Comparison of the Performance of Point-of-Care and Device Analyzers to Hospital Laboratory Instruments

Wanvisa Boonlert, Porntip H. Lolekha, Gerald J Kost, Somsak Lolekha

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Point-of-care testing plays an important role in critical care medicine. This study evaluated the performance of the OPTI CCA and OMNI 9 critical care analyzers by comparing them to our currently used routine instruments (Stat Profile Ultra C, CRT, Dimension RxL, and Cell Dyn). The authors used least squares linear regression, the correlation coefficient, mean bias, and Student t test for data analysis. Three levels of aqueous control material were used to perform within-run and between-day evaluation of imprecision, as well as recovery studies, and arterial whole-blood and plasma obtained from critically ill patients were used to perform the comparison study. For within-run and between-day imprecision, the coefficients of variation of analyte measurements obtained with the OPTI and OMNI were within acceptable limits, and the recovery of analytes was close to 100%. Most comparison results from the OPTI and OMNI correlated well with results from currently used routine instruments. Most analytes on the OPTI and OMNI showed acceptable agreement with small mean biases, except pO2Na+and Cl-. Therefore, users should check these analytes and consider the potential clinical significance of such bias. Otherwise, the OPTI CCA and the OMNI 9 are suitable for analysis of samples from patients in critical care.

Original languageEnglish (US)
Pages (from-to)172-178
Number of pages7
JournalPlastic Surgical Nursing
Issue number3
StatePublished - 2003


  • accuracy
  • evaluation
  • imprecision
  • point-of-care testing
  • precision

ASJC Scopus subject areas

  • Surgery
  • Medical–Surgical
  • Advanced and Specialized Nursing


Dive into the research topics of 'Comparison of the Performance of Point-of-Care and Device Analyzers to Hospital Laboratory Instruments'. Together they form a unique fingerprint.

Cite this