe CAP with risk factors
Add Pseudomonas cover
Pip-tazo, cefepime, meropenem β same trigger criteria
Clinical Pearls
Duration: minimum 5 days; stop when afebrile 48 h and clinically stable. Most uncomplicated CAP = 5 days. Pseudomonas/MRSA: 7 days.
Steroids in severe CAP: hydrocortisone 200 mg/d Γ 4β8 d significantly reduced mortality in CAPE COD β consider for ICU-level CAP without contraindication (active infection like flu/aspergillus = caution).
Influenza/COVID coinfection: add oseltamivir during flu season; do not delay antibiotics in bacterial pneumonia.
Follow-up imaging: only if symptoms persist or in patients at risk for malignancy. Routine repeat CXR not recommended.
Vaccinate on discharge: pneumococcal (PCV20 or PCV15+PPSV23), influenza, COVID, RSV (β₯60), Tdap.
Aspiration pneumonia: empiric anaerobic coverage (e.g., amox-clav, ampicillin-sulbactam) not required for most β recent evidence supports standard CAP regimens unless empyema/lung abscess.
CURB-65No more "HCAP"
Key Trials
CAPE CODNEJM 2023Hydrocortisone 200 mg/d in severe CAP (n=800) β 28-d mortality 6.2% vs 11.9% (HR 0.46). Major shift toward steroid use in severe CAP.
Uranga et al.JAMA IM 20165-day antibiotic course (if afebrile 48 h, clinically stable) was non-inferior to physician-determined duration β underpins minimum 5-day rule.
ProHOSP / ProACTJAMA 2009 / NEJM 2018Procalcitonin-guided antibiotic algorithms safely reduced antibiotic exposure in some settings, but no mortality benefit in US ED population β guidelines don't recommend PCT to start antibiotics.
Aliberti / Webb HCAP studiesClin Infect Dis 2013β2018HCAP definition over-treated patients without improving outcomes; risk-factor-based MRSA/Pseudomonas screening more accurate β eliminated HCAP in 2019 ATS/IDSA.
Postma et al.NEJM 2015Ξ²-lactam monotherapy non-inferior to Ξ²-lactam + macrolide or fluoroquinolone in non-severe inpatient CAP β supports stepwise approach.
5
Acute Kidney Injury
Based on KDIGO 2012 AKI Guideline (and 2023 KDIGO consensus updates)
Learning Objective: Stage AKI by KDIGO criteria, work through pre-renal / intrinsic / post-renal, and know absolute indications for emergent dialysis.
KDIGO Definition
β SCr β₯0.3 mg/dL within 48 h, or
β SCr β₯1.5Γ baseline within 7 days, or
UOP <0.5 mL/kg/h Γ 6 h.
Stage 1: 1.5β1.9Γ Β· Stage 2: 2β2.9Γ Β· Stage 3: β₯3Γ or SCr β₯4 or RRT.
Intrinsic: ATN (FENa >2%, muddy brown casts), AIN (eos, drug β PPIs, Ξ²-lactams, NSAIDs), GN (RBC casts), vascular.
Post-renal: obstruction β get a bladder scan and renal ultrasound.
Workup
UA + microscopy, urine Na/Cr, FENa (FEUrea if on diuretic).
Bladder scan, renal US (rule out obstruction).
Med review: hold nephrotoxins (NSAIDs, ACEi/ARB, contrast, aminoglycosides). Renally dose all meds.
If GN suspected: complement, ANA, ANCA, anti-GBM, SPEP/UPEP, HIV/HCV, biopsy.
Emergent Dialysis β "AEIOU"
Acidosis (refractory, pH <7.1)
Electrolytes (refractory hyperkalemia)
Ingestions (toxic alcohols, salicylates, lithium)
Overload (refractory pulmonary edema)
Uremia (pericarditis, encephalopathy, bleeding)
Clinical Pearls
Hyperkalemia order: stabilize membrane (calcium gluconate 1β2 g) β shift K (insulin 10 U + D50, Ξ²-agonist, bicarb if acidotic) β remove K (loop diuretic, K-binder like patiromer/SZC, dialysis). Calcium does not lower K β get an ECG.
Contrast-associated AKI risk is lower than historically taught β do not withhold indicated contrast in critical illness; isotonic IVF is preferred prophylaxis (PRESERVE trial).
STARRT-AKI: no benefit to "early" RRT initiation in absence of urgent indications. Wait for AEIOU.
