a b s t r a c t

Objective: To determine the validity of plasma lactate in the emergency department for the early detection of tis- sue hypoperfusion in septic patients.

Materials and methods: Longitudinal descriptive study. Non probabilistic sampling for convenience. Plasma lac- tate levels were determined in patients admitted to the emergency department with systemic inflammatory re- sponse data and clinical suspicion or documented infection. Follow-up was seven days. Complications were considered if the patients presented septic shock, severe sepsis, entry to intensive care or death.

Results: Ninety patients were included. The mean age was 57.4 +- 20.31. Fifty five percent (n = 49) were women. 25% (n = 22) of the patients showed complications. Plasma lactate levels were 1.55 mmol/L in uncomplicated patients and 3.72 mmol/L for complicated patients (p b 0.001). The area under the ROC curve was 0.72 (95% CI, 0.575-0.829). The cutoff point that best described the relationship with the probability of complications was that set at 4.2 mmol/L. The variables studied that showed a significant association with the probability of complications were edema (p = 0.004), and infections of the respiratory tract (p = 0.037). A model that included lactate levels, using as adjustment variables edema and the presence of low respiratory tract infection explained between 0.234 and 0.349 of the dependent variant, correctly classifying 80% of the cases.

Conclusion: Plasma lactate is useful in emergency departments as a predictive test for the early detection of pa- tients with tissue hypoperfusion that evolve to severe sepsis, septic shock or death.

(C) 2017

Introduction

Sepsis is one of the major health problems. Up to 19 million cases per year around the world are estimated [1]. The rapidity with which the initial management is established is determinant in the morbi-mortality in the short, medium and long term [2].

In 2001, approximately 750,000 cases of severe sepsis were esti- mated annually in the United States; with a mortality rate of 28.6% [3, 4]. In Mexico, an annual incidence of 11,183 cases (27.3%) was esti- mated in 2009, with an estimated mortality of 30.4% [5].

The Consensus of the Surviving Sepsis Campaign of 2012, defined it as: the presence (probable or documented) of infection along with sys- temic manifestations of infection [2]. When the systemic response is perpetuated, hypotension, hypoperfusion and white organ damage may develop [6]. However, some patients report global tissue hypoxia,

* Corresponding author at: Av. De las Gondolas 416 Int. 10B, Valle del Alamo, Guadalajara, Jalisco, Mexico.

E-mail address: stat5a@hotmail.com (A. Marin-Medina).

evidenced by a lactate greater than or equal to 4 mmol/L in the presence of normotension, a condition referred to as cryptic shock [7]. The evi- dence supports that early intervention and diagnosis result in a signifi- cant reduction in morbidity and mortality [6].

Recently, attention has focused on lactate as a biomarker and even as a therapeutic target [8].

Under normal circumstances, tissues can be considered as lactate producers or metabolizers [9]. The arterial plasma concentration of lac- tate reflects the balance between production, and its consumption/ clearance. This concentration is generally b 2 mmol/L [10]. Most of the situations that lead to excess production and reduction in clearance are predominantly pathological, reflecting tissue hypoxia or non- hypoxic tissue injury [8].

However, although we found lactate levels N 4 mmol/L in septic patients, as a clear indication of tissue hypoperfusion; there is no clarity in the intervention that should be done in the patient with hyperlactatemia but without hemodynamic alteration [6].

According to the management guidelines, an elevated lactate is sufficient to diagnose shock, regardless of whether or not there is

https://doi.org/10.1016/j.ajem.2017.12.068

0735-6757/(C) 2017

hypotension. Sepsis with elevated levels of lactate (N 4 mmol/L) is asso- ciated with high mortality, and is an indication for initiating manage- ment protocols and packages. It has also been reported an inverse relationship between Lactate clearance and mortality in severe sepsis and septic shock [11].

Although a formal characterization of the possible progression from an infection without systemic manifestations to septic shock or death has not been achieved; traditionally, the natural process of sepsis has been understood as a linear sequence, covering different clinical stages [12].

The present study seeks to know if the measurement of plasma lactate can be used as a prognostic marker of complications in septic patients admitted to an emergency department.

