a b s t r a c t

Objective: This study describes emergency department (ED) sepsis patients with non-critical serum venous Lactate levels (LAC b 4.0 mmol/L) who suffered in-hospital mortality and examines LAC in relation to sur- vival times.

Methods: An ED based retrospective cohort study accrued September 2010 to August 2014. Inclusion criteria were ED admission, LAC sampling, N 2 systemic inflammatory response syndrome criteria with an infectious source (sepsis), and in-hospital mortality. Kaplan-Meier curves were used for survival estimates. An a priori sub-group analysis for patients with repeat LAC within 6 hours of initial sampling was undertaken. The primary outcome was time to in-hospital death evaluated using rank-sum tests and regression models.

Results: One hundred ninety-seven patients met inclusion criteria. pulmonary infections were the most common (44%) and median LAC was 1.9 mmol/L (1.5, 2.5). Thirteen patients (7%) died within 24 hours and 79% by <=28 days. Median survival was 11 days (95% CI, 8.0-13). Sixty-two patients had repeat LAC sampling with 14 (23%) and 48 (77%) having decreasing increasing levels, respectively. No significant differences were observed in treat- ment requirements between the LAC subgroups. Among patients with decreasing LAC, median survival was 24 days (95% CI, 5-32). For patients with increasing LAC median survival was significantly shorter (7 days; 95% CI, 4-11, P = .04). Patients with increasing LAC had a non-significant trend toward reduced survival (HR = 1.6 95% CI, 0.90-3.0, P = .10).

Conclusions: In septic ED patients experiencing in-hospital death, non-critical serum venous lactate may be uti- lized as a risk-stratifying tool for early mortality, while increasing LAC levels may identify those in danger of more rapid deterioration.

(C) 2015

Introduction

Sepsis, defined as a Systemic Inflammatory Response Syndrome in the setting of infection, represents a wide spectrum of pathol- ogy and is a common emergency department (ED) diagnosis [1,2]. Se- vere sepsis is defined by the presence of acute organ dysfunction secondary to infection, while the sub-classification of septic shock oc- curs with concomitant hypotension not reversed by fluid resuscitation therapies. In the United States more than 750,000 patients with severe sepsis are treated in EDs annually [2]. Contemporary trials demonstrate

Abbreviations: LAC, serum venous lactate; NCLAC, non-critical serum venous lactate; SIRS, systemic inflammatory response syndrome; WBC, white blood cell.

? Financial disclosure: No financial support was provided for this study.

?? Conflict of interest statement: None of the authors has any conflicts of interest. All au- thors had full access to all study data and had final responsibility for the decision to submit

for publication.

* Corresponding author at: SUNY Downstate Medical Center, Department of Emergency Medicine, 450 Clarkson Ave, Brooklyn, NY 11203, USA. Tel.: +1 718 245 4790; fax: +1 718 245 4799.

E-mail address: [email protected] (A.R. Aluisio).

a mortality rate of approximately 20% to 30% among those with septic shock [3-5]. ED patients admitted to hospitals with uncomplicated or non-severe sepsis (having no clinical evidence of shock or acute organ dysfunction) compose a less well-studied population [6]. This clinical sub-group however accounts for a large proportion of the sepsis- related morbidity and mortality per annum [7-10]. Identification of pa- tients with non-severe sepsis who will have poor outcomes is difficult as there is often a lack of early clinical indicators and hemodynamic insta- bility may not manifest until hours or days after the initial ED evaluation [11-13]. Research has shown that ED patients with uncomplicated sep- sis have a one in five chance of developing severe sepsis or septic shock within 72 hours of presentation, and with this Disease progression there is an associated increased risk of mortality [11].