Cardiorenal syndrome: diurese aggressively (loop Β± thiazide, "sequential nephron blockade") even with rising Cr β congestion drives the AKI.
Hepatorenal syndrome looks like pre-renal but doesn't respond to volume β treat with vasoconstrictor + albumin (see Talk 9).
AKI markedly increases risk of CKD β arrange nephrology follow-up post-discharge; resume renin-angiotensin therapy when stable.
Pre Β· Intrinsic Β· PostAEIOU for HD
Key Trials
PRESERVENEJM 2018IV bicarbonate vs saline + N-acetylcysteine vs placebo in 5,177 high-risk CKD patients getting angiography β no difference. Killed the routine "renal protection cocktail." Use isotonic crystalloid alone.
STARRT-AKINEJM 2020Accelerated vs standard RRT initiation in severe AKI (n=3,019) β no mortality benefit; accelerated arm had more dialysis dependence at 90 days. Wait for clinical indication.
AKIKINEJM 2016Confirmed STARRT-AKI's finding earlier: early vs delayed RRT showed no mortality difference; nearly half in delayed arm never needed RRT.
SMARTNEJM 2018Balanced crystalloids (LR/Plasma-Lyte) vs 0.9% saline in 15,802 ICU patients β reduced major adverse kidney events (14.3% vs 15.4%). Avoid hyperchloremic acidosis from saline.
BaSICS / PLUSJAMA 2021 / NEJM 2022Confirmed and refined SMART β balanced crystalloids preferred for resuscitation in critical illness, especially in sepsis and TBI subgroups.
EMPA-KIDNEYNEJM 2023Empagliflozin in CKD (with or without DM, with or without albuminuria) β 28% reduction in CKD progression/CV death. SGLT2i now a foundational kidney-protective therapy.
6
Sepsis & Septic Shock
Based on Surviving Sepsis Campaign 2021 Guidelines & SCCM Hour-1 Bundle
Learning Objective: Recognize sepsis early, execute the Hour-1 bundle, and know when to escalate to vasopressors and steroids.
Definitions (Sepsis-3)
Sepsis: life-threatening organ dysfunction from dysregulated host response β SOFA β β₯2 with suspected infection.
Septic shock: sepsis + vasopressor requirement to keep MAP β₯65 AND lactate >2 despite resuscitation.
Screening: qSOFA (RR β₯22, AMS, SBP β€100) or NEWS β qSOFA is for screening, not diagnosis.
Hour-1 Bundle
Measure lactate (re-measure if >2).
Blood cultures Γ 2 before antibiotics (do not delay abx >45 min).
Broad-spectrum antibiotics within 1 h for septic shock; within 3 h for possible sepsis without shock.
30 mL/kg crystalloid for hypotension or lactate β₯4 (balanced crystalloid preferred).
Vasopressors if MAP <65 after fluids β start norepinephrine first (peripheral access OK to start).
Beyond the First Hour
Add vasopressin (0.03 U/min) when norepi >0.25β0.5 Β΅g/kg/min.
Hydrocortisone 200 mg/d if persistent shock on pressors (Class 2b).
"Fluid responsiveness" matters more than blindly giving 30 mL/kg in patients with HF/CKD β consider passive leg raise, IVC variation, dynamic indices.
Albumin reasonable adjunct in patients receiving large volumes of crystalloid (SAFE, ALBIOS β neutral but safe).
Mortality drops with each hour antibiotics are not delayed in septic shock β get cultures fast, but do not let them delay therapy.
Stress-dose steroids: hydrocortisone 200 mg/d IV (50 mg q6 h or continuous) in pressor-dependent septic shock. APROCCHSS adds fludrocortisone 50 Β΅g PO/d.
VTE prophylaxis, glucose 140β180, lung-protective ventilation (Vt 6 mL/kg IBW, plateau <30) all part of the bundle.
Re-evaluate antibiotics daily β narrow once organism known; stop if no infection identified after 48β72 h.
Lactate Β· Cultures Β· Abx Β· Fluids Β· PressorsNorepi first
Key Trials
ARISE / ProCESS / ProMISeNEJM 2014β2015Three large RCTs across 3 continents: protocolized "Early Goal-Directed Therapy" (Rivers) was not superior to usual care β simplified the bundle and dropped CVP/ScvOβ targets.
ADRENALNEJM 2018Hydrocortisone 200 mg/d in septic shock (n=3,800) β no mortality benefit at 90 d, but faster shock resolution and shorter ICU stay.
