Shock in the surgical patient — classify, resuscitate, defend the trials

Examiner probes

1. Classify him aloud — category, Sepsis-3 status, and the score you use where he is now.^[1][2]
2. Resuscitate him: your fluid, your evidence for and against balanced crystalloids, and where EGDT sits in your answer.^{[15][11][14][20]}
3. Pressors and targets: first agent, pressure floor, perfusion endpoints — and what you do with his lactate at hour 18 if perfusion is normal.^{[22][23][27][9]}
4. He has a tense abdomen with murky drain output on pressors — when do you operate, and what is the operation trying to achieve?^[8][1]

Model responses

1. Distributive (septic) shock by Sepsis-3 — qSOFA 3 on the ward, SOFA in ICU. Four categories share supply-demand mismatch; this is the vascular compartment — relative hypovolaemia from redistribution, needing vasoconstrictors plus fluids.^[1] He meets shock physiology: pressor-imminent hypotension with lactate 4.8 after fluids — the pressor-plus-lactate-above-2 combination with mortality above 40%.^[2] Score qSOFA 3 here (rate 26, drowsy, systolic 86) — and move him to ICU where SOFA takes over, since qSOFA predicts best on the intermediate unit (AUC 0.82) and SOFA once ICU-involved.^[6]

2. Balanced-crystalloid challenges with dynamic testing — EGDT retired, volumes neutral. Balanced first (SMART kidney-event signal), honestly held: PLUS null and BEST-Living small — so modest gains expected, saline kept for head injury.^[15] EGDT answered in one breath: Rivers 30.5% versus 46.5% single-centre promise, then concordant multicentre negatives ending in PRISM (24.9% versus 25.4%, odds ratio 0.97, higher costs) — resuscitate to perfusion with reassessment, not to venous numbers.^[11][14] Volume strategy alone never won either: CLOVERS separated 2,134 mL with no mortality difference — so challenges with responsiveness testing, early pressors, stop when perfusion answers.^[20]

3. Norepinephrine first to MAP 65+, perfusion endpoints, lactate honesty at hour 18. Dopamine doubles arrhythmia (24.1% versus 12.4%) without mortality gain; norepinephrine carries NNT 9.^[22][23] Then perfusion — mentation, urine, refill — because ANDROMEDA Bayesian analysis gives above-90% posterior probability that perfusion-targeting beats lactate-targeting (odds ratio 0.61).^[27] At hour 18 with normal perfusion and persistent lactate: that is the metabolic second phase, not flow — do not pour more fluids.^[9]

4. Resuscitate then operate urgently — source control is the treatment. Surgical sepsis is infection plus organ dysfunction treated by source control, antibiotics and organ support — and murky drain output with a tense abdomen on pressors is an operative abdomen once resuscitation allows.^[8] The operation eliminates contamination definitively (washout, leak assessment, diversion by findings), not tentatively — with the postoperative leak-hunting and hour-1 detail owned by the sibling sepsis topic.