Serum venous lactate (LAC) is a physiologic marker that has prognostic value for mortality in ED patients with sepsis [13,14]. LAC N 4 mmol/L is the most commonly used clinical parameter in defining severe sepsis and levels above this critical value are associated with in- creased mortality [2,14]. The utility of LAC in relation to morbidity and mortality outcomes among patients with non-severe sepsis is less well defined. In an ED based prospective study enrolling nonelderly patients

http://dx.doi.org/10.1016/j.ajem.2015.10.006

0735-6757/(C) 2015

A.R. Aluisio et al. / American Journal of Emergency Medicine 34 (2016) 170-173 171

with stable hemodynamics and elevated LAC N 2 mmol/L was associated with increased need for pharmacologic vasopressor support and me- chanical ventilation by 72 hours of hospitalization [15]. A systematic re- view examining patients with intermediate lactate levels (2.0-3.9 mmol/L) found a 28-day mortality rate of 15% for ED patients with un- complicated sepsis and normotension [14]. Furthermore, a retrospec- tive study of patients admitted with sepsis and initial LAC levels in the intermediate range found that those who had improvement in their LAC on repeat sampling within 12 hours had a 30-day mortality rate of 13.3% as compared to 24.7% among patients who failed to have im- provement [16]. Although guidelines do not recommend repeat sam- pling in non-severe sepsis these findings suggest that serial LAC sampling may provide clinically important information in this sub- population. While these studies show a significant mortality risk for pa- tients with non-critical serum venous lactate (NCLAC) levels and that LAC clearance may be associated with reduced mortality, they fail to de- fine the utility of LAC levels in prognosticating survival time in such high-risk cohorts. This study aims to describe the characteristics of ED patients with NCLAC levels and sepsis who died during hospitalization and evaluate the role of initial, and trends in LAC in relation to in- hospital survival.

Methods

Study setting and design

This retrospective cohort study was carried out at an urban teaching hospital. The ED has an annual census of approximately 140000 visits. Documentation in the hospital is done using an electronic medical re- cord (EMR) that allows for comprehensive review of all health informa- tion. Institutional review board approval was obtained before the study.

Inclusions and exclusion criteria and data acquisition

ED patients meeting the following criteria were included in the study: (1) admission through the ED, (2) N 2 SIRS criteria at time of triage and a documented infectious source (sepsis) [2], (3) NCLAC on initial ED serologic evaluation, and (4) death during the index hospital- ization. The study reviewed all hospital admissions from the ED that occurred between 1 September 2010 (the initiation date of the current EMR system) and 31 August 2014. The hospital’s EMR (Quadramed; Quadramed Corporation, Reston, VA) was queried for patients that died during the defined time period. After obtaining the list of all deaths, the medical records were screened for patients meeting all other inclu- sion criteria. Trained research personnel using a structured data acquisi- tion protocol extracted data from the EMR, which were inputted into a secure database with ten percent double entered and showing an error rate of less than 5%. Information pertaining to patient demographics, past medical history, the index presentation, ED vital signs, initial labo- ratory results, treatments, cause of death, date and time of ED arrival, and date and time of death were obtained.

Outcome definition

The primary outcome of in-hospital survival time was derived from the difference in time from ED presentation (ED EMR time-stamped ar- rival time) to time of in-hospital death. Time of death was based on the time that was recorded in the EMR disposition note. Time variables were extracted to include the hour and minute of the time point of in- terest. For subjects dying in less than 24 hours fractional days were cal- culated based on survival hours and minutes.

Statistical methods

Data were explored using descriptive statistics with frequencies and corresponding percentages for categorical variables and medians with

interquartile ranges (25% and 75% quartiles) or 95% confidence intervals (95% CI) for continuous variables. Statistical associations between com- parison variables were assessed using Mann-Whitney U or Wilcoxon rank-sum tests for continuous variables and Pearson ?² or Fisher exact tests for categorical variables. Kaplan-Meier curves were used to define survival estimates with median values and corresponding 95% CI calculated. Based on prior research showing potential prognostic util- ity of LAC trends among septic patients, an a priori subgroup analysis of patients with repeat LAC within 6 hours of ED presentation was under- taken [16-18]. The subgroups were stratified by increasing versus de- creasing repeat LAC levels. The differences in median survival time were assessed using Wilcoxon rank-sum test and Cox proportional haz- ards models. All analysis was performed using STATA version 11.0 (College Station, TX).

Results

Cohort characteristics

During the study period a total of 440 ED patients with NCLAC levels died during index hospitalizations. Among these, 197 patients (45%) met inclusion criteria and were analyzed. The majority of patients (79%) died <= 28 days of initial presentation. Thirteen patients (7%) died in <= 24 hours from time of ED presentation (Fig. 1).