APROCCHSSNEJM 2018Hydrocortisone + fludrocortisone in septic shock β 90-d mortality 43.0% vs 49.1% (RR 0.88). Established add-on steroids in severe shock.
CLOVERSNEJM 2023Restrictive (early pressors) vs liberal (more fluid) strategy in septic shock β no 90-d mortality difference. Either approach acceptable; tailor to patient.
SMARTNEJM 2018Balanced crystalloids reduced major adverse kidney events vs saline in critically ill (see also Talk 5). Particularly relevant in sepsis.
VICTAS / VITAMINSJAMA 2020 / 2020Vitamin C + thiamine + hydrocortisone "Marik cocktail" β no benefit. Cooled enthusiasm for high-dose vitamin C in sepsis.
7
COPD Exacerbation
Based on GOLD 2025 Report & ATS/ERS guidelines on COPD exacerbations
Learning Objective: Recognize and grade an AECOPD, deliver the bronchodilatorβsteroidβantibiotic core, and know when to escalate to NIV.
Severe: requires hospitalization or ED visit; ABG with respiratory acidosis.
Use the new Rome severity classification: dyspnea (VAS), RR, HR, SpOβ/FiOβ, CRP, ABG.
Treatment Core β "ABCs of COPD"
Albuterol + ipratropium nebs q1β4 h.
Steroids: prednisone 40 mg PO Γ 5 days (REDUCE trial β no benefit beyond 5 d).
Antibiotics if β₯2 cardinal symptoms with purulence, or mechanical ventilation: azithromycin / doxycycline / amox-clav Γ 5 d. Cover Pseudomonas if frequent exacerbations or recent abx.
Don't be afraid of oxygen β but titrate. Hypoxia kills faster than hypercapnia. AVOID Venturi-based runaway FiOβ in COβ retainers.
Frequent exacerbator phenotype (β₯2/yr or 1 hospitalization) β escalate to triple therapy (LABA/LAMA/ICS) per GOLD; mortality benefit shown with budesonide-containing triple (ETHOS).
Consider azithromycin prophylaxis 250 mg daily in select non-smokers with frequent exacerbations despite optimal therapy (Albert NEJM 2011).
Add eosinophil count: peripheral eos β₯300 favors ICS-responsive phenotype; eos <100 β ICS unlikely to help and may β pneumonia risk.
Pulmonary rehab within 4 weeks of hospitalization reduces readmission and mortality (Class I).
Long-term Oβ if PaOβ β€55 or SpOβ β€88% at rest β only intervention besides smoking cessation proven to reduce mortality in COPD.
SABA Β· Steroids Β· Antibiotics Β· BiPAPSpOβ 88β92%
Key Trials
REDUCEJAMA 20135 days vs 14 days of prednisone 40 mg in AECOPD β non-inferior re-exacerbation; less steroid exposure. Basis for short-course steroids.
Plant et al. (BiPAP)Lancet 2000NIV in hypercapnic AECOPD on the ward β reduced intubation (15% vs 27%) and in-hospital mortality (10% vs 20%). Use early in pH 7.25β7.35.
Anthonisen criteriaAnn Intern Med 1987Original RCT defining cardinal symptoms (dyspnea, sputum volume, purulence) β antibiotics benefit those with β₯2. Still the basis of antibiotic decision today.
IMPACTNEJM 2018Triple therapy (fluticasone/umeclidinium/vilanterol) vs LAMA/LABA β fewer moderate-severe exacerbations; mortality signal favoring triple.
ETHOSNEJM 2020Triple ICS/LAMA/LABA (budesonide-based) vs LAMA/LABA β 49% reduction in all-cause mortality at higher ICS dose. Confirmed mortality benefit of triple therapy.
Albert (Macrolide)NEJM 2011Daily azithromycin 250 mg in COPD patients with prior exacerbation β reduced exacerbation frequency (HR 0.73). Watch for QTc prolongation and hearing loss.
8
Upper & Lower GI Bleeding
Based on ACG 2021 UGIB & 2023 LGIB guidelines Β· AASLD variceal bleeding guidance
Learning Objective: Resuscitate first, localize the bleed, give the right empiric drug for the suspected etiology, and know endoscopy timing.
Localize
UGIB (proximal to ligament of Treitz): hematemesis, coffee-ground emesis, melena, BUN/Cr >30. Common: PUD, varices, Mallory-Weiss, esophagitis, malignancy, AVM.
LGIB: hematochezia (but brisk UGIB can present as hematochezia + shock β always consider NG lavage or EGD first if unstable).
