The relationship between elevated prehospital point-of-care lactate measurements, intensive care unit admission, and mortality: A retrospective review of adult patients.

OBJECTIVE: To determine whether prehospital point-of-care lactate (pLA) is associated with mortality, admission, and duration of hospital stay. DESIGN: A retrospective clinical audit, where elevated lactate was defined as ≥2 mmol/L. SETTING: The ambulance service and primary referral hospital in the Australian Capital Territory from 1st July 2014 to 30th June 2015. PARTICIPANTS: Adult patients (≥18 years) who had pLA measured and were transported to the primary referral hospital. MAIN OUTCOME MEASURES: Mortality, admission, and duration of hospital stay. RESULTS: Two hundred fifty-three patients with a median pLA of 2.5 mmol/L (interquartile range [IQR]: 1.5-3.7) were analysed. Overall mortality was 8.3%; 68% were admitted to the hospital; 8.3% to the intensive care unit (ICU). pLA was non-significantly higher in those who died compared to survivors (3.5 [IQR: 2.75-5.85] vs 2.4 [1.5-3.6]; W = 1631.5; p = 0.053). pLA was higher for those admitted to the hospital (2.9 [1.9-3.9] vs 2.0 [1.4-3.1]; W = 5094.5, p = 0.001) and the ICU (3.2 [2.4-5.7] vs 2.4 [1.5-3.6]; W = 1578.5; p = 0.008). There was no relationship between pLA and duration of stay. Considered as a screening tool, at a cut-off of 2.5 mmol/L, pLA had a likelihood ratio+ of 1.61 for mortality and 1.44 for ICU admission; the odds ratio for mortality was 3.76 (95% confidence interval = 1.30, 13.89). CONCLUSIONS: Elevated prehospital lactate was associated with significantly increased ICU and hospital admissions. There may be value in pLA as a screening tool.

A 12-Month Clinical Audit Comparing Point-of-Care Lactate Measurements Tested by Paramedics with In-Hospital Serum Lactate Measurements.

OBJECTIVE: Prehospital point-of-care lactate (pLA) measurement may be a useful tool to assist paramedics with diagnosing a range of conditions, but only if it can be shown to be a reliable surrogate for serum lactate (sLA) measurement. The aim of this study was to determine whether pLA is a reliable predictor of sLA. METHODS: This was a retrospective study of adult patients over a 12-month period who had pLA measured by paramedics in an urban Australian setting and were transported by ambulance to a tertiary hospital where sLA was measured. Patients were excluded if they suffered a cardiopulmonary arrest at any time, had missing data, or if sLA was not measured within 24 hours of arrival. Levels of agreement were determined using methods proposed by Bland and Altman. RESULTS: A total of 290 patients were transported with a pLA recorded. After exclusions, there were 155 patients (55.0% male; age 71 [SD=18] years) remaining who had sLA recorded within 24 hours. Elevated pLA (>2.0mMol/L) was associated with sLA measurement (76.1% vs 23.9%; OR 3.18; 95% CI, 1.88-5.37; P<.0001). Median time between measurements was 89 minutes (IQR=75). Overall, median pLA was higher than sLA (3.0 [IQR=2.0] mMol/L vs 1.7 [IQR=1.3]; P<.001). Bland-Altman analysis on all participants showed a mean difference of 1.48 mMol/L (95% CI, -3.34 to 6.31). Normal pLA was found to be a true negative in 82.9% of cases, and elevated pLA was a true positive in 48.3% of cases. When the time between measurements was less than 60 minutes (n=25), normal pLA predicted normal sLA with 100% accuracy, with a false-positive rate of 18.2%. As time between measurements increased, accuracy diminished and the false-positive rate increased. CONCLUSIONS: Overall, the level of agreement between pLA and sLA was poor. Accuracy of pLA diminished markedly as the time between the two measurements increased. It may be possible to use pLA as a screening tool; when considered this way, pLA performed much better, though larger prospective trials would be needed to confirm this. Swan KL , Keene T , Avard BJ . A 12-month clinical audit comparing point-of-care lactate measurements tested by paramedics with in-hospital serum lactate measurements. Prehosp Disaster Med. 2018;33(1):36-42.