The median age for the cohort was 67 years (quartiles: 55, 79) with a range from 19 to 99 years. There were slightly more male than female subjects (54% vs 46%). The majority of patients (93%) had documented pre-existing comorbidities. The most frequently observed SIRS parame- ters were elevated heart rate (84%), elevated band count (72%) and ab- normal White Blood Cell count (67%). An initial systolic blood pressure

<= 90 mmHg was found in 13% of subjects at triage, all of which improved to N 90 mmHg with ED therapies. The median initial ED LAC level was

1.9 mmol/L (1.5, 2.5). In the cohort ED laboratory evaluations found that electrolytes and Creatinine levels were within normal ranges and that the median hemoglobin level was 10 g/dL. The most common infec- tious sites were pulmonary (44%) and genitourinary (19%). Nearly all patients (99%) had antibiotics administered during ED treatment. Ap- proximately half of the cohort necessitated treatment with endotrache- al intubation or pharmacologic vasopressor support and 44% required intensive care unit admission (Table 1).

Fig. 1. Study flow diagram.

172 A.R. Aluisio et al. / American Journal of Emergency Medicine 34 (2016) 170-173

Table 1

Baseline characteristics of the enrolled cohort

Characteristic n (%)/median (25%, 75%)

Age (y) 67 (55, 79)

Sex

Male 107 (54%)

Female 90 (46%)

Documented prior comorbidities 183 (93%)

Initial systolic blood pressure b 90 (mmHg) 26 (13%)

Initial heart rate N 90 (beats/min) 166 (84%) Initial temperature N 38 or b36 (degrees Celsius) 62 (31%) Initial respiratory rate N 20 (breaths/min) 108 (54%)

WBC count N 12,000 or b 4,000 (x10^{^9}/L) 132 (67%)

Bands N 10 (%) 142 (72%)

LAC? (mmol/L) 1.9 (1.5, 2.5)

Hemoglobin (g/dL) 10.2 (8.5, 12.2)

Platelet count (uL^-1) 241,000 (141,000, 365,000)

International normalized ratio 1.3 (1.0, 1.6)

Serum creatinine (mg/dL) 1.4 (0.9, 2.4)

Serum HCO₃ (mmol/L) 22 (17, 25)

Serum glucose (mg/dL) 124 (97, 196) Infectious source

Pulmonary 87 (44%)

Genitourinary 37 (19%)

Intra-abdominal 20 (10%)

Occult 20 (10%)

Skin and soft tissue 16 (8%)

Blood 14 (7%)

Central nervous system 2 (1%)

Osteomyelitis 1 (0.5%)

Antibiotics administered in ED 194 (99%)

Required endotracheal intubation 115 (58%)

Required vasopressor support 98 (50%)

Admitted to intensive care unit 86 (44%)

Primary cause of death infection/sepsis 139 (71%)

* WBC, white blood cell.

Survival time

Among patients that died within 24 hours of ED presentation, the median age was 77 years (65, 82). Approximately three quarters of pa- tients (77%) with death by <= 24 hours had prior documented comorbid- ities. The median initial ED LAC level in this sub-population was 2.2 mmol/L (1.8-2.5). Similar to the overall cohort, the most common infec- tious sites were pulmonary (46%) and genitourinary (23%).

Total time at risk among patients meeting inclusion criteria was 3553 hospital days. The Kaplan-Meier survival curve for the overall co- hort is illustrated in Fig. 2. The median survival time is demarcated by the vertical hashed line and corresponds to 11 days (95% CI, 8-12). The in-hospital survival time ranged from 5 hours to 155 days.

Fig. 2. Kaplan-Meier curve representing in-hospital survival time of the cohort. ?Hashed line demarcates the median survival time among the overall cohort of 11 days (95% CI 8.0-13).

