Fluid responsiveness and renal replacement timing in septic shock after laparotomy

CVP of 13 mmHg does not reliably predict fluid responsiveness, and the IVC variation is also unhelpful here because the patient is in atrial fibrillation and receiving positive-pressure ventilation. The echocardiogram does not suggest gross hypovolaemia, the ScvO2 is adequate and there are bilateral pulmonary opacities, so blind fluid loading risks worsening gas exchange and oedema. I would use a dynamic test: a passive leg raise with measurement of cardiac output by echocardiography or a calibrated device, or a 250 mL crystalloid challenge over 10 minutes with a direct stroke volume or cardiac output response. A rise of more than 10% would suggest fluid responsiveness. If neither is available, I would use capillary refill time, lactate trend and the response of the blood pressure to a small, time-limited challenge rather than a large bolus. In this case I would not give a 500 mL or 1 L bolus empirically; I would give a small 250 mL challenge while measuring the response, and stop if there is no improvement. Fluid stewardship is important because the patient already has risk factors for fluid overload and AKI.

There is no absolute emergency indication at present: potassium is 5.2 mmol/L, pH is 7.28 and there is no refractory pulmonary oedema or overt uraemic complication. The AKIKI and STARRT-AKI trials showed no clear benefit from early CRRT initiation in critically ill patients with acute kidney injury, so I would not start it solely because creatinine is 210 µmol/L or urine output is low. However, this patient has persistent oliguria below 0.5 mL/kg/h, rising creatinine from a baseline of 78 µmol/L, fluid overload risk and septic shock requiring noradrenaline. I would start CRRT if he develops refractory hyperkalaemia, severe metabolic acidosis, pulmonary oedema resistant to diuretics, uraemic encephalopathy or pericarditis, or if oliguria persists with progressive fluid accumulation. In a haemodynamically unstable patient, continuous therapy is preferred over intermittent haemodialysis because it avoids rapid osmotic and volume shifts. Regional citrate anticoagulation would be my default unless there is severe liver failure, and I would review drug dosing for antibiotics and sedatives during CRRT.

I would first treat the precipitant: source control, antibiotics, careful fluid assessment and correction of magnesium and potassium. In a septic patient with a ventricular rate of 132/min, rate control is reasonable if he is haemodynamically stable, but I would avoid verapamil or diltiazem because of negative inotropy and the risk of hypotension, especially with concomitant beta-blockade. Amiodarone is a safer first-line agent in this setting, given as a loading dose followed by infusion, with monitoring for QT prolongation and hypotension. If he becomes haemodynamically unstable with hypotension, ischaemia or pulmonary oedema, I would perform synchronised electrical cardioversion. I would also consider whether the AF is contributing to the low cardiac output and whether any electrolyte abnormality, such as hypokalaemia or hypomagnesaemia, is perpetuating it. Anticoagulation decisions should be individualised and deferred until bleeding risk after laparotomy is clarified. The underlying message is that treating sepsis and correcting physiology often converts AF to sinus rhythm without aggressive rate control.