Helpful Only When Elevated: Initial Serum Lactate in Stable Emergency Department Patients with Sepsis Is Specific, but Not Sensitive for Future Deterioration

Abstract

Background

Early emergency department (ED) identification of septic patients at risk of deterioration is critical. Lactate is associated with 28-day mortality in admitted patients, but little evidence exists on its use in predicting short-term deterioration.

Objective

Our aim was to determine the role of initial serum lactate for prediction of short-term deterioration in stable ED patients with suspected sepsis.

Methods

We conducted a prospective cohort study of adult ED sepsis patients. Venous lactate was obtained within 2 h of ED arrival. Main outcome was subsequent deterioration (defined as any of the following: death, intensive care admission > 24 h, intubation, vasoactive medications for > 1 h, or noninvasive positive pressure ventilation for > 1 h) within 72 h. Patients meeting any endpoint within 1 h of arrival were excluded.

Results

Nine hundred and eighty-five patients were enrolled, of whom 84 (8.5%) met the primary outcome of deterioration. Initial lactate ≥ 4.0 mmol/L had a specificity of 97% (95% confidence interval [CI] 94–100%), but a sensitivity of 27% (95% CI 18–37%) for predicting deterioration, with positive and negative likelihood ratios of 10.7 (95% CI 6.3–18.3) and 0.8 (95% CI 0.7–0.9), respectively. A lower threshold of lactate (≥2.0 mmol/L) had a sensitivity of 67% (95% CI 55–76%) and specificity of 66% (95% CI 63–69%), with corresponding positive and negative likelihood ratios of 2.0 (95% CI 1.7–2.3) and 0.5 (95% CI 0.4–0.7).

Conclusions

High ED lactate is predictive of subsequent deterioration from sepsis within 72 h, and may be useful in determining disposition, but low lactate is not effective in screening stable patients at risk of deterioration.

Introduction

1. Why is this topic important?

1. Detection and risk stratification of emergency department (ED) patients with infection and sepsis is necessary to optimize treatment and disposition. While high lactate has been traditionally useful for identification of patients with evidence of shock at risk of 28-day mortality, it is unclear how well lactate risk stratifies patients who are initially stable, and whether low lactate levels are adequate to rule out future short-term deterioration.

2. What does this study attempt to show?

1. This prospective cohort study aimed to identify the role of initial serum lactate in the prediction of short-term deterioration among patients presenting to the ED and not requiring critical care interventions within the first hour. Both admitted and discharged patients were included.

3. What are the key findings?

1. While an initial lactate ≥ 4.0 mmol/L was highly predictive of short-term deterioration (specificity of 97%; positive likelihood ratio of 10.7), a lactate ≥ 2.0 mmol/L did not adequately capture patients at risk of deterioration (sensitivity of 67%; negative likelihood ratio of 0.5).

4. How is patient care affected?

1. Emergency physicians will often encounter septic patients who are hemodynamically stable, and must appropriately risk stratify and disposition them, based on their risk of short-term deterioration. While a lactate ≥ 4.0 mmol/L suitably identifies patients at high risk of short-term deterioration (and thereby necessitating disposition to a highly monitored setting), a lactate < 2.0 mmol/L does not reliably rule out short-term deterioration in these patients.

In the emergency department (ED), identification of patients with sepsis is critical, as it allows for the prompt initiation of appropriate resuscitation, including fluids and antibiotics (1). Once diagnosis is made and treatment is initiated, risk stratification (i.e., determining risk of future deterioration) of this patient population is difficult and ultimately paramount in determining the disposition of the patient (2). While many patients with severe sepsis or septic shock will require critical care in the intensive care unit (ICU), some patients are stable enough for management on the wards. However, patients who subsequently deteriorate on the wards and then require ICU admission have worse outcomes than those who are admitted directly to the ICU from the ED 3, 4, 5. It is therefore important to identify patients at risk of short-term deterioration at an early stage (6). Recent work in this area has focused on the utility of early warnings scores, and decision tools, such as the quick Sequential Organ Failure Assessment (qSOFA) (7). Unfortunately, the early data show that such tools have variable accuracy when tested in the ED setting, both prospectively and retrospectively, and include variables that are already suggestive of deterioration, such as hypotension 8, 9, 10, 11.

Serum biomarkers present an interesting avenue for investigation in risk stratification of ED patients with sepsis and, to date, > 180 different biomarkers have been linked with this disease process (12). Perhaps the biomarker of greatest interest to emergency physicians is serum lactate 13, 14. In the septic ED patient, lactate elevation occurs for several reasons, but the most common is inadequate tissue oxygenation (15). Most of the evidence surrounding the utilization of lactate in ED septic patients comes from patients admitted to hospital, where a lactate level ≥ 4.0 mmol/L is highly specific for in-hospital mortality (16). Hyperlactatemia is also a poor prognostic sign in these admitted patients, independent of evidence of organ hypoperfusion or shock, and regardless of the presumed etiology 14, 17, 18, 19. Furthermore, improved outcomes in patients with septic shock have been associated with clearance of lactate 20, 21. However, little is known regarding the use of lactate in risk stratification of ED patients not admitted to hospital, or about the ability of this biomarker to predict short-term deterioration (an important decision for the ED clinician in determining disposition). Furthermore, there are few published data on low lactate levels (i.e., < 2.0 mmol/L) and the prognostic value of a low lactate level in ruling out deterioration in ED patients with sepsis. Several studies have espoused the idea of cryptic shock, suggesting that patients who are hemodynamically stable, yet with evidence of hyperlactatemia, may in fact be in true shock, despite what would appear to be blood pressure capable of providing adequate organ perfusion 22, 23. Given this, we aimed to test the ability of initial ED serum lactate in predicting future deterioration among initially stable ED patients (defined as those not requiring endotracheal intubation, initiation of vasopressors or inotropes, or ICU admission, within the first hour following ED arrival). Additionally, with regard to patients with intermediate lactate levels (2.0–3.9 mmol/L), we sought to determine any differences between those who deteriorated and those who did not, as patients in this group are often the most difficult to prognosticate, and up to one-quarter may progress to septic shock (24).