Comparison of automated devices UX-2000 and SediMAX/AutionMax for urine samples screening: A multicenter Spanish study.
Clin Biochem
; 50(12): 714-718, 2017 Aug.
Article
em En
| MEDLINE
| ID: mdl-28188739
OBJECTIVES: In this study we aim to compare UX2000 (Sysmex Corp, Japan) and SediMAX/AutionMax (Arkray Factory Inc., Japan), totally automatized analyzers, against Fuchs-Rosenthal counting chamber, the gold standard technique for sediment analysis. DESIGN AND METHODS: Urine samples of 1454 patients from three Spanish hospitals were assessed for red and white blood cells (RBC; WBC) using three different techniques: flow cytometry, image-based method and Fuchs-Rosenthal counting chamber. Test strip results were subjected to concordance evaluation. Agreement was assessed by Cohen's weighted kappa for multinomial results. Sensitivity (SE) and specificity (SP) were calculated. RESULTS: The categorization of the results showed that UX-2000 had higher concordance over SediMAX for WBC (0.819 vs. 0.546) and similar for RBC (0.573 vs. 0.630). For RBC, UX-2000 had higher SE (92.7% vs. 80.3%) but lower SP (77.1% vs. 87.4%), and showed higher both SE (94.3% vs. 76.7%) and SP (94.7% vs. 88.2%) for WBC. Inter-devices test strip agreement was substantial (kappa>0.600) for all variables except for bilirubin (kappa: 0.598). Intra-device test strip agreement was similar for UX2000 and SediMAX with regard to RBC (kappa: 0.553 vs. 0.482) but better for UX2000 with regard to WBC (0.688 vs. 0.465). CONCLUSIONS: Both analyzers studied are acceptable for daily routine lab work, even though SediMAX is easier to use in laboratories thanks to its lower maintenance procedure. UX-2000 has shown to have better concordance with the gold standard method. However, it needs some improvements such as an image module in order to decrease manual microscopy review for urine samples.
Palavras-chave
Texto completo:
1
Base de dados:
MEDLINE
Assunto principal:
Urinálise
/
Contagem de Eritrócitos
/
Automação Laboratorial
/
Contagem de Leucócitos
Idioma:
En
Ano de publicação:
2017
Tipo de documento:
Article