Capillary and venous point-of-care testing of lactate concentrations in the emergency department: a prospective observational study

Abstract

BackgroundHyperlactatemia is a sensitive marker for early identification of patients who are critically ill. Rapid capillary lactate measurement using a handheld device can hasten the identification of patients at risk. We aimed to compare capillary lactate with venous blood lactate assessed using a blood gas analyser in the emergency department. MethodsThis was a prospective observational study of patients presenting to the emergency department in Hong Kong SAR. Venous and fingertip capillary blood samples were collected for lactate analysis. Venous lactate concentrations measured using a point-of-care blood gas analyser was used as a reference method (VL-Ref). Capillary lactate concentrations were measured using two handheld analysers: (1) Nova StatStrip Xpress (CL-Nova); and (2) Lactate Scout+ (CL-Scout+). Venous lactate measurements were also done using two handheld analysers (VL-Nova and VL-Scout+). Bland-Altman analysis was used to determine the agreement between different methods of measurement. Ethics approval was obtained from the Institutional Review Board of the Chinese University of Hong Kong (CREC 2015.624). Written consent was obtained either from the patient or a relative in all cases. The study is registered with ClinicalTrials.gov (NCT02694887). FindingsBetween March and July, 2016, we recruited 240 patients (aged 18 years or older) who were triage category 3 (urgent) presenting to the emergency department. Of 240 patients (mean age 69·9 years; 54% men), lactate concentrations were in the range 0·70–5·38 mmol/L (mean 1·96mmol/L). 151 patients (63%) had lactate concentrations (VL-Ref) less than 2 mmol/L, and 89 patients (37%) had lactate concentrations of at least 2 mmol/L. The results of a Bland-Altman analysis showed that percentage of mean difference were −6·2% between VL-Ref and CL-Scout+, 29·9% between VL-Ref and CL-Nova, with 95% limits of agreement of −79·3 to 66·8% and −43·8 to 103·6%, respectively. Regarding venous blood lactate, the results showed bias values of 13% between VL-Ref and VL-Scout+, 55·3% between VL-Ref and VL-Nova, with limits of agreement being −23·9 to 49·9% and 20·7 to 90·0%, respectively. InterpretationAn overall low systemic bias were observed in CL-Scout+ (bias −6·2%) and VL-Scout+ (bias 13%), suggesting the potential clinical utility of the Scout+ handheld analyser. However, the wide range of limits of agreement implies poor precision of using fingertip capillary blood for lactate testing. Further studies are needed to validate the use of capillary lactate in emergency settings. FundingDirect Grant for Research, The Chinese University of Hong Kong.