emiumslcbt(OalM1t1ftts

et

S7

RA

The Journal of Emergency Medicine, Vol. 38, No. 5, pp. 578–581, 2010Copyright © 2010 Elsevier Inc.

Printed in the USA. All rights reserved0736-4679/$–see front matter

doi:10.1016/j.jemermed.2007.11.099

OriginalContributions

POINT-OF-CARE TESTING AT TRIAGE DECREASES TIME TO LACTATE LEVELIN SEPTIC PATIENTS

Munish Goyal, MD, Jesse M. Pines, MD, MBA, Byron C. Drumheller, BA, and David F. Gaieski, MD

Department of Emergency Medicine, University of Pennsylvania, Philadelphia, PennsylvaniaReprint Address: Munish Goyal, MD, Department of Emergency Medicine, Ground Ravdin, 3400 Spruce Street, Philadelphia,

PA 19104

itie

sbibssat(h

(�rtvImp

dtti

Abstract—Early recognition of elevated lactate levelsay hasten the detection of time-sensitive illness. We stud-

ed a method to measure lactate levels in septic patientssing a point-of-care (POC) device at Emergency Depart-ent triage. A convenience sample of adult patients with

epsis was enrolled. Consenting patients received a fingertipactate measurement using a POC device. Treating clini-ians were blinded to study and POC test results. Wholelood lactate levels were drawn at the discretion of thereating physician. A total of 238 patients were eligible; 15465%) consented, 5 left without treatment after consenting.f the remaining 149, 44 (30%) patients received both POC

nd whole blood lactates. Median time from triage to POCactate result was 21 min (interquartile range [IQR] 14–25).

edian time from triage to whole blood lactate result was72 min (IQR 119–256). The time difference between POCest result and whole blood lactate result was 151 min (IQR01–247). Triage POC fingertip lactate is a feasible methodor reducing time to identification of lactate levels in pa-ients with sepsis. Larger studies are needed to determinehe ability of POC lactate measurement to aid in the risktratification of septic patients. © 2010 Elsevier Inc.

Keywords—sepsis; lactate; triage; point-of-care sys-ems; emergency medicine

INTRODUCTION

epsis continues to challenge clinicians. An estimated51,000 annual cases led to over 215,000 deaths

ECEIVED: 1 September 2006; FINAL SUBMISSION RECEIVED

CCEPTED: 15 November 2007

578

n the United States alone. As the population continueso age, the overall incidence of sepsis is expected toncrease because sepsis is more common in thelderly (1).

Sepsis is defined as the systemic inflammatory re-ponse to infection. Although several parameters haveeen used to define inflammation, two or more systemicnflammatory response syndrome (SIRS) criteria haveeen quoted in the recent literature (2,3). Within thepectrum of sepsis, lactate level is a tool used totratify patient illness severity. Lactate is a marker ofnaerobic respiration, often elevated in the setting ofissue hypoxia. An elevated Emergency DepartmentED) lactate level is an indicator of increased likeli-ood of death (3).

A recent ED trial of early goal-directed therapyEGDT) for patients with severe sepsis and a lactate level

4 mmol/L or septic shock showed a 16% absoluteeduction in 28-day mortality, indicating the time sensi-ivity of the sepsis syndrome (2). Subsequently, the Sur-iving Sepsis Campaign and the Institute for Healthcaremprovement published recommendations that lactateeasurement be available within minutes to help identify

atients eligible for EGDT (4).We sought to utilize a fingertip point-of-care (POC)

evice for the measurement of blood lactate level at EDriage in patients with suspected sepsis. We hypothesizedhat the use of a POC device would decrease time todentification of blood lactate levels.

arch 2007;

: 6 M

S

TnAIia

S

Thait

aarccrsaepw

S

A1aiuwaumecuecgptp

M

TllsgNaDl1trlccnmdR

D

Drt(s

TtpcasTwwMP1fi

Tt

Point-of-Care Lactate at Triage 579

METHODS

tudy Design

his was a prospective, observational study of a conve-ience sample of adult patients conducted from May tougust 2005. The study protocol was approved by the

nstitutional Review Board for Human Research at ournstitution. Written informed consent was obtained fromll patients before enrollment.

tudy Setting and Population

he study was conducted in a 700-bed urban tertiary careospital with a 56-bed ED and approximately 55,000dult patient visits per year. The triage station and wait-ng room are located in a single enclosed area apart fromhe main treatment area.

Inclusion criteria consisted of a suspected infectionnd two or more of the following SIRS criteria: temper-ture � 38°C or � 36°C, heart rate � 90 beats/min, andespiratory rate � 20 breaths/min. The remaining SIRSriterion was unavailable at the time of enrollment be-ause the patients were evaluated at triage, before labo-atory draw. Eligible patients 18 years or older werecreened and approached for study enrollment immedi-tely after triage. Patients arriving by ambulance werenrolled only if they were sent to the triage station. Allatients arriving by ambulance that bypassed triage andent directly to a room were excluded.

tudy Protocol

ll patients presenting to triage between 11:00 a.m. and:00 a.m., Monday–Friday, were screened by researchssistants using the ED computerized tracking and chart-ng system. Consenting patients received a finger-pricksing a disposable lancet. Blood lactate concentrationas measured with a previously validated Lactate Pro™

nalyzer (LT-1710, Arkray Inc., Kyoto, Japan), whichses a 5-�L sample, has a range from 0.8 to 23.3mol/L, and provides results in 60 s (5). The device

mploys single-use test strips containing an enzyme-oated electrode and was calibrated every 8 h while inse. The treating Emergency Physicians were blinded toach patient’s study participation, data collection pro-ess, and fingertip lactate test results. The study investi-ators did not influence the clinical decision-makingrocess. Patients were included in the data analysis if thereating clinicians obtained a whole blood lactate level as

art of their clinical evaluation. o

easurements

he primary outcome was time from triage to POCactate result vs. time from triage to whole bloodactate result obtained via standard laboratory analy-is. Whole blood lactate level was obtained using arey-top tube (BD Vacutainer™ [BD, Franklin Lakes,J] sodium fluoride 10 mg/potassium oxalate 8 mg)

nalyzed on a Vitros® 950 analyzer (Ortho-Clinicaliagnostics, Rochester, NY). The Vitros® 950 ana-

yzer processes whole blood lactate in approximately2 min. Triage vitals signs, infectious source, triageime, triage complaint, triage classification, time tooom placement, patient demographics, diagnoses,aboratory test results, and disposition were also re-orded. Brief health history and physical examinationhecklists were completed based upon the triageurse’s documented assessment. Data were docu-ented on collection forms and then entered into

atabase software (Access 2000, