The accuracy and timeliness of a Point Of Care lactate measurement in patients with Sepsis

Fatene Ismail, William G. Mackay, Andrew Kerry, Harry Staines, Kevin D. Rooney

The aims of this study were to a) compare the lactate measurement of a Point of Care (POC) handheld device to near patient blood gas analysers, and b) determine the differential reporting times between the analysers. A two-staged study; method comparison and prospective observational stages, was conducted. For the first stage, blood samples were analysed on the i-STAT handheld device and the near patient blood gas analysers (GEM 4000 and OMNI S). Results were compared using Pearson correlation coefficient and Bland-Altman tests. For the second stage, we examined the differential reporting times of the POC device compared to the near patient blood gas analysers in two Scottish hospitals. Differential reporting times were assessed using Mann-Whitney test and descriptive statistics were reported with quartiles. Highly significant Pearson correlation coefficients (0.999 and 0.993 respectively) were found between i-STAT and GEM 4000 and OMNI S. The Bland-Altman agreement method showed bias values of -0.03 and -0.24, between i-STAT and GEM 4000 and OMNI S respectively. Median time from blood draw to i-STAT lactate results was 5 min (Q1-Q3 5-7). Median time from blood draw to GEM 4000 lactate results was 10 min (Q1-Q3 7.75-13). Median time from blood draw to OMNIS lactate results was 11 min (Q1-Q3 8-22). The i-STAT was significantly quicker than both the GEM 4000 and the OMNIS (each p-value < 0.001). In addition, 18 of our study samples were sent to the central laboratory for analysis due to a defect in the lactate module of OMNI S. The median time for these samples from blood draw to availability of the central laboratory results at the clinical area was 133 min. The POC handheld device produced accurate, efficient and timely lactate measurements with the potential to influence clinical decision making sooner.