Drowsy inpatient with a sodium of 121 drawn above a running infusion
An inpatient on the surgical ward has 0.45% sodium chloride with 5% dextrose running at 75 mL/h through a right forearm cannula. At 06:00 the overnight nurse requests "stat electrolytes" because the patient has become drowsy. The phlebotomist draws from the right antecubital fossa, about 4 cm above the cannula, using a 5 mL serum separator tube; the tourniquet is released after 90 seconds and the sample reaches the laboratory at 06:40. Results: sodium 121 mmol/L, potassium 2.9 mmol/L, chloride 86 mmol/L, bicarbonate 24 mmol/L, glucose 68 mg/dL, total protein 4.1 g/dL, albumin 2.2 g/dL, calcium 7.4 mg/dL. Every chemistry result from yesterday was within reference limits.
1. What single explanation accounts for this pattern? Identify the results that support it.
Read the worked answer
The specimen is diluted by the infusing fluid, drawn from a vein proximal to the cannula.
The discriminating feature is that total protein and albumin have both fallen - 4.1 g/dL and 2.2 g/dL - alongside the electrolytes. A genuine hyponatraemia of 121 mmol/L does not lower albumin and total protein; when they decline together with sodium, potassium, chloride, and glucose, the plasma water itself has been diluted. The bicarbonate is normal, consistent with 0.45% saline plus dextrose, which contains no buffer. Sodium and chloride falling together, with no anion gap, also fits dilution rather than a true disorder.
The reported glucose of 68 mg/dL is itself a hint: dextrose-containing fluid passing the draw site usually raises the glucose, but the analyser result here is at the low end, which argues against sampling pure infusate and for a partially diluted specimen.
The clinical danger is that a sodium of 121 in a drowsy patient can precipitate hypertonic saline.
2. The patient is symptomatic. How do you handle the result and the clinician?
Read the worked answer
Do not release 121 mmol/L as a final result. The patient is symptomatic, so silence is also unsafe - the clinician is acting on the assumption that a number exists.
Steps in order:
- Telephone the nurse and the ordering provider now, stating that the sample was drawn above a running infusion, that the results are likely falsely low, and that treatment should not be based on them.
- Recollect from the opposite arm, or from a site distal to the cannula and away from the infusion, and process immediately.
- Release the original panel only as a held/flagged result with an explicit comment, never as a clean value on the flowsheet.
- If the redraw also shows a low sodium, then the hyponatraemia is real and treatment is appropriate - the redraw is what makes that judgement possible.
- Document the notification, the time, and who was spoken to.
Remember the patient could genuinely be hyponatraemic; the job is to remove the artefact, not to dismiss the symptom.
3. The ward manager asks what process change would stop this recurring. What do you propose?
Read the worked answer
Attack it at the point of collection and at the point of ordering.
- Mark the IV site. A dated sticker at the cannula insertion site, plus a note in the chart, so anyone drawing blood can see which arm to avoid.
- Rule with a rationale: never draw proximal to a running infusion; use the contralateral arm, or a site distal to the cannula if the other arm is inaccessible. Staff follow rules they understand far better than rules they memorise.
- Order entry flag. Where the laboratory information system allows, flag patients with active infusions on the draw list for the collector.
- Feedback loop. The specimen was collected by a person and accepted by the laboratory without question - the 4.1 g/dL protein next to a low sodium should have triggered a call before reporting, so build that check into the pathologist review.
- Audit contaminated-chemistry recollections monthly and share the count with the ward; what gets measured gets managed.