How should a hemolyzed potassium sample be interpreted and what steps are needed to obtain an accurate result?

Hemolyzed Potassium Sample Management

A hemolyzed potassium sample should NOT be automatically rejected or repeated in most clinical scenarios—instead, interpret the result in clinical context, considering the degree of hemolysis, patient risk factors, and whether the elevated value would change management. The traditional practice of reflexively rejecting hemolyzed samples causes significant delays in critical care without clear patient benefit.

Understanding the Problem

Hemolysis causes falsely elevated potassium levels because intracellular potassium (which is 30-40 times higher than serum levels) leaks from ruptured red blood cells into the sample. However, pseudohyperkalemia from hemolysis must be distinguished from true hyperkalemia before treatment decisions are made 1.

Key laboratory considerations:

• Plasma potassium is typically 0.1-0.4 mEq/L lower than serum due to platelet potassium release during coagulation 1
• Hemolysis index (H-index) thresholds vary widely between institutions
• Complications such as pseudohyperkalemia from hemolysis, repeated fist clenching, and poor phlebotomy techniques need assessment prior to hyperkalemia treatment 1

When to Repeat vs. When to Accept the Result

Do NOT automatically repeat if:

1. The patient is healthy with no risk factors for true hyperkalemia 2

   • In pediatric ED patients with hemolyzed hyperkalemia, 97.9% had normal potassium on repeat testing 2
   • Only 2% had true hyperkalemia, and all had underlying conditions predisposing to hyperkalemia 2

2. BUN and creatinine are normal 2

   • Normal renal function makes true hyperkalemia unlikely in otherwise healthy patients

3. The hemolysis index is only mildly elevated

   • Recent data shows no clear benefit from avoiding specimens with H-index >20, and this practice causes average delays of 3 hours 23 minutes for critical results 3
   • Context-specific thresholds are more appropriate than universal strict cutoffs 4

DO repeat the sample if:

1. Patient has high-risk conditions:

   • CKD (especially advanced stages)
   • Heart failure
   • Diabetes
   • Taking RAASi medications (ACE inhibitors, ARBs, aldosterone antagonists)
   • History of hyperkalemia 1

2. ECG changes are present:

   • Peaked T waves
   • Prolonged QRS complexes
   • Though note: ECG findings are highly variable and not as sensitive as laboratory testing 1

3. Clinical symptoms suggest true hyperkalemia:

   • Severe muscle weakness
   • Cardiac arrhythmias

4. The result would trigger immediate treatment:

   • If potassium >6.5 mEq/L and you're considering IV calcium or insulin/glucose, confirm with a non-hemolyzed sample first unless the patient is clinically unstable

Practical Algorithm

Step 1: Assess the clinical context

• Does the patient have risk factors for hyperkalemia?
• Are there ECG changes or symptoms?
• What is the degree of hemolysis (H-index)?

Step 2: For LOW-RISK patients (healthy, normal renal function):

• Accept the hemolyzed result if potassium is normal
• Do not repeat if mildly elevated unless it would change management
• Document "likely pseudohyperkalemia from hemolysis"

Step 3: For HIGH-RISK patients:

• Obtain a careful repeat sample with proper technique
• Consider point-of-care testing if available
• Monitor ECG while awaiting results if severely elevated

Step 4: If repeat sampling is needed:

• Use proper phlebotomy technique (no fist clenching, appropriate tourniquet time)
• Process specimen promptly
• Consider plasma rather than serum (lower baseline values) 1

Critical Pitfalls to Avoid

1. Treating pseudohyperkalemia as real hyperkalemia causes unnecessary interventions (insulin-induced hypoglycemia, calcium administration) and inappropriate medication discontinuation (especially RAASi drugs that provide mortality benefit) 1

2. Excessive specimen rejection creates dangerous delays in critical care settings, particularly in NICUs and EDs where hemolysis rates approach 50-76% with strict thresholds 4

3. Correction formulas using free hemoglobin are unreliable and may produce inaccurate results—qualitative comments about likelihood of abnormality are more appropriate 5

4. Whole blood analysis (point-of-care devices) carries significant risk for undetected hemolysis, with bias ranging from -1.0 to 4.0 mmol/L compared to plasma 6

Context-Specific Thresholds

Laboratories should implement setting-specific H-index thresholds rather than universal strict cutoffs 4:

• NICU/ED: Higher thresholds acceptable (H-index ≤300) with interpretive guidance
• Inpatient/Outpatient: Moderate thresholds (H-index ≤100)
• This approach reduces unnecessary rejections to <3% while maintaining clinical safety 4