Characteristics and survival time based on repeat LAC trends

There were 62 patients that had repeat LAC levels within 6 hours of their initial ED LAC testing. Among this group, 14 (23%) patients had de- creasing LAC levels and 48 (77%) had increasing levels. For patients with decreasing LAC trends, the median initial levels were significantly higher than those of patients with increasing LAC at 2.6 versus 1.9 mmol/L, respectively (P = .02). Median in-hospital survival among pa- tients with a decrease in repeat LAC levels was 24 days (95% CI, 5-32). For patients with increasing LAC levels the median survival time was significantly reduced at 7 days (95% CI, 4-11) as compared to patients with down trending levels (P = .04). No significant differences were ob- served in need for treatment with endotracheal intubation, pharmaco- logic vasopressor support or required ICU admission between groups with divergent repeat LAC levels (Table 2). Patients with increasing LAC levels showed a non-significant trend toward reduced survival time and earlier in-hospital mortality risk as compared to those with decreasing levels (HR, 1.6; 95% CI, 0.9-3.0; P = .10).

Discussion

This study demonstrates that ED patients with sepsis and NCLAC with in-hospital mortality had a median survival time of 11 days. Deaths within the first 24 hours of presentation were uncommon. These find- ings indicate that NCLAC levels may serve as a useful prognostic marker against early in-hospital death in similar ED populations. In addition, a trend toward reduced in-hospital survival time was observed with in- creasing LAC levels repeated within six hours suggesting that serial sampling may aid in further risk stratification among ED patients with sepsis and initial NCLAC results.

This cohort study is one of the largest to examine ED patients admit- ted without severe sepsis or septic shock with an outcome of in-hospital mortality. The protracted median survival time of 11 days in this popu- lation has not been reported previously. This is an important finding as it shows that early mortality is rare among non-severe septic ED pa- tients with NCLAC levels. The characteristics of this cohort are similar to prior studies of patients with sepsis and NCLAC levels in relation to demographics, underlying infectious etiologies and presenting LAC levels [19-21]. A previous study of ED patients with NCLAC levels re- ported 3-day mortality rates ranging from 4% to 12%, which are similar to the rate of early death (24-hour survival) observed in this cohort [22]. These congruencies support validity in the representativeness of the co- hort and the potential usefulness of NCLAC as a prognostic marker against early in-hospital death in similar ED patients.

The role of serial LAC in patients with severe sepsis for risk stratifica- tion and evaluation of Treatment response is well documented and rec- ommended by international guidelines [2,23]. There is, however, limited evidence for serial assessments of LAC levels among patients non-severe sepsis as a parameter to predict clinical trajectories. A retro- spective cohort study of patients with sepsis and intermediate lactate levels demonstrated a ten percent absolute increase in 30-day mortality for patients who failed to clear their LAC at 12 hour repeat testing [16]. This prognostic finding in LAC trends has also been reported in other

Table 2

Characteristics of patients with 6-hour repeat serum venous lactate testing

Characteristic	Decreasing LAC	Increasing LAC	P
	(n = 14)	(n = 48)
Initial ED LAC (mmol/L)*	2.6 (2.2, 3.1)	1.9 (1.4, 2.5)	.02
Change in LAC (%)	-41 (-47, -30)	157 (41, 269)	b.001
Survival time (days)*	24 (5-32)?	7 (4-11)?	.04
Died b 24 hours	1 (7%)	3 (6%)	.65
Required endotracheal intubation	12 (86%)	29 (60%)	.07
Required vasopressor support	9 (64%)	27 (58%)	.45
Admitted to intensive care unit	9 (64%)	25 (52%)	.31

A.R. Aluisio et al. / American Journal of Emergency Medicine 34 (2016) 170-173 173

smaller studies [14,24]. In the patients studied in the reported on popu- lation, up trending LAC levels at six hours were found to be associated with a significant 3-fold reduction in survival time with non- parametric testing. In regression analysis a non-significant trend to- wards reduced survival was found for patients with up trending as com- pared down trending LAC levels. This lack of statistical significance in the observed survival difference likely stems for insufficient power owing to the small number of patients who had serial LAC sampling available for analysis. Even though the Cox regression results relating to Lactate clearance were non-significant they are concordant with the existing body of literature [14,16,24] and when viewed in concert sug- gest that serial LAC sampling in septic patients with initial NCLAC levels may have utility as a prognostic parameter in the acute setting. In ED patients with severe sepsis and septic shock serial LAC monitoring has been associated as an independent predictor for reduction in the likeli- hood of death however the role of serial sampling in non-severe sepsis is not as clearly defined [23]. Prospective studies assessing the prognos- tic utility of repeat sampling in sepsis patients with initial NCLAC levels are needed to better evaluate clinical benefits and effects on patient outcomes.

