Considerable differences between methods for plasma D‐dimer in the diagnosis of venous thrombosis: Increased positive prediction using elevated discriminatory limits

Abstract

Four analytical methods for plasma D‐dimer have been compared in a group of 82 consecutive patients with suspected venous thrombosis, 22 of whom subsequently received a positive clinical diagnosis. Differences in diagnostic power between methods were observed. In three of the four methods tested, the use of elevated discriminatory limits (rather than conventional 95‐percentile upper reference limits) resulted in stronger positive predictive values for thromboembolism. We suggest that this approach may result in better diagnostic performance of the D‐dimer method.

• D‐dimer
• discrimination limit
• receiver‐operator characteristics
• thrombosis

Acknowledgements

We express our gratitude to Margaretha Bloom and Marlene Lindgren for excellent technical assistance and to Monika Johansson and Andreas Hillarp for helpful discussions of the manuscript.

Disclosure of potential conflict of interest

Christina Hall has previously worked for Haemochrom AB, the distributor of Chromogenix reagents in Sweden. Haemochrom AB also provided Chromogenix reagents.