Clinical concordance assessment should be an integral component of laboratory method comparison studies: A regression transference of routine clinical data approach
Kalaria, Tejas Fenn, Jonathan Sanders, Anna Ford, Clare Gama, Rousseau
Kalaria, Tejas
Fenn, Jonathan
Sanders, Anna
Ford, Clare
Gama, Rousseau
Affiliation
New Cross Hospital; The Dudley Group NHS Foundation Trust
Other Contributors
Publication date
2022-02-17
Subject
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
Manufacturer-provided assay-specific reference intervals may not compensate for between assay differences leading to method related clinical discordance. Not all laboratories, however, assess clinical concordance as part of method comparison studies. We assessed the utility of method comparison data combined with routine clinical data to assess method related clinical concordance in the diagnosis and management of subclinical hypothyroidism (SCH). Passing-Bablok method comparison regression analysis was performed for both thyroid stimulating hormone (TSH) and free thyroxine (fT4) from 100 samples analysed by Abbott and Roche methods. Primary care samples indicative of SCH were identified from the laboratory information system (LIMS) of two laboratories (one with Roche methods and one with Abbott) over four months. For Roche and Abbott TSH and fT4 results, the Passing-Bablok regression equations were used to predict Abbott and Roche results respectively. The predicted results were interpreted using manufacturer-provided assay-specific reference intervals and compared to those of the previous SCH sample exchange study. On laboratory method comparison, Roche TSH and fT4 results were 30 ± 13% and 16 ± 7% higher compared to Abbott assays respectively. Of those with results indicative of SCH by Roche assays, 76.8% would have had normal thyroid function using predicted Abbott assay results and 46.9% of those with results indicative of SCH by Abbott assays would have had a biochemical indication for levothyroxine replacement using predicted Roche assay results. The results from the regression-based approach were comparable to the previous SCH sample exchange study. A regression-based approach using routine method comparison data and real-world clinical data identifies potential clinical discordance. We suggest that clinical concordance assessment be an integral component of laboratory method comparison studies.
Citation
Kalaria T, Fenn J, Sanders A, Ford C, Gama R. Clinical concordance assessment should be an integral component of laboratory method comparison studies: A regression transference of routine clinical data approach. Clin Biochem. 2022 May;103:25-28. doi: 10.1016/j.clinbiochem.2022.02.008. Epub 2022 Feb 17
Type
Journal Article