Common LGIB: diverticulosis, angiodysplasia, ischemic colitis, hemorrhoids, malignancy, post-polypectomy.
Risk Stratify
Glasgow-Blatchford for UGIB: score 0β1 β outpatient management possible.
Oakland score for LGIB: β€8 may be safe for discharge.
Predictors of severe bleed: hemodynamic instability, ongoing bleeding, transfusion need, low Hb, comorbidities, anticoagulation.
Consider early TIPS within 72 h in high-risk variceal bleeders (Child-Pugh B with active bleeding at endoscopy or Child-Pugh C β€13) β reduces rebleeding and mortality.
Restart anticoagulation as soon as hemostasis is durable β usually ~7 days, balanced against thrombotic risk; resume DOACs over warfarin where possible.
ASA for secondary CV prevention should be resumed within ~3β5 days after hemostasis (don't permanently stop) β withdrawal triples MACE risk (Sung 2010).
Stigmata of recent bleeding on EGD (Forrest classification): active bleeding, visible vessel, adherent clot β endoscopic hemostasis (clip/cautery/injection) + IV PPI 72 h.
LGIB workup if colonoscopy negative in ongoing bleed: tagged RBC scan, CT angio, or push enteroscopy/capsule for small bowel sources.
Reverse anticoagulation cautiously: 4F-PCC for warfarin, idarucizumab (dabigatran), andexanet alfa (apixaban/rivaroxaban β variable availability). Vitamin K for warfarin always added.
Resuscitate β Localize β Drug β ScopeHb < 7 transfuse
Key Trials
VillanuevaNEJM 2013Restrictive (Hb <7) vs liberal (Hb <9) transfusion in 921 UGIB patients β restrictive improved 6-week survival (95% vs 91%) and lowered rebleeding. Transformed transfusion thresholds.
TRIGGERLancet 2015Cluster RCT in UK confirming feasibility and safety of restrictive Hb <8 strategy in UGIB β supports restrictive approach broadly.
Lau (Pre-EGD PPI)NEJM 2007High-dose IV PPI before EGD in UGIB β reduced active bleeding stigmata at endoscopy and need for endoscopic therapy. Doesn't reduce rebleeding/mortality.
Early TIPS (GarcΓa-PagΓ‘n)NEJM 2010Pre-emptive TIPS within 72 h in high-risk variceal bleed β 1-year survival 86% vs 61%. Drove early-TIPS recommendation in AASLD/Baveno.
Sort (SBP albumin)NEJM 1999Albumin 1.5 g/kg + 1 g/kg in SBP β reduced renal failure (10% vs 33%) and in-hospital mortality (10% vs 29%). Key to current SBP protocol.
Chavez-Tapia (Antibiotics in cirrhosis GIB)Cochrane 2010Prophylactic antibiotics in cirrhotic patients with GI bleeding β reduced bacterial infection, rebleeding, and mortality. Ceftriaxone 1 g IV Γ 7 d standard.
9
Decompensated Cirrhosis
Based on AASLD Practice Guidance on Cirrhosis (2021) Β· Acute-on-Chronic Liver Failure (2024)
Learning Objective: Recognize and treat the four classic decompensations β variceal bleed, ascites/SBP, hepatic encephalopathy, and hepatorenal syndrome.
Decompensations
Variceal hemorrhage
Ascites & SBP
Hepatic encephalopathy
Hepatorenal syndrome (HRS-AKI)
Each carries dramatically worse prognosis β assess for transplant eligibility (MELD 3.0, Na included).
Ascites & SBP
Diagnostic paracentesis on every admission and any clinical deterioration.
SAAG β₯1.1 = portal HTN. Total protein <1 β SBP risk.
SBP: ascitic PMN β₯250 β ceftriaxone 2 g IV Γ 5 d + albumin 1.5 g/kg day 1, 1 g/kg day 3 (reduces HRS & mortality).
Lactulose titrated to 2β3 soft stools/day (1st line).
Rifaximin 550 mg BID for recurrence prevention (HE-2 trial).
Check ammonia is not diagnostic β treat clinically. Identify and reverse precipitant.
HRS-AKI
AKI in cirrhosis with no other cause; bland UA, FENa <1%, no improvement with 48 h albumin challenge (1 g/kg/d).
Terlipressin + albumin (CONFIRM trial) β first-line in US since 2022.
Alternatives: norepinephrine + albumin, or midodrine + octreotide + albumin (outside ICU).
Definitive treatment = liver transplant.
