Day 4 of septic shock: antibiotic de-escalation and de-resuscitation in a patient who is now fluid-overloaded
66-year-old woman, day 4 of septic shock from ascending cholangitis treated with ERCP and a biliary stent. Norepinephrine was stopped 12 hours ago; MAP 72 mmHg, lactate 1.4 mmol/L (from 6.0), creatinine 2.1 mg/dL (baseline 1.2). She remains febrile at 38.7 °C with a WBC of 18.4 × 10⁹/L. Bile cultures grew pan-susceptible E. coli; two sets of blood cultures are negative at 72 hours. She is on meropenem 1 g every 8 hours and vancomycin. Net fluid balance is +9 L, weight is up 12 kg, and she has bibasilar crackles with SpO₂ 89% on 6 L nasal cannula. CRP 172 mg/L, procalcitonin 4.1 ng/mL (was 22). CT shows a small residual fluid collection with the stent in place and no biliary dilatation.
1. Should you de-escalate antibiotics today? State exactly what you would stop, what you would keep, and for how long.
Read the worked answer
Yes - de-escalate today
- The pathogen is known: pan-susceptible E. coli from bile, with source control achieved by the stent.
- Vancomycin has no target. There is no MRSA, no line infection and no confirmed pneumonia, and blood cultures are negative at 72 hours. Continuing it adds nephrotoxicity and Clostridioides difficile risk.
- Meropenem adds nothing against a susceptible E. coli. Narrow to ceftriaxone 2 g daily, or ampicillin if susceptibility allows the narrowest option. If an ESBL had been reported, ertapenem or meropenem would be justified - it was not.
Do not be misled by the numbers
Persistent fever, CRP 172 and a procalcitonin of 4.1 (down from 22) are not reasons to stay broad. A falling procalcitonin with a controlled source supports de-escalation. Her fever and hypoxia are at least as likely to reflect atelectasis and 12 kg of fluid gain as uncontrolled infection.
Duration
For intra-abdominal infection with adequate source control, 5–7 days of an appropriate agent is sufficient; a history of shock does not justify extending it.
Bottom line: stop vancomycin, switch meropenem to ceftriaxone, plan a 5–7 day total course, and require another explanation for the fever.
2. She is hypoxic with 12 kg of weight gain. How do you remove the fluid safely, and what would make you choose ultrafiltration instead of diuretics?
Read the worked answer
Decide first whether this is fluid or infection
SpO₂ 89% on 6 L with crackles, +9 L balance and 12 kg weight gain after shock has resolved is fluid overload until proven otherwise. Lung ultrasound showing B-lines and a plethoric IVC supports that; a focal consolidation with air bronchograms would add antibiotics and physiotherapy to the plan.
How to remove it
- Furosemide 40 mg IV bolus, then an infusion of 5–10 mg/h titrated to urine output, targeting negative balance of 1–1.5 L/day, more if tolerated.
- Check MAP, creatinine, potassium and urine output every 4–6 hours. A modest creatinine rise during successful decongestion is acceptable and often transient; oliguria with hypotension is not.
- Do not use CRRT ultrafiltration for fluid alone: there is no renal indication, and it adds a dialysis catheter and circuit. Reserve it for diuretic resistance, worsening kidney injury or hyperkalaemia.
Supportive measures
Sit her up, physiotherapy and incentive spirometry, and high-flow nasal oxygen if she remains hypoxic while diuresis takes effect. Reassess in 24 hours - falling weight with improved SpO₂ confirms the diagnosis.
Bottom line: diurese promptly and monitor renal function, treating the hypoxia as a volume problem with a defined trial before adding antibiotics.
3. On day 7 she spikes to 39.2 °C with hypotension and worsening abdominal pain. How do you approach this, and would you broaden antibiotics first?
Read the worked answer
Think before broadening
Recurrent hypotension and abdominal pain a week after stenting suggests stent occlusion or recurrent cholangitis, a residual abscess, or a new complication. Alternatives include C. difficile colitis from antibiotics, a central-line infection, and drug fever.
Workup
- CT abdomen and pelvis with contrast, and ultrasound of the biliary tree.
- Repeat blood cultures before new antibiotics; stool testing for C. difficile if there is diarrhoea.
- Blood gas, lactate, liver enzymes and bilirubin to assess the biliary tree.
- Review the stent: if occluded, she needs ERCP and stent exchange, and antibiotics alone will fail.
Antibiotic decisions
Draw cultures, then broaden empirically while source control is arranged. If the previous E. coli was susceptible, returning to a narrow agent after drainage is often correct. Do not escalate to carbapenem-resistant coverage without evidence of a resistant organism.
Bottom line: source control is the treatment. Imaging and biliary drainage come before another antibiotic, and duration should be reset from the date of effective drainage.