A flagged specimen and an in-range result

A basic metabolic panel comes back on a routine inpatient specimen. The potassium result reads 4.8 mmol/L, comfortably inside this method's inpatient reference interval, and nothing about the number itself looks wrong. The analyzer has also posted a moderate hemolysis index on the same specimen. Hemolysis can release intracellular potassium from red cells and falsely raise the measured potassium result. The ordering nurse asks a plain question: is this potassium real, or is it wrong in a direction that happens to look normal?

Does a method detect the intended measurand reliably, and what else in the specimen can change the result? Those are two different failure modes. A method can be working exactly as designed and still report a biased number because something in the specimen, not the analyte itself, is changing the signal the method reads.

Answering the nurse requires two separate pieces of knowledge. First, what does this specific method actually measure, and how sensitive is it to hemolysis at this index level? Second, what is the difference between a method-level property like interference and a population-level property like diagnostic sensitivity, so the technologist does not reach for the wrong tool when explaining the result?

Analytical performance terms describe what a measurement procedure does, and diagnostic performance terms describe how a test performs across a population of patients with and without a condition. Investigating a specimen when hemolysis, icterus, lipemia, biotin, or a heterophile antibody is suspected of changing a result works through both.

Before answering whether a result is trustworthy, name whether the question is about detection, about interference, or about diagnostic performance, because each has a different answer and a different next step.