Clinical Pearls
"Bad meds in cirrhosis": NSAIDs, aminoglycosides, IV contrast (relative), benzos (use sparingly β prefer oxazepam/lorazepam, no active metabolites), high-dose acetaminophen (limit 2 g/d if active drinker β but acetaminophen is preferred analgesic over NSAIDs).
Vaccinate: Hep A & B, pneumococcal, flu, COVID, RSV, Tdap. All cirrhotics should also receive a one-time PPSV23 + age-appropriate PCV.
HCC screening: abdominal US Β± AFP every 6 months in all cirrhotics β improves survival when caught early.
Refer for transplant at MELD 3.0 β₯ 15, refractory ascites, hepatopulmonary syndrome, recurrent SBP, HRS, or recurrent variceal bleeding despite TIPS.
SBP prophylaxis: norfloxacin/cipro daily for prior SBP; ceftriaxone Γ 7 d during any GIB; consider in low-protein ascites + advanced disease.
Coagulopathy in cirrhosis is "rebalanced" β INR doesn't reflect bleeding risk; prophylactic FFP not recommended.
Tap every admitAlbumin saves kidneys
Key Trials
Sort (SBP)NEJM 1999Cefotaxime + IV albumin vs cefotaxime alone for SBP β reduced renal failure (10% vs 33%) and in-hospital mortality (10% vs 29%). Albumin (1.5 g/kg D1, 1 g/kg D3) is now standard.
HE-2 (Bass)NEJM 2010Rifaximin 550 mg BID for HE recurrence prevention (n=299) β 58% reduction in HE breakthrough (HR 0.42). Standard secondary prophylaxis with lactulose.
CONFIRMNEJM 2021Terlipressin + albumin vs placebo + albumin in HRS-1 β HRS reversal 29% vs 16%. Led to 2022 FDA approval. Watch for respiratory failure and ischemia.
ANSWERLancet 2018Long-term IV albumin (40 g BIW for 18 months) in decompensated cirrhosis with ascites β reduced 18-month mortality (38%β22%) and complications. Use varies by region.
Moreau (CANONIC, ACLF)Gastroenterology 2013Defined Acute-on-Chronic Liver Failure as a distinct syndrome with high short-term mortality. Drove early ICU triage and transplant evaluation.
Early TIPS (GarcΓa-PagΓ‘n)NEJM 2010See Talk 8 β pre-emptive TIPS in high-risk variceal bleed reduced rebleeding and mortality.
10
Atrial Fibrillation with RVR
Based on 2023 ACC/AHA/ACCP/HRS Guideline for the Diagnosis and Management of AFib
Learning Objective: Decide rate vs rhythm control, choose the right rate-control drug for the comorbidity, and stratify stroke and bleeding risk for anticoagulation.
Cardioversion: if AFib <48 h or on therapeutic AC β₯3 wk or TEE rules out clot.
2023 guidelines emphasize early rhythm control within 1 year (EAST-AFNET 4) β reduces CV events.
Catheter ablation Class I for symptomatic paroxysmal AFib refractory to drugs; Class 2a for persistent.
Anticoagulation
CHAβDSβ-VASc (men β₯2, women β₯3) β anticoagulate. Score 1 (men) or 2 (women) β consider.
DOACs > warfarin in non-valvular AFib (Class I).
Warfarin still preferred in moderate-severe mitral stenosis and mechanical valves.
Use HAS-BLED to identify modifiable bleeding risks β not to withhold AC.
LAA occlusion (Watchman) for those with high stroke + contraindication to long-term AC.
Clinical Pearls
2023 guideline: AFib is now staged (1β4) like cancer/HF β with focus on lifestyle (weight loss to BMI <27, OSA treatment, alcohol <3 drinks/week, aerobic exercise) as primary therapy. LEGACY trial: 10% weight loss β 6Γ more AFib-free survival.
Don't slow the rate in septic AFib aggressively β fix the underlying problem first; Ξ²-blockers can drop CO.
Diltiazem drip works fast: 0.25 mg/kg bolus β 5β15 mg/h drip; bridge to PO once controlled.
WPW + AFib: avoid AV nodal blockers (BB, CCB, digoxin, adenosine) β use procainamide or cardiovert. Wide irregular rhythm = think this.
"Pill in pocket" approach: flecainide or propafenone PO + AV nodal blocker for select paroxysmal AFib without structural heart disease.
