Article Text

Download PDFPDF
  1. D Datta1,
  2. J Grahamslaw1,
  3. AJ Gray1,
  4. C Walker1,2,
  5. C Graham3
  1. 1Emergency Medicine Research Group Edinburgh, Department of Emergency Medicine, Royal Infirmary of Edinburgh, NHS Lothian, Edinburgh, Midlothian, UK
  2. 2St John's Hospital at Howden, Department of Critical Care, Livingstone, West Lothian, UK
  3. 3Wellcome Trust Clinical Research Facility, University of Edinburgh, Edinburgh, Midlothian, UK


    Objectives & Background Sepsis is a common condition with around 30% hospital mortality. Measurement of lactate levels is an important part of the sepsis bundle. Although peripheral venous lactate (PV-LACT) is widely measured, most research has used arterial lactate (A-LACT). The interchangeability between the measurements is debatable, with no research supporting the widespread use of PV-LACT instead of A-LACT. If we can demonstrate that PV-LACT has good agreement it would significantly reduce patient discomfort and the potential risks of arterial sampling for a large group of acutely unwell patients, whilst potentially reducing costs to the healthcare system.

    Table 1

    Analysis of PV-LACT cut-offs to predict A-LACT≥2

    The aim of this study is to determine if peripheral venous lactate measurement gives the same results as arterial lactate measurement in septic patients attending the Emergency Department (ED).

    Methods We performed a prospective observational cohort study of 304 consecutive consented patients presenting with sepsis to a single UK NHS ED (110,000 adult attendances annually) between October 2013 and March 2014. Arterial and peripheral venous samples were taken within 20 minutes of each other, and measured in mmol/L. Bland-Altman analysis was used to compare the level of agreement between PV-LACT and A-LACT. Receiver operating characteristic (ROC) curves were created to delineate optimal cut-off points to predict arterial lactate ≥2 mmol/L.

    Results Bland-Altman analysis noted a mean difference (PV-LACT-A-LACT) of 0.4 mmol/L, with 95% limits of agreement from −0.4 to 1.2 (figure 1). The majority of PV-LACT samples were higher than A-LACT samples, with a trend to a greater difference as lactate concentration increased. Only 7 A-LACT samples had a concentration ≥4 mmol/L. ROC analysis showed that we have a 100% ability to predict (i) A-LACT of <2 with a PV-LACT cut-off of 2.05, and (ii) A-LACT of ≥2 with a PV-LACT cut-off of 3.5 (table 1).

    Conclusion This study shows good clinical agreement between PV-LACT and A-LACT. PV-LACT measurements <2 mmol/L predict A-LACT <2 mmol/L with 100% accuracy. We recommend the routine use of peripheral venous lactate in the management of septic patients presenting to the ED. This will allow faster screening of this important group of patients in EDs (as well as other healthcare settings) where arterial lactate analysis or appropriately trained staff are not available. We can also improve patient care by reducing patient risk and discomfort.

    Figure 1

    Bland–Altman plot of agreement between peripheral venous (PV-LACT) and arterial (A-LACT) lactate.

    • emergency departments

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.