Sepsis is defined as a clinical syndrome characterised by the presence of both infection and a systemic inflammatory response. Sepsis may result in organ dysfunction, shock and death (Harrison et al, 2006; Wang et al, 2010). The definition of severe sepsis is sepsisinduced organ dysfunction or tissue hypoperfusion, and septic shock is defined by severe sepsis plus hypotension that persists despite fluid resuscitation (Levy et al, 2003; Dellinger et al, 2008).
Severe sepsis is common, and has an estimated incidence of one to three cases per 1 000 population per year, with a rate that has been increasing over two decades (Seymour et al, 2010). Severe sepsis accounts for 27% of admissions to intensive care units (ICU) in England, Wales and Northern Ireland. Patients with severe sepsis account for 46% of all ICU bed days, with a median hospital stay of four weeks. Mortality in patients with severe sepsis remains high at 30–50% (Harrison et al, 2006).
The Surviving Sepsis Campaign (SSC) was launched in 2002 and focused primarily on reducing mortality of severe sepsis by 25%. The SSC produced evidence-based guidelines for the management of severe sepsis that were published in 2004 and updated in 2008. The guidelines were produced in two care bundles: the resuscitation bundle, to be applied in the first six hours of diagnosis, and the management bundle, applicable within 24 hours of diagnosis (Robson et al, 2009).
The resuscitation bundle involves a number of simple therapeutic interventions that should be implemented within three hours of sepsis being recognised. The interventions encompass the measurement of serum lactate, taking blood cultures prior to antibiotics administration, broad spectrum antibiotics within one hour, and administration of 30 ml/kg crystalloid for hypotension or lactate equal to or greater than 4 mmol/L. More advanced resuscitative measures—early goal-directed therapy (EGDT)—include the application of vasopressors for ongoing hypotension unresponsive to fluid challenges, the maintenance of central venous pressure of 8 mm Hg or higher, and central venous oxygen saturation greater than 70% (SSC, 2012).
Non compliance with the six hour resuscitation bundle has been shown to be associated with a twofold increase in mortality, despite similarities in age and severity of sepsis (Gao et al, 2005). The SSC and critical care community have targeted clinical staff outside of critical care to implement a resuscitation bundle known as the ‘sepsis six’. The ‘sepsis six’ consists of (Robson et al, 2009):
The pre-hospital phase of sepsis care provides the earliest opportunity for sepsis identification and treatment. Pre-hospital treatment remains variable with emergency medical services (EMS) providers administering IV fluid in less than 40% of patients in septic shock (Seymour et al, 2010). Studnek et al (2012) note that there is very little literature discussing the role of EMS on the early care of sepsis; however, it is estimated that EMS provide care for approximately 50–60% of patients with severe sepsis (Wang et al, 2010; Studnek et al, 2012). Furthermore EMS provide initial care for the highest acuity patients with hemodynamic instability, severe sepsis and septic shock (Wang et al, 2010). Patients with sepsis identified by EMS staff receive a reduced time to EGDT in the ED (Halimi et al, 2011; Studnek et al, 2012). The identification of haemodynamic shock in sepsis is widely acknowledged as a vital step towards improving survival. In pre-hospital care this is particularly challenging (Pearse, 2009).
A proportion of patients in septic shock maintain normal blood pressure, but manifest global tissue hypoxia indicated by lactate levels higher than 4 mmol/L. This is often referred to as cryptic shock (Hanudel et al, 2008; Jansen et al, 2008; Puskarich et al, 2010). Patients in cryptic shock have been found to have an equal mortality to those patients with overt shock (Puskarich et al, 2010). Patients with elevated lactate levels experiencing sepsis may not display dramatic signs and symptoms (Goyal et al, 2010). Seymour et al (2010) observed hypotension in less than 25% of patients transported with severe sepsis. The measurement of lactate can be particularly useful in identifying septic patients with cryptic shock (Robson et al, 2009).
Within the spectrum of sepsis, lactate measurement is a tool commonly used to stratify patient illness severity. Lactate is a marker of anaerobic respiration, often elevated in the presence of tissue hypoxia. An elevated lactate level in the emergency department (ED) indicates an increased likelihood of mortality (Goyal et al, 2010). A trial by Rivers et al (2001) of EGDT in the ED for patients with severe sepsis and a lactate level of higher than 4 mmol/L or septic shock, revealed a 16% reduction in 28-day mortality. Following the trial by Rivers et al, the SSC and the Institute for Healthcare Improvement recommended early lactate measurement to help identify patients eligible for EGDT (Institute for Healthcare Improvement, 2012). A lactate level of greater than 2 mmol/L is one of the diagnostic criteria for organ dysfunction qualifying as severe sepsis. A lactate level of higher than 4 mmol/L persisting despite fluid resuscitation defines shock (Levy et al, 2003).
Research has shown point-of-care (POC) fingertip lactate testing of sepsis patients in the ED to be time saving, assists in identifying critical patients on arrival to the ED, and correlates well with whole blood lactate levels (Goyal et al, 2010; Gaieski et al, 2011). Practitioners in the ED often use elevated serum lactate as a measure of cryptic shock (Wang et al, 2010).