Reversible AFib: post-op (especially cardiothoracic), thyrotoxicosis-induced, alcohol "holiday heart," sepsis-induced β still anticoagulate based on CHAβDSβ-VASc; high recurrence risk.
Stable? Β· Rate vs Rhythm Β· CHAβDSβ-VAScFind PIRATES
Key Trials
EAST-AFNET 4NEJM 2020Early rhythm control (within 1 yr of dx) vs usual care in 2,789 patients β 21% reduction in CV death/stroke/HF/ACS hospitalization (HR 0.79). Shifted philosophy toward early rhythm control.
CASTLE-AFNEJM 2018Catheter ablation vs medical therapy in AFib + HFrEF β 38% reduction in death or HF hospitalization (HR 0.62). Class I indication for ablation in this population.
RACE IINEJM 2010Lenient (HR <110) vs strict (HR <80) rate control β non-inferior at 3 years for CV death/morbidity. Drove move to lenient targets in stable patients.
ARISTOTLENEJM 2011Apixaban vs warfarin in non-valvular AFib β fewer strokes (HR 0.79), less major bleeding (HR 0.69), and lower all-cause mortality (HR 0.89). DOACs preferred over warfarin.
RE-LY / ROCKET-AF / ENGAGENEJM 2009β2013Dabigatran, rivaroxaban, edoxaban each non-inferior or superior to warfarin for stroke prevention with similar/less bleeding. Reinforced DOAC class effect.
PROTECT-AF / PREVAILJAMA 2014 / JACC 2014Watchman LAA closure vs warfarin β non-inferior for stroke prevention; option for patients with high stroke risk and contraindication to long-term anticoagulation.
LEGACYJACC 201510% weight loss in obese AFib patients β 6Γ higher AFib-free survival than minimal weight loss. Supports lifestyle as primary therapy in 2023 guideline.
11
Alcohol Withdrawal Syndrome
Based on ASAM 2020 Clinical Practice Guideline on Alcohol Withdrawal Management
Learning Objective: Predict severity, pick a benzodiazepine strategy, and prevent the lethal complications β DTs, seizures, and Wernicke's.
Timeline
6β12 h: minor β tremor, anxiety, insomnia, GI upset, mild autonomic.
Adequate benzodiazepines save lives in DTs β under-treatment kills more often than over-sedation.
Phenobarbital is gaining ground β single-dose 10 mg/kg loading or symptom-triggered dosing shortens LOS, reduces ICU need, and reaches both GABA-A and reduces glutamatergic activity. Know your hospital's protocol.
Don't use antipsychotics alone β they lower seizure threshold and don't address GABA deficit. Adjunct only if hallucinations remain after adequate benzo loading.
"Wet brain" is preventable: Wernicke triad = ophthalmoplegia, ataxia, encephalopathy β permanent Korsakoff (anterograde amnesia, confabulation) if untreated.
Outpatient detox is reasonable for low-risk patients: PAWSS <4, no prior DTs/seizures, no severe medical/psych comorbidity, supportive home, can return for daily checks.
Don't miss alternative diagnoses: sepsis, intracranial bleed (falls common), hypoglycemia, hepatic encephalopathy, hyperthyroidism, sympathomimetic toxicity β all can mimic AWS.
Address co-occurring tobacco use disorder β high overlap, motivation often comes simultaneously.
PAWSS Β· CIWAThiamine before glucose
Key Trials
SaitzJAMA 1994Symptom-triggered (CIWA-driven) vs fixed-schedule chlordiazepoxide in alcohol withdrawal β reduced total benzodiazepine use (100 vs 425 mg) and treatment duration (9 h vs 68 h). Foundational for CIWA-based dosing.
RosensonJ Emerg Med 2013Single IV phenobarbital 10 mg/kg in ED for AWS β reduced ICU admission (8% vs 25%) and rate of complications. Drove protocol changes toward phenobarbital-based pathways.
GoldCrit Care Med 2007Phenobarbital + benzo for severe AWS in ICU β reduced ICU LOS and need for mechanical ventilation vs benzo alone.
COMBINEJAMA 2006Naltrexone + medical management vs placebo in AUD (n=1,383) β naltrexone improved % days abstinent. Established naltrexone as first-line MAT.
Mason (Acamprosate)Cochrane 2010Acamprosate reduced relapse to any drinking (RR 0.86); preferred in liver disease since not hepatically metabolized.
Caine et al. (Wernicke)JNNP 1997Demonstrated under-recognition of Wernicke's; only 16% of confirmed cases had classic triad. Drove low-threshold high-dose IV thiamine.