Materials and method

This observational study was designed to determine whether the measurement of plasma lactate as a result of tissue hypoperfusion in septic patients can be used as a prognostic marker for complications in an emergency department. It was carried out in the period from May to September 2013 in the Emergency Service of the Regional General Hospital No. 46 of the Mexican Institute of Social Security in Guadala- jara, Jalisco. The study was previously authorized by the Hospital Re- search Committee (1306).

Inclusion criteria

Patients older than 18 years with at least two signs of systemic in- flammatory response (SIRS) according to the general and inflammatory variables proposed by the Surviving Sepsis Campaign: International Guidelines for Management of Severe Sepsis and Septic Shock: 2012, which are detailed in Table 1.

Criteria for non-inclusion

Patients presenting with chronic renal failure on renal replacement therapy, patients with hepatic failure at any stage, and patients on biguanide therapy.

Exclusion criteria

Patients who presented to the emergency department with a diag- nosis of severe sepsis or septic shock, patients who presented with some SIRS criteria, suspected infection or documented infection; Or a di- agnosis of sepsis, and which rapidly evolved to severe sepsis or septic shock before sampling for serum lactate and patients who met inclusion criteria and who were definitively transferred to another hospital to continue their study driving.

Table 1

Diagnostic criteria for sepsis.

Documented or suspected infection, and some of the following:

General variables

Fever (N 38.3 ?C)

Hypothermia b36 ?C)

Heart rate N 90 beats per min, or more than two SDs above normal for age. Tachypnea

Altered mental state

Significant edema or a positive fluid balance (N 20 mL/kg in 24 h) Hyperglycemia (plasma glucose N 140 mg/dL) in the absence of diabetes

Inflammatory variables

Leukocytosis (leukocyte count N 12,000) Leukopenia (leukocyte count b4000)

Normal leukocytes with N 10% immature forms

Abbreviations: SD, standard deviation; ?C, degrees centigrade; ml, milliliters; kg, kilo- grams; mg, milligrams; dl, deciliter.

Intervention

Patients were identified and classified in the first contact area of the emergency department, the clinical data were collected and the venous blood sample was collected for plasma lactate determination. The blood sample was analyzed using the VITROS(R) Chemistry Products LAC DT slides, to quantitatively determine the lactate concentration in plasma using the VITROS(R) 5600 Integrated System Biochemistry Analyzer, based on colorimetric spectrophotometry. Measuring range 0.5- 12 mmol/L (reference range 0.6-2.1 mmol/L).

Data collection and analysis

We identified patients who presented complications and those who did not. The independent variables studied were chosen from the set of signs, symptoms and laboratory data included as part of the systemic in- flammatory response and exposed in the last update of the Surviving Sepsis Campaign; as part of the selection criteria for these variables, the ease of collection in an emergency department was considered, as well as the feasibility and opportunity with which a laboratory study could be processed.

To determine the utility of lactate values, patients were followed up for 7 days. Possible complications were ICU admission, death, and pro- gression to severe sepsis or septic shock according to the medical diag- nosis at the time of the evaluation.

In order to establish the validity and usefulness of determination of lactate plasma levels, inferential statistics were used in the cases that merited (Student’s t, square chi, area under the ROC curve and bivariate logistic regression calculation). A significant p b 0.05 was considered.

Results

We included 98 patients who requested medical attention in the adult emergency department. All patients had at least two diagnostic criteria for systemic inflammatory response, more suspicion of infec- tion, or documented infection. Table 2 shows the distribution by sex, co- morbid, some clinical characteristics and basal lactate levels.

The usefulness of plasma lactate for the early detection of tissue hy- poperfusion in septic patients was assessed through the ability of this test to discriminate those patients likely to present complications (evo- lution to severe sepsis or septic shock, ICU admission or death for any reason) within seven days of your hospitalization.

