Critical potassium after a delayed, haemolysed sample from an IV arm
At 02:00 in a provincial hospital, a 58-year-old man with type 2 diabetes is admitted with vomiting and confusion. He has a fever of 38.4°C and blood pressure 92/54 mmHg. A venous blood sample is drawn from the arm with the IV infusion running; the specimen arrives in the laboratory 4 hours later, having been kept in a hot transport box at approximately 34°C. Results: sodium 131 mmol/L, potassium 6.9 mmol/L, chloride 98 mmol/L, bicarbonate 12 mmol/L, glucose 28 mmol/L, creatinine 210 µmol/L, urea 18 mmol/L, haemoglobin 140 g/L, white cells 18 ×10^9/L. The sample is visibly haemolysed (index 3+). The analyser flags a critical potassium. The clinician calls and asks whether to give intravenous calcium and insulin–dextrose immediately. A repeat sample is requested. What should the laboratory do before and while the repeat is taken, and how should the critical result be communicated?
1. Interpret the potassium and sodium results. What pre-analytical and analytical explanations must be excluded, and what should be reported now?
Read the worked answer
Interpretation. The first result is not reliable for potassium: haemolysis index 3+ releases intracellular potassium, and a sample drawn from an arm with an IV line and delayed 4 hours at 34°C adds pre-analytical error. However, the patient has acidosis (bicarbonate 12 mmol/L), renal impairment (creatinine 210 µmol/L) and hyperglycaemia, so true hyperkalaemia is plausible. Do not dismiss the value.
Immediate action. Telephone the clinician now, state that K 6.9 mmol/L is critical but may be falsely raised, and advise clinical assessment (ECG, muscle weakness, acid–base status). If ECG shows peaked T waves or widened QRS, treatment should not wait for the repeat. Collect a new sample from the opposite arm, without IV contamination, minimal tourniquet time, plain tube, rapid transport at ambient temperature, and analyse within 30 minutes. Record the phone call, the read-back and the caveat in the LIS. Do not simply cancel the result.
2. The repeat sample, collected from the opposite arm without an IV line and analysed within 30 minutes, gives potassium 6.1 mmol/L and sodium 136 mmol/L; haemolysis index 1+. How does this change your interpretation and what quality-improvement action is indicated?
Read the worked answer
Repeat result. K 6.1 mmol/L with Na 136 mmol/L and haemolysis index 1+ is still abnormal but much more plausible. The fall in potassium and correction of sodium suggest that the first sample was contaminated by IV fluid and haemolysis; however, K 6.1 mmol/L in a patient with acidosis and renal impairment can still be true hyperkalaemia, so treat the patient, not the number. Correlate with ECG, urine output, bicarbonate and creatinine, and repeat again after treatment.
Quality action. Open a non-conformity report. Review phlebotomy practice (avoid drawing above a running IV, separate tube, correct order of draw), transport time and temperature, and centrifuge/analyser settings. Retrain staff, audit haemolysis rates, and document corrective action. If the first result was already released, issue an amended report with a comment.
3. The first sample's potassium was entered into the LIS and appeared on the ward report before the laboratory called. What are the risks and the correct reporting practice for critical results?
Read the worked answer
Risk. A critical potassium released only through the LIS can be missed or acted on without the laboratory's caveat. It also creates a duplicate, confusing record if the amended result is not clearly linked.
Correct practice. Follow the written critical-value policy: phone the responsible clinician or ward within the defined time (for example 30 minutes), read back the result and the caveat, record the name of the person receiving it, the time and the action taken. The LIS report should carry a comment that the first specimen was haemolysed and that a repeat is in progress; later add the repeat result and an amended report. Audit the episode to confirm that communication occurred and that the clinician understood the uncertainty. Critical-value reporting is a patient-safety process, not a clerical step.