Meta-analyzed data estimate mortality rates for ED patients with sepsis and NCLAC levels at approximately 15%. This rate represents a substantial burden of death in comparison to other common presenting Disease states [25]. Risk stratification of this frequent ED presentation for appropriate disposition and treatment is crucial to effectively combat morbidity and mortality in this subpopulation [8,9]. The findings of this study add to the knowledgebase for that risk stratification however, given the methods these results are primarily hypothesis generating and additional research is needed to better explore and validate the results.

Limitations

This study must be interpreted in the context of certain limitations. Given the retrospective design, the results are subject to selection and misclassification bias. The investigators used a systematic data extrac- tion protocol in order to attempt to minimize such biases. In addition, during the period of accruement there was a trial concurrently enrolling patients with septic shock in the same practice setting, which likely en- hanced practitioner awareness of septic patients and should have served to reduce misclassifications [3]. This cohort was derived from a single center, which may reduce the generalizability of the findings. However, given the similarities between the population reported, and those from prior studies, this is less likely [19-21].

The studied cohort used clinical findings in conjunction with NCLAC as the serological screening marker for non-severe sepsis. Although LAC is the most frequently used ED biomarker to assess for organ hypoper- fusion there are additional laboratory tests for septic classification which can be used, and as such patients with alternative serological pa- rameters defining severe sepsis may have been included in the popula- tion [2]. Analysis of serologic trends in the cohort did not indicate that this occurred however inclusion is possible and must be taken into ac- count with the results. The available study data did not allow for tempo- ral analysis of factors relating to In-hospital treatment that may have confounded the results and affected survival time and this is a major limitation. However, in ED settings providers must make decisions without knowing the subsequent clinical course and observational data such as that presented here represent realistic clinical practice and provides pragmatic information applicable to patient care [26].

Conclusions

This study demonstrates that septic patients with NCLAC who died during their hospitalization had protracted survival times. Early death (within 24 hours) was not common in this population. These findings support the utility of LAC measurement in the ED setting as a risk strat- ification tool against early in-hospital mortality. In addition, the data in

conjunction with prior research suggests that patients with initial NCLAC that were increasing within 6-hour repeat sampling had re- duced in-hospital survival time as compared to those with decreasing levels. This finding puts forth a potential role for repeat LAC sampling in patients with initial NCLAC for survival prognostication and warrants further prospective investigation in this high-risk and common ED patient population.