Table 2

Clinical and demographic characteristics of patients

Variable Values

Age^a 59 (+-20.31)

Sex^b

Men 54 (49)

Women 46 (41)

Comorbidities^b

Systemic Arterial hypertension 41 (37)

Diabetes mellitus type 2 30 (27)

Clinical features^a

Systolic blood pressure (mm Hg) 114 (+-20.73)

Diastolic blood pressure (mm Hg) 70 (+-11.93)

Mean blood pressure (mm Hg) 86 (+-13.73)

Heart rate (bpm) 103 (+-16.39)

Respiratory rate (bpm) 23 (+-5.66)

Significant edema^b 27 (30)

Temperature (?C) 36 (+-0.94)

Oxygen saturation by pulse oximetry (%) 95 (+-5.98)

Glasgow Coma Scale (score) 15 (+-2.83) Laboratory studies^a

Plasma lactate (mmol/L) 1.4 (+-2.32)

Abbreviations: mmol/L, millimoles per liter; mmHg, millimeters of mercury; bpm, breaths per minute; bpm, beats per minute; ?C, degrees Celsius.

^a Median and standard deviation.

^b Percentage.

It was observed that 25% (n = 22) of the patients showed complica- tions. Lactate levels were significantly higher in patients with complica- tions with a mean of 3.72 +- 3.94 mmol/L than in the uncomplicated group with a mean of 1.65 +- 1.10 mmol/L. The mean difference be- tween complicated and uncomplicated patients in relation to lactate levels was – 2.06, 95% CI, -3.83 to – 0.29, respectively. The sample was statistically significant at t = 2,42, (p = 0.024). The model had a regular discrimination capacity ROC 0.72, 95% CI, 0.575-0.829 (Fig. 1). The cutoff point that best described the likelihood of complications was set at 4.2 mmol/L, as shown in Table 3.

Among the general SIRS variables, statistical significance was ob- served with edema (p = 0.004), and for Infection sites, only infection of the respiratory tract with pneumonia (p = 0.037). This is shown in Table 2.

Within the general inflammatory response variables, the heart rate N 90 beats per minute was the most common, followed by tachypnea and the presence of edema. In turn leukocytosis N 12,000 was the most frequent within the inflammatory variables. This is shown in Table 4.

A logistic regression model that included lactate levels was used, using as adjustment variables the presence of significant edema and the presence of Low Respiratory Tract Infection with Pneumonia, the model was significant, explaining between the 0.234-0.349 of the variant Dependent, and correctly classifying 80% of the cases. The lactate level was the most strongly associated with the prediction of complications.

Discussion

Since 1996, Bakker et al. [13] have studied plasma lactate levels as predictors of the development of multiple organ failure; concluding that not only the high initial plasma lactate but the duration of lactatemia have an important prognostic value in septic shock.

Other studies have shown that the initial increase in lactate has a di- rect relationship with mortality independently of systemic dysfunction [14]. And that patients with high lactate clearance have a higher survival rate than those with low clearance, showing that the sustained time of lactate clearance better predicts mortality [15]. The detection and erad- ication of global tissue hypoxia even before normalization of vital signs shows a beneficial impact in relation to mortality in severe sepsis and septic shock; and inadequate clearance during the early phase of resus- citation implies organic dysfunction and an increase in mortality [11].

In the present study, it was observed that 25% of the patients showed complications, with significantly higher lactate levels in the complicated

Fig. 1. ROC curve for lactate value (mmol/L) as predictor of complications in septic patients.

Table 3

Predictive ability of plasma lactate complications.

Values	Lactate >= 4.20 mmol/L	Lower limit	Upper limit
Prevalence of complications	24.44%	16.26%	34.84%
Patients correctly diagnosed	76.67%	66.35%	84.67%
Sensitivity	22.73%	8.69%	45.82%
Specificity	94.12%	84.86%	98.10%
Positive predictive value	55.56%	22.65%	84.66%
Negative predictive value	79.01%	68.26%	86.96%
Positive probability ratio	3.86	1.14	13.14
Negative probability ratio	0.82	0.65	1.04

Abbreviations: mmol/L, millimoles per liter.

patients than in the uncomplicated ones. The ROC curve set at 0.72 de- termines a plasma lactate value of 4.2 mmol/L as the cutoff point that best describes the likelihood of complications.

