Hyperchloremic acidosis and AKI after massive saline resuscitation: what to buffer and when to dialyse
58-year-old man, day 3 after emergency laparotomy for perforated sigmoid diverticulitis. Creatinine 3.4 mg/dL (baseline 1.0), urine output 20 mL/h, norepinephrine 0.05 µg/kg/min. He has received 9 L of 0.9% sodium chloride since admission. Arterial blood gas on FiO₂ 0.40: pH 7.19, PaCO₂ 30 mmHg, HCO₃⁻ 11 mmol/L. Chemistry: Na 138, Cl 119, K 5.6, lactate 1.6 mmol/L, albumin 2.4 g/dL, glucose 140 mg/dL. Abdomen is soft, drain output minimal, no ascites. He is awake but tachypneic at 28 breaths/min with tidal volumes of 550 mL, and says he feels exhausted.
1. Give the complete acid-base analysis, including the albumin-corrected anion gap and the delta ratio. What does the pattern tell you about cause and about his immediate risk?
Read the worked answer
Step 1 - anion gap
AG = Na − (Cl + HCO₃) = 138 − (119 + 11) = 8 mmol/L. Corrected for hypoalbuminaemia: 8 + 2.5 × (4.0 − 2.4) ≈ 12 mmol/L, i.e. not elevated.
Step 2 - pattern
Normal-gap hyperchloraemic metabolic acidosis. Chloride 119 with bicarbonate 11 is a non-gap acidosis, and lactate 1.6 does not explain it. After 9 L of 0.9% saline this is the expected iatrogenic hyperchloraemic acidosis of large-volume chloride loading, compounded by kidney injury that limits chloride excretion.
Step 3 - respiratory compensation
Winter's formula: expected PaCO₂ = 1.5 × 11 + 8 = 24.5 ± 2, i.e. 22.5–26.5 mmHg. Measured PaCO₂ is 30 mmHg - higher than expected. There is therefore a superimposed respiratory acidosis: he is not clearing CO₂ adequately for the degree of metabolic acidosis, and his reported exhaustion suggests he is tiring.
Step 4 - delta ratio
(12 − 12)/(24 − 11) = 0, confirming no meaningful unmeasured anion accumulation; the acidosis is chloride-driven.
Bottom line: saline-related non-gap acidosis with an inadequate respiratory response. The respiratory component, not the bicarbonate number, is what can decompensate in the next few hours.
2. Is bicarbonate indicated at pH 7.19? If the pH fell to 7.10, would your answer change, and what else would you change first?
Read the worked answer
At pH 7.19 - no
In a normal-gap acidosis with compensatory tachypnoea, bicarbonate buys little. Every millimole generates CO₂ that must be exhaled, and his PaCO₂ is already inappropriately high, so buffer would worsen hypercapnia. It also adds a sodium load, lowers ionised calcium, and shifts potassium intracellularly while total-body potassium may be depleted.
What actually helps
- Stop 0.9% saline; use a balanced crystalloid (lactated Ringer's or Plasma-Lyte) if more resuscitation is needed.
- Treat the cause of kidney injury: perfusion, nephrotoxin avoidance, review the abdomen and drain for a missed source.
- Address the respiratory component - assess for fatigue, atelectasis and whether non-invasive or invasive ventilation is needed to sustain minute ventilation.
- Monitor potassium and ECG; treat hyperkalaemia medically if changes appear.
If pH fell to 7.10
That is a different problem. With oliguria (20 mL/h) and potassium 5.6, the better answer is usually renal replacement therapy, which corrects acidosis, potassium and volume simultaneously. If RRT is not immediately available, bicarbonate can bridge - but only alongside a plan to increase minute ventilation.
Bottom line: fix the chloride load and the ventilation; reserve bicarbonate and dialysis for an acidosis that is not responding, not for 7.19.
3. When would you start renal replacement therapy in this patient, and what specifically would make you start tonight?
Read the worked answer
Established indications
Refractory acidosis (pH <7.10), refractory hyperkalaemia (>6.5 mmol/L or ECG changes not responding to medical therapy), dialysable intoxication, refractory volume overload, and uraemic complications such as encephalopathy, pericarditis or bleeding. He meets none right now: potassium 5.6 without ECG change, pH 7.19 with a treatable cause, and he is awake.
What would make me start tonight
- Potassium climbing above 6.5 mmol/L or any ECG change despite calcium, insulin/dextrose and a binder.
- pH falling below 7.10 with a PaCO₂ he cannot clear.
- Oliguria progressing to anuria with worsening overload and hypoxia that diuretics cannot address.
- Uraemic encephalopathy or a pericardial rub.
How I would monitor
Hourly urine output, electrolytes every 6 hours, daily weights and fluid balance. He is KDIGO stage 3 by creatinine (3.4 against a baseline of 1.0), so the question is timing rather than whether. Trials of accelerated versus standard initiation have not shown mortality benefit from starting early without a complication, and early initiation exposes him to catheter and circuit harms.
Bottom line: correct reversible factors and watch closely; start CRRT when a complication appears, not on the creatinine value alone.