Oliguria and rising creatinine after emergency laparotomy
Mr K is a 68-year-old man on day 3 after emergency laparotomy for perforated diverticulitis. He required 4 hours of intra-operative noradrenaline and is now in the ICU, extubated and awake, with a sinus tachycardia of 104/min, MAP 72 mmHg maintained by noradrenaline 0.08 µg/kg/min, and a lactate that has fallen from 4.2 to 1.6 mmol/L. His urine output over the last 8 hours is 0.28 mL/kg/h. Creatinine has risen from a baseline of 92 µmol/L to 208 µmol/L; potassium 5.4 mmol/L, bicarbonate 19 mmol/L. He weighs 96 kg. Medications include piperacillin–tazobactam, vancomycin (trough 22 mg/L), furosemide 40 mg twice daily, and enoxaparin. Bladder scan is 60 mL. He has had no contrast in 72 hours. There is a central line and an arterial line; CVP is 11 mmHg.
1. Stage the AKI using KDIGO criteria and set out the first four diagnostic steps you would take, explaining the reasoning behind each.
Read the worked answer
KDIGO staging uses the worst of creatinine or urine output. Creatinine 208/92 = 2.26× baseline, which is stage 2 (2.0–2.9×). Urine output 0.28 mL/kg/h for 8 h satisfies <0.5 mL/kg/h for ≥6 h (stage 1) but not yet <0.3 mL/kg/h for ≥24 h (stage 3); if oliguria persists, he will reclassify as stage 3. So: stage 2 now, with a trajectory to stage 3.
First four steps:
- Volume assessment and fluid responsiveness - not CVP. Use passive leg raise or stroke volume variation if sinus rhythm and adequate tidal volume; look at capillary refill, lactate trend and cumulative balance. Rationale: prerenal components are reversible, and fluid overload worsens outcomes.
- Urine microscopy and dipstick - muddy brown casts suggest ATN; white cells or eosinophils raise acute interstitial nephritis (piperacillin–tazobactam). FeNa is uninterpretable on furosemide; use fractional excretion of urea or clinical context instead.
- Renal ultrasound - a bladder scan of 60 mL does not exclude obstruction, especially with a catheter or after pelvic surgery. Rationale: obstruction is the one rapidly reversible cause.
- Medication and haemodynamic review - stop nephrotoxins, check creatine kinase and uric acid, and assess sepsis and cardiac output. Rationale: a single treatable driver may be present.
Document daily creatinine, urine output, weight and fluid balance.
2. Two of his current medications are contributing to the trajectory. Identify them and justify what you would change, including how you would monitor.
Read the worked answer
The two main contributors are vancomycin (trough 22 mg/L, above the 15–20 mg/L target and nephrotoxic, especially with piperacillin–tazobactam) and furosemide. Piperacillin–tazobactam itself is associated with AKI, particularly in combination with vancomycin, although confounding by illness severity is debated. Furosemide does not prevent or treat AKI; in a patient who may be hypovolaemic it reduces effective circulating volume and worsens prerenal injury.
Plan:
- Stop furosemide unless there is documented fluid overload with hypoxaemia. If diuresis is needed later, use a monitored trial and reassess.
- Review the indication for vancomycin. If MRSA is not confirmed and there is no line infection or pneumonia, stop it. If it must continue, dose by AUC (target 400–600 mg·h/L) rather than trough alone, and avoid concomitant piperacillin–tazobactam if a safer alternative exists.
- De-escalate piperacillin–tazobactam once cultures and source control are known; use the narrowest effective regimen.
- Avoid NSAIDs and iodinated contrast; review enoxaparin dose for renal function.
- Monitor creatinine and urine output at least 12-hourly, and consider cystatin C if creatinine is confounded by muscle mass.
The principle is nephrotoxin stewardship: every drug should have a current indication, a dose adjusted for eGFR, and a planned stop date.
3. He remains oliguric and the team asks about starting renal replacement therapy before he gets sicker. Argue the case for and against early initiation and state your decision with criteria for revisiting it.
Read the worked answer
Indications for urgent RRT are refractory hyperkalaemia, severe metabolic acidosis, fluid overload with pulmonary oedema, uraemic encephalopathy or pericarditis, and certain poisonings. He has K 5.4 and bicarbonate 19 - neither is refractory, and he is not yet uraemic. There is no absolute indication.
The case for early RRT: it might limit fluid accumulation, control acidosis and potassium, and avoid a crisis out of hours; observational data link fluid overload to worse outcomes.
The case against: two large RCTs (AKIKI and STARRT-AKI) found no mortality benefit from early initiation, and more patients remained dialysis-dependent at 90 days in STARRT-AKI; catheter complications and intradialytic hypotension are real. Early RRT also commits staff and machine time.
My decision: watchful waiting with clear triggers - K >6.5 mmol/L or rising despite treatment, pH <7.2 with bicarbonate <12, refractory fluid overload with hypoxia, uraemic complications, or oliguria beyond 72 h with worsening azotaemia. Reassess at 12 and 24 h, document the plan, and do not place vascular access just in case without a trigger.