Recently, Londono et al. [17] have determined the usefulness of lac- tate as a prognostic factor in mortality and suggest that lactate value in patients admitted for emergencies with infection and without hypoten- sion is independently and significantly associated with 28-day mortal- ity, with an increase in the risk of death of 16% for each mmol/L of lactate above the initial values detected in blood.

Likewise, the results of Shapiro et al. [16] in patients with suspicion of infection in the emergency department show a higher mortality in patients with hyperlactatemia (28.4%) than those with normal lactate (4.0%), showing that the lactate measurement can be of great help to identify those patients at high risk of complications. Serum lactate greater than or equal to 4.0 mmol/L showed a sensitivity of 36% and specificity of 92% as predictor of death at three days; with an area under the curve for lactate levels as a predictor of death of 0.67, and sim- ilarly, Howell et al. [17] in their study on serum lactate as a predictor of mortality in patients with infection in the emergency department obtained an area under the ROC curve of 0.87 for lactate levels as a pre- dictor of in-hospital death.

In contrast to the previously mentioned studies, in our study, mor- tality was not only evaluated as a potential risk, but was included as one of the possible complications. In addition, unlike the findings of Sha- piro et al. [16] and Howell et al. [17] in our study, only patients without

Table 4 Clinical characteristics observed in septic patients in relation to the general and inflamma- tory variables of SIRS and identified infection sites.

Variable	Values
General variables of SIRSa
Heart rate N 90 bpm	89 (80)
Tachypnea	72 (65)
Significant edema	30 (27)
Altered mental state	26 (23)
Fever N 38.3 ?C	16 (14)
Hypothermia b36 ?C	6 (5)
Hyperglycemia N 200 mg/dl in the absence of diabetes mellitus	6 (5)
Inflammatory variables of SIRSa

Leukocytosis N 12,000	67 (60)
Leukopenia b4000	2 (2)
Normal leucocytes with N 10% immature forms. Site of documented or suspected infectiona Infection of the respiratory tract with pneumonia	1 (1) 48.8 (44)
Urinary tract infection	45.5 (41)
Skin and/or soft tissue infection	22.2 (20)
Intra-Abdominal infection	20.0 (18)
Infection of the gastrointestinal tract	11.1 (10)
Surgical Wound infection	5.5 (5)
Neuroinfection	3.3 (3)
Bone and/or joint infection	2.2 (2)
Genital tract infection	1.1 (1)
Infection of the bloodstream associated with a catheter	1.1 (1)

Abbreviations: SIRS, systemic inflammatory response syndrome; mg/dl, milligrams per deciliter; ?C, degrees centigrade; bpm, beats per minute; mmol/L, millimoles per liter.

^a Median and standard deviation.

hypotension were considered; the purpose of this study was to evaluate the capacity of the test for the early detection of patients susceptible to complications in order to perform a Timely intervention and, conse- quently, to impact their short, medium and long term prognosis in both morbidity and mortality.

The accuracy index that shows the ROC curve of our patients, established at 0.72, classifies the accuracy of the test as regular in its ability to discriminate individuals with a plasma lactate >= 4.2 mmol/L and with probability of complications. For the cut-off point stipulated in 4.2 mmol/L, plasma lactate shows a non-significant sensitivity, how- ever, its predictive value lies in its high specificity, qualifying it with a high potential to detect those patients with a diagnostic test b 4.2 mmol/L and low probability of complications. According to speci- ficity, the negative predictive value is representative, which also indi- cates a high effectiveness in distinguishing the group of patients with a plasma lactate b 4.2 mmol/L and a low risk of complications. The odds ratios indicate an expectation of complication almost four times greater in those patients with a lactate >= 4.2 mmol/L (positive), in view of a low complication perspective in patients with a determination b 4.2 mmol/L. These results are similar to those described by Shapiro et al. [16] who describe the association of plasma lactate levels and the risk of in-hospital death in patients with suspected infection, although un- like this, in our study, mortality was considered as part of possible com- plications, and extended the time span for three to seven days follow- up. Likewise, in the study conducted by Londono et al. [18] the useful- ness of lactate was determined as a predictive factor of mortality at 28 days, establishing an Odds Ratio of 4.08 in those patients with plasma lactate greater than or equal to 4 mmol/L.

