Measurement uncertainty in laboratory reports: A tool for improving the interpretation of test results

Abstract

Background: Measurement uncertainty (MU) estimation has been introduced by ISO 15189 for the accreditation of clinical laboratories. Although MU reporting is not required, its inclusion in medical reports is of potential assistance to physicians in results interpretation.

Methods: MU reporting was evaluated with respect to different test purposes, namely comparison with reference intervals (RI), patient monitoring or comparison with clinical decision limits. Clinical Biochemistry, Hematology, Coagulation and Clinical Immunology measurands were used as examples. Assuming Gaussian RI distribution, the probability of retesting due to MU was determined by simulations. Significant MU variations were compared against the reference change value (RCV) and clinical decision limits.

Results: Three potential scenarios emerged for RI. For 12 measurands, depending on the MU interval, a potential change in results interpretation was found only for Sodium and S-Protein. On considering only the results within RI, simulations confirmed that up to 8.6% of MU intervals encompassed the RI limits, thus potentially leading to retesting. For tests used in patient monitoring, significant MU variations were comparable to those calculated by RCV, with the exception of CEA. For tests results evaluated with respect to clinical decision limits, on including MU, the clinical interpretation may be improved (e.g. for tPSA).

Conclusion: The findings made in the present study, which considers real MU data and hypothetical results obtained for a series of measurands, support the concept that MU may aid the physician's interpretation thus ensuring reliable clinical decision making.

Keywords: Clinical decision point; ISO15189:2012; Laboratory reports; Measurement uncertainty; Reference change value; Reference intervals.