References

1. Rangel-Frausto MS, Pittet D, Costigan M, Hwang T, Davis CS, Wenzel RP. The natural history of the Systemic Inflammatory Response Syndrome . A prospective study. JAMA 1995;273(2):117-23.
2. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and sep- tic shock: 2012. Crit Care Med 2013;41(2):580-637.
3. Pro CI, Yealy DM, Kellum JA, Huang DT, Barnato AE, Weissfeld LA, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med 2014;370(18):1683-93.
4. Gaieski DF, Edwards JM, Kallan MJ, Carr BG. Benchmarking the incidence and mor- tality of severe sepsis in the United States. Crit Care Med 2013;41(5):1167-74.
5. Mouncey PR, Osborn TM, Power GS, Harrison DA, Sadique MZ, Grieve RD, et al. Trial of early, goal-directed resuscitation for septic shock. N Engl J Med 2015;372(14):1301-11.
6. Ter Avest E, de Jong M, Brummer I, Wietasch GJ, Ter Maaten JC. outcome predictors of uncomplicated sepsis. Int J Emerg Med 2013;6(1):9.
7. Dombrovskiy VY, Martin AA, Sunderram J, Paz HL. Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: a trend analysis from 1993 to 2003. Crit Care Med 2007;35(5):1244-50.
8. Alberti C, Brun-Buisson C, Burchardi H, Martin C, Goodman S, Artigas A, et al. Epide- miology of sepsis and infection in ICU patients from an international multicentre co- hort study. Intensive Care Med 2002;28(2):108-21.
9. McCaig L, Nawar E. Advance Data from Vital and Health Statistics. National Ambula- tory Medical Care Survey: 2005 Emergency Department Summary. Washington DC: US Department of Health and Human Services; 2006 372.
10. Whippy A, Skeath M, Crawford B, Adams C, Marelich G, Alamshahi M, et al. Kaiser Permanente’s performance improvement system, part 3: multisite improvements in care for patients with sepsis. Jt Comm J Qual Patient Saf 2011;37(11):483-93.
11. Glickman SW, Cairns CB, Otero RM, Woods CW, Tsalik EL, Langley RJ, et al. Disease progression in Hemodynamically stable patients presenting to the emergency de- partment with sepsis. Acad Emerg Med 2010;17(4):383-90.
12. Kennedy M, Joyce N, Howell MD, Lawrence Mottley J, Shapiro NI. Identifying infected emergency department patients admitted to the hospital ward at risk of clinical dete- rioration and intensive care unit transfer. Acad Emerg Med 2010;17(10):1080-5.
13. Mikkelsen M, Miltiades A, Gaieski D, Goyal M, Fuchs B, Shah C, et al. Serum lactate is associated with mortality in severe sepsis independent of organ failure and shock. Crit Care Med 2009;37:1670-7.
14. Puskarich MA, Kline JA, Summers RL, Jones AE. Prognostic value of incremental lac- tate elevations in emergency department patients with suspected infection. Acad Emerg Med 2012;19(8):983-5.
15. Musikatavorn K, Thepnimitra S, Komindr A, Puttaphaisan P, Rojanasarntikul D. Venous lactate in predicting the need for intensive care unit and mortality among nonelderly sepsis patients with stable hemodynamic. Am J Emerg Med 2015;33(7):925-30.
16. Liu V, Morehouse JW, Soule J, Whippy A, Escobar GJ. Fluid volume, lactate values, and mortality in sepsis patients with intermediate lactate values. Ann Am Thorac Soc 2013;10(5):466-73.
17. Jones AE, Shapiro NI, Trzeciak S, Arnold RC, Claremont HA, Kline JA, et al. Lactate clearance vs Central venous oxygen saturation as goals of early sepsis therapy: a ran- domized clinical trial. JAMA 2010;303(8):739-46.
18. Puskarich MA, Trzeciak S, Shapiro NI, Albers AB, Heffner AC, Kline JA, et al. Whole blood lactate kinetics in patients undergoing quantitative resuscitation for severe sepsis and septic shock. Chest 2013;143(6):1548-53.
19. Puskarich MA, Trzeciak S, Shapiro NI, Arnold RC, Heffner AC, Kline JA, et al. Prognos- tic value and agreement of achieving lactate clearance or central venous oxygen sat- uration goals during early sepsis resuscitation. Acad Emerg Med 2012;19(3):252-8.
20. Puskarich MA, Trzeciak S, Shapiro NI, Arnold RC, Horton JM, Studnek JR, et al. Association between timing of antibiotic administration and mortality from septic shock in patients treated with a quantitative Resuscitation protocol. Crit Care Med 2011;39(9):2066-71.
21. Trzeciak S, Dellinger RP, Chansky ME, Arnold RC, Schorr C, Milcarek B, et al. Serum lactate as a predictor of mortality in patients with infection. Intensive Care Med 2007;33(6):970-7.
22. 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;45(5):524-8.
23. Dettmer M, Holthaus CV, Fuller BM. The impact of serial lactate monitoring on emer- gency department resuscitation interventions and clinical outcomes in severe sepsis and septic shock: an observational cohort study. Shock 2015;43(1):55-61.
24. Nguyen HB, Rivers EP, Knoblich BP, Jacobsen G, Muzzin A, Ressler JA, et al. Early lac- tate clearance is associated with improved outcome in severe sepsis and septic shock. Crit Care Med 2004;32(8):1637-42.
25. Puskarich MA, Illich BM, Jones AE. Prognosis of emergency department patients with suspected infection and intermediate lactate levels: a systematic review. J Crit Care 2014;29(3):334-9.
26. DiPietro NA. Methods in epidemiology: Observational study designs. Pharmacother- apy 2010;30(10):973-84.