When performing the data analysis, no statistical significance was

observed between complicated and uncomplicated patients with re- spect to the pathological antecedents, nor in the inflammatory variables of SIRS; However, within the general SIRS variables, statistical signifi- cance was observed with edema (p = 0.004). Regarding infection sites, only infection of the respiratory tract with pneumonia (p = 0.037) showed statistical significance. These results contrast with that reported by Puskarich et al. [7] who report a significant correlation be- tween the presence of tachycardia and tachypnea with 28-day mortal- ity, with plasma lactate N 4 mmol/L, the relationship described is also applicable for other variables studied Is the case of the presence of Sys- temic Arterial Hypertension and type 2 diabetes mellitus. Although it should be noted that in our study, uncomplicated patients were taken into account, unlike the study by Puskarich et al. [7] in which patients were not discriminated between those who presented hypotension; in addition to only the physiological variables related to SIRS were studied. Regarding the sites of infection detected in this study, the prevalence found coincides with that reported by Abad et al. [19] who showed pleuropulmonary infections as the main site of infection, followed by intra-abdominal and urinary tract infections. It also coincides with the findings of Rezende et al. [20] who report a higher incidence in lung in- volvement (55%), urinary tract (22.2%), intrabdominal infections (11%), others (8.3%) and unknown site (3.5%). In Mexico, there are no studies that identify the most common sites of infection in septic patients in the emergency department; however, Carrillo-Esper et al. [5] identified in Mexican Intensive Care Units that intra-abdominal infections were the most prevalent (47%), followed by pulmonary (33%), soft tissue (8%), renal 7%) and miscellaneous (5%), which differs with our findings,

and can be attributed to the provenance of the individuals studied.

The incidence of plasma lactate levels >= 4.0 mmol/L was 10%, higher than reported by Howell et al. [17] who studied the relationship of these levels solely with mortality. Comparing patients with plasma lactate

>= 4.2 mmol/L with those who presented complications through the chi-square test, statistical significance was observed, with p = 0.022, being able to present complications such as severe sepsis, ICU admission or death from any cause within seven days of hospitalization; this ob- servation coupled with those performed by Puskarich et al. [7] who found no difference in mortality among patients recruited with septic

shock manifested in relation to those with cryptic shock (lactate N 4 mmol/L) revealed the need to detect this group of septic patients with occult hypoperfusion manifested by elevated plasma lactate levels. According to the Surviving Sepsis Campaign: International Guide- lines for Management of Severe Sepsis and Septic Shock: 2012, the de- termination of serum lactate in patients with severe sepsis or with septic shock is proposed, not only as a diagnostic test, but even as a ther- apeutic guide; however, its use has not been standardized in septic pa- tients presenting to the emergency department without clinical data of tissue hypoperfusion or overt septic shock. Our study proposes the de- termination of plasma lactate levels as a prognostic marker of complica- tions in septic patients admitted to an emergency department even without clinical data of tissue hypoperfusion or overt shock, who could benefit from the early Resuscitation protocol by decreasing their

morbidity and mortality.

Therefore, lactate acquires great utility in the screening of septic pa- tients who enter the emergency department. Elevated plasma lactate may alert physicians to the presence of subclinical tissue hypoperfusion and/or incipient complication.

Conclusion

In hospitalized patients presenting to an Emergency Service with SIRS and clinical suspicion of infection or documented infection, a single venous lactate measurement provides substantial prognostic informa- tion regarding the probability of complications at 7 days. Our results support serum lactate as a useful risk stratification tool for decision making in emergency departments.

Conflict of interest

No conflict of interest.

Funding

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors.