Pearse (2009) commented that pre-hospital lactate testing showed promise in identifying patients with haemodynamic shock and correlated with inhospital values as expected. The research by Jansen et al (2008) suggested a lactate value of 3.5 mmol/L as the optimal cutoff value for mortality prediction. The authors also concluded that lactate values offered an opportunity for early detection of cryptic shock and pre-hospital resuscitation. Pearse (2009) considered that lactate values would be used in a specific treatment algorithm, but identified that the accuracy of peripheral venous and capillary blood samples must be carefully considered. Van Beest et al (2009) also concluded that pre-hospital lactate testing was feasible and related to patient outcome. The authors suggested that further studies evaluate whether treatment of shock based on lactate values would improve outcomes.
The out-of-hospital phase of sepsis care is under researched (Seymour et al, 2010; Band et al, 2011). Additional research into non-invasive diagnostic modalities, such as pre-hospital temperature and lactate measurement to aid more timely and accurate diagnosis of sepsis by EMS personnel, has been recommended (Wang et al, 2010; Studnek et al, 2012).
Literature search aims
Research question
In patients presenting with sepsis, severe sepsis and septic shock, could lactate testing assist with pre-hospital identification?
Search strategy
A detailed search strategy is presented in Appendix 1. The search was conducted during April 2013 and identified four pieces of primary research.
Review of primary research
A summary of each piece of research is presented in Appendix 2.
There were four pieces of primary research identified that included lactate monitors to assist with identifying sepsis pre-hospital. Of these four papers, three were limited to a research forum abstract, and it has not been possible to obtain the full paper to date. The largest of these four studies included 219 patients, and of these, 183 were included in analysis.
Three pieces of research identified a strong correlation between pre-hospital lactate levels and ED lactate levels (Hokanson et al, 2012; Shiuh et al, 2012; Guerra et al, 2013). Hokanson et al (2012) identified that at higher lactate levels of >4 mmol/L, the correlation was lower at 0.33 (p=0.46). The authors of this study highlighted that four out of these seven patients had received fluid pre-hospitally.
Shiuh et al (2012) found that in utilising a pre-hospital sepsis protocol, which included POC lactate testing, pre-hospital clinicians correctly identified 76.7% of sepsis alert patients and 74.2% of sepsis advisory patients. These patients had a time to antibiotics of 59 minutes and 81 minutes respectively. There was an absence of a control group to compare intervention times, incidence of ICU admission and mortality. The authors concluded that the pre-hospital sepsis protocol was accurate, with potential for significant time savings in identification and treatment of sepsis patients and reductions in morbidity and mortality. Hokanson et al (2012) were in agreement, demonstrating that paramedics can use hand-held lactate POC devices to evaluate sepsis patients more effectively.
In the study by Hokanson et al (2012), only 9 of the admitted 39 study patients received a diagnosis of sepsis. Pre-hospital lactate values ranged from 1–9.8 mmol/L. There were significant advantages in time to lactate values when lactate was measured pre-hospitally. Median time intervals from pre-hospital lactate values to ED venipuncture were 67 minutes, and to actual ED measured lactate results, 116 minutes.
Guerra et al (2013) reported that clinicians trained in the sepsis screening tool and POC lactate monitors identified 47.8% of severe sepsis patients correctly. Interestingly, five of the unidentified patients' vital signs and venous lactate did not confirm hypoperfusion, but on arrival at ED, reassessment confirmed severe sepsis. In this study there were an additional eight unidentified patients with cryptic shock, but a late introduction of venous lactate monitors to the study meant these patients were not identified as septic by pre-hospital clinicians. Guerra et al (2013) recognised that the small sample size and late introduction of lactate monitors to the study limited the reliability of the findings.
Elevated lactate was found to be a predictor of ICU admission and mortality in the study by Hanudel et al (2008). Lactate levels were only available on 54 of 134 patients with sepsis. 31.5% of these patients had an elevated lactate level, with 20.37% maintaining a normal blood pressure, indicating the level of cryptic shock in the sample.
The research reviewed shows that pre-hospital lactate testing has been shown to reduce time to lactate values in septic patients, and has been successfully used pre-hospitally to identify sepsis patients when incorporated into a pre-hospital sepsis screening tool. Pre-hospital lactate testing has contributed to time savings in both the identification and treatment of patients presenting with sepsis, severe sepsis and septic shock. Lactate values have also been shown to correlate well with ED values.
The literature search has identified only four pieces of primary research all originating from the United States of America. Of this research, small sample sizes and variable availability of lactate testing has limited the strength of the research base in this area.
Proposals for future research
Conclusions
Sepsis is becoming increasingly common and has an extremely significant mortality rate. The pre-hospital phase of sepsis care is crucial as it is the first opportunity to identify sepsis and begin treatment.
The identification of haemodynamic shock in sepsis pre-hospital is challenging, especially where patients present with cryptic shock. Lactate testing is commonly used in other healthcare settings to identify sepsis and to stratify illness severity in sepsis. The literature search has shown some positive findings when lactate measurement has been utilised pre-hospital. The requirement for further research in this area within the UK ambulance service has been clearly identified. Research is required to further assess the benefit of pre-hospital lactate testing in the identification of septic patients presenting to the ambulance service.