Guided example: one hemolyzed pool, two analytes, two outcomes

A serum pool is split into two aliquots. The control aliquot receives 100 microliters of vehicle with no interferent. The test aliquot receives 100 microliters of a hemolysate working solution at a concentration chosen to reproduce a moderate hemolysis index, following the paired-aliquot design CLSI EP07 describes. Both aliquots come from the same base pool, receive the same added volume, and are analyzed under the same run conditions, so any difference between them is attributable to hemolysis rather than to unequal dilution.

Analyte 1, modeled on potassium, reads 4.2 mmol/L in the control aliquot and 5.1 mmol/L in the hemolyzed aliquot. Hemolysis releases intracellular potassium from red cells, producing this false-high potassium result. Applying the EP07 percent-interference formula, (5.1 − 4.2) / 4.2 × 100 = +21.4%; the table and chart show +21% only when rounding to whole percentages. This laboratory has declared an allowable-interference criterion of ±10% for Analyte 1, a local, director-approved value stated for this method, not a universal cutoff. The hemolyzed result for Analyte 1 exceeds that declared criterion.

Analyte 2, modeled on a sodium-type analyte known to be hemolysis-tolerant, reads 140 mmol/L in the control aliquot and 141 mmol/L in the hemolyzed aliquot. The same formula gives about +0.7% interference, well inside the ±10% criterion declared for this method. The identical hemolyzed specimen that unacceptably biased Analyte 1 leaves Analyte 2 essentially unaffected.

Both aliquots carried the same hemolysis index, because both came from the same hemolyzed pool. If a laboratory suppressed or qualified every result on a hemolyzed specimen using one panel-wide rule, it would either needlessly withhold a perfectly reliable Analyte 2 result or, in the opposite failure, release a biased Analyte 1 result because the flag looked tolerable for the panel as a whole. Assay-specific thresholds, verified individually, are what let a laboratory report Analyte 2 and hold or qualify Analyte 1 from the same tube.

This is also why an unflagged specimen for Analyte 1 is not automatically safe. If the laboratory's hemolysis index alert level for Analyte 1 had been set above the true point where +10% interference is crossed, a specimen could clear the flag and still carry a biased result. The declared criterion has to be verified against the analyte's actual behavior, not assumed from the index alone.

The same specimen can be reportable for one analyte and unreportable for another at the identical hemolysis index, so an interference decision is made analyte by analyte, never panel by panel.