References

1. Adhikari NK, Fowler RA, Bhagwanjee S, Rubenfeld GD. Critical care and the global burden of critical illness in adults. Lancet 2010 Oct 16;376(9749):1339-46.
2. Dellinger NP, et al. Surviving sepsis campaign: international guidelines for manage- ment of severe sepsis and septic shock: 2012. Crit Care Med 2013 Feb;41(2): 580-637.
3. Raghavan M, Marik PE. Management of sepsis during the early “olden hours”. J Emerg Med 2006 Aug;31(2):185-99.
4. Napoli AM, Machan JT, Forcada A, Corl K, Gardiner F. Tissue oxygenation does not predict central venous oxygenation in emergency department patients with severe sepsis and septic shock. Acad Emerg Med 2010 Apr;17(4):349-52.
5. Carrillo-Esper R, Carrillo-Cordova JR, Carrillo-Cordova LD. Epidemiological study of sepsis in Mexican intensive care units. Cir Cir 2009 Jul-Aug;77(4):301-8 [279-85].
6. Zapata ML, Jaimes F. Fisiopatologia, importancia y utilidad del lactato en pacientes con sepsis. Iatreia 2010 Sep;23(3):278-85.
7. Puskarich MA, Trzeciak S, Shapiro NI, Heffner AC, Kline JA, Jones AE. Outcomes of pa- tients undergoing early sepsis resuscitation for cryptic shock compared with overt shock. Resuscitation 2011 Oct;82(10):1289-93.
8. Okorie ON, Dellinger P. Lactate: biomarker and potential therapeutic target. Crit Care

   Clin 2011 Apr;27(2):299-326.

   Handy J. Lactate — the bad boy of metabolism, or simply misunderstood? Curr Anaesth Crit Care 2006;17(1-2):71-6.

9. Phypers B, Pierce JT. Lactate physiology in health and disease. Contin Educ Anaesth Crit Care Pain 2006 Jun;6(3):128-32.
10. Nguyen HB, Loomba M, Yang JJ, Jacobsen G, Shah K, Otero RM, et al. Early lactate clearance is associated with biomarkers of inflammation, coagulation, apoptosis, organ dysfunction and mortality in severe sepsis and septic shock. J Inflamm (Lond) 2010 Jan 28;7:6.
11. Leon AL, et al. Clinical course of sepsis, severe sepsis, and septic shock in a cohort of infected patients from ten Colombian hospitals. BMC Infect Dis 2013 Jul 24;13:345.
12. Bakker J, Gris P, Coffernils M, Kahn RJ, Vincent JL. Serial Blood lactate levels can pre- dict the development of multiple organ failure following septic shock. Am J Surg 1996 Feb;171(2):221-6.
13. Mikkelsen ME, Miltiades AN, Gaieski DF, Goyal M, Fuchs BD, Shah CV, et al. Serum lactate is associated with mortality in severe sepsis independent of organ failure and shock. Crit Care Med 2009 May;37(5):1670-7.
14. Porras-Garcia W, Ige-Afuso M, Ormea-Villavicencio A. Depuracion de lactato como indicador pronostico de mortalidad en pacientes con sepsis severa y choque septico. Rev Soc Peru. Med Interna 2007;20(4):132-8.
15. Shapiro NI, Howell MD, Talmor D, Nathanson LA, Lisbon A, Wolfe RE, et al. Serum lactate as a predictor of mortality in emergency department patients with infection. Ann Emerg Med 2005 May;45(5):524-8.
16. Howell MD, Donnino M, Clardy P, Talmor D, Shapiro NI. Occult hypoperfusion and mortality in patients with suspected infection. Intensive Care Med 2007 Nov;33 (11):1892-9.
17. Londono J, Leon AL, Rodriguez F, Barrera L, de la Rosa L, Dennis R, et al. Lactato serico en urgencias como factor pronostico en pacientes con sepsis sin hipotension. Med Clin (Barc) 2013 Sep;141(6):246-51.
18. Abad CL, Kumar A, Safdar N. Antimicrobial therapy of sepsis and septic shock – when are two drugs better than one? Crit Care Clin 2011 Apr;27(2):e1-27.
19. Rezende E, Silva Jr JM, Isola AM, Campos EV, Amendola CP, Almeida SL. Epidemiol- ogy of severe sepsis in the emergency department and difficulties in the initial assis- tance. Clinics (Sao Paulo) 2008 Aug;63(4):457-64.