Pocket Medicine

PULMONARYQ

SEPSIS

Fluids & vasoactive drugs

•  Early goal-directed therapy (“Rivers Protocol,” NEJM 2001;345:1368), confirm. trials pending

Insert arterial & central venous lines (NEJM 2007;356:e21; PAC not needed) and ✓ MAP, CVP & ScvO2 (central venous O2 sat, nl 60–80%) which measures O2 consumption vs. delivery (less invasive than mixed venous) w/ low SCVO2 → ↓ O2 delivery (↓ SaO2, nl SaO2 but ↓ CO or anemia) or excessive O2 consumption

Target MAP ≥65 mmHg, CVP 8–12 mmHg, & UOP ≥0.5 mL/kg/h using fluid (eg, 500 mL NS q30min) and vasopressors as needed

Target ScvO2 ≥70% using PRBCs & inotropes (dobutamine, ↑ dose as needed q15min)

When done w/in first 6 h for severe sepsis & septic shock, 42% ↓ mortality

•  Lactate clearance (≥20% / 2 h) as effective as ScvO2 to guide resuscitation (  JAMA 2010;303:739)

•  Crystalloid better than colloid for resuscitation (NEJM 2004;350:2247 & 2012;367:124 & 1901)

•  Norepi ↓ arrhyth. & mort. c/w dopamine (NEJM 2010;362:779; Crit Care Med 2012;40:725)

•  Vasopressin added to low-dose norepinephrine not superior to high-dose norepinephrine (NEJM 2008;358:877); consider if HoTN refractory to catecholamine vasopressors

•  Use PRBC w/ caution, may ↑ mortality/morbidity, ↑ risk of ARDS (Crit Care Med 2005;33:1191); ∴ goal Hb 7 unless active cardiac ischemia (NEJM 1999;340:409)

•  After early resuscitation, if ALI/ARDS, target CVP 4–6 mmHg as additional fluids may be harmful → ↑ ventilator/ICU days (NEJM 2006;354:2564; Chest 2008;133:252)

•  Pulse pressure variation >13% with respiration → likely volume-responsive (Chest 2008;133:252); only validated in passive, intubated Pts and studied in higher tidal volumes

Antibiotics

•  Start empiric IV abx w/in 1 h of recognition of severe sepsis or septic shock; every hour delay in abx admin a/w 8% ↑ in mortality (Crit Care Med 2006;34:1589)

•  If possible, obtain 2 sets of BCx before urgently starting abx (but do not delay abx)

•  Typically want broad gram-positive and gram-negative coverage, including MRSA and highly resistant gram-negative bacilli ± anaerobes

Steroids (NEJM 2003;348:727; JAMA 2009;301:2362)

•  Cortisol secretion helps predict mortality, but treatment of adrenal insufficiency is unproven (  JAMA 2000;283:1038; NEJM 2008;358:111)

•  Earlier study showed possible mortality benefit w/in 8 h of severe septic shock (SBP <90 for >1 h despite fluids & pressors) if post ACTH stim cortisol Δ ≤ 9 µg/dL (  JAMA 2002;288:862)

•  No mortality benefit to early (<72 h) empiric corticosteroids in all Pts w/ septic shock, regardless of ACTH stim; faster resolution of shock, more superinfection (NEJM 2008;358:111)

•  ? hydrocortisone 50–100 q6–8h ± fludrocortisone 50 µg daily in septic shock refractory to fluids & pressors, regardless of ACTH stim (Crit Care Med 2008;36:296)

Activated protein C

•  No longer FDA approved, no improvement in mortality (NEJM 2012;366:2055)

Intensive glycemic control (NEJM 2010;363:2540)

•  No evidence of improved outcomes in MICU population w/ intensive glycemic control

•  Intensive glycemic control to goal 80–110 mg/dL in surgical ICU population → mortality benefit, greatest if >3-d ICU stay (NEJM 2001;345:1359)

•  Repeat studies suggest intensive glycemic control → either no Δ or ↑ increased mortality, and definite ↑ hypoglyc. (  JAMA 2008;300:933; NEJM 2006;354:449; 2008;358:125; 2009;360:1283)

•  Hypoglycemia associated with mortality (NEJM 2012;367:1108)

•  Reasonable to keep glc <150 mg/dL in severe sepsis, using validated protocol (Crit Care Med 2008;36:296)