151. Accuracy of the HumaSensplus point-of-care uric acid meter using capillary blood obtained by fingertip puncture
- Author
-
Jean-Pierre Riveline, Jean-François Gautier, Anna Abrahamsson, Tiphaine Vidal-Trecan, Adam Platt, Jean-Marie Launay, Pierre Clerson, Jeffrey N. Miner, Thomas Bardin, Glen Hughes, Stéphanie Fabre, and Pascal Richette
- Subjects
Male ,Letter ,lcsh:Diseases of the musculoskeletal system ,Gout ,Intraclass correlation ,Capillary blood ,Point-of-Care Systems ,Hyperuricemia ,Punctures ,Hematocrit ,Point-of-care meter ,03 medical and health sciences ,chemistry.chemical_compound ,0302 clinical medicine ,Uric acid meter ,Linear regression ,Medicine ,Humans ,030212 general & internal medicine ,Point of care ,030203 arthritis & rheumatology ,Blood Specimen Collection ,medicine.diagnostic_test ,Receiver operating characteristic ,business.industry ,Reproducibility of Results ,Middle Aged ,medicine.disease ,Confidence interval ,ROC Curve ,chemistry ,Point-of-care ,Uric acid ,Female ,lcsh:RC925-935 ,business ,Nuclear medicine ,Research Article - Abstract
Background The uric acid (UA) level in patients with gout is a key factor in disease management and is typically measured in the laboratory using plasma samples obtained after venous puncture. This study aimed to assess the reliability of immediate UA measurement with capillary blood samples obtained by fingertip puncture with the HumaSensplus point-of-care meter. Methods UA levels were measured using both the HumaSensplus meter in the clinic and the routine plasma UA method in the biochemistry laboratory of 238 consenting diabetic patients. HumaSensplus capillary and routine plasma UA measurements were compared by linear regression, Bland-Altman plots, intraclass correlation coefficient (ICC), and Lin’s concordance coefficient. Values outside the dynamic range of the meter, low (LO) or high (HI), were analyzed separately. The best capillary UA thresholds for detecting hyperuricemia were determined by receiver operating characteristic (ROC) curves. The impact of potential confounding factors (demographic and biological parameters/treatments) was assessed. Capillary and routine plasma UA levels were compared to reference plasma UA measurements by liquid chromatography-mass spectrometry (LC-MS) for a subgroup of 67 patients. Results In total, 205 patients had capillary and routine plasma UA measurements available. ICC was 0.90 (95% confidence interval (CI) 0.87–0.92), Lin’s coefficient was 0.91 (0.88–0.93), and the Bland-Altman plot showed good agreement over all tested values. Overall, 17 patients showed values outside the dynamic range. LO values were concordant with plasma values, but HI values were considered uninterpretable. Capillary UA thresholds of 299 and 340 μmol/l gave the best results for detecting hyperuricemia (corresponding to routine plasma UA thresholds of 300 and 360 μmol/l, respectively). No significant confounding factor was found among those tested, except for hematocrit; however, this had a negligible influence on the assay reliability. When capillary and routine plasma results were discordant, comparison with LC-MS measurements showed that plasma measurements had better concordance: capillary UA, ICC 0.84 (95% CI 0.75–0.90), Lin’s coefficient 0.84 (0.77–0.91); plasma UA, ICC 0.96 (0.94–0.98), Lin’s coefficient 0.96 (0.94–0.98). Conclusions UA measurements with the HumaSensplus meter were reasonably comparable with those of the laboratory assay. The meter is easy to use and may be useful in the clinic and in epidemiologic studies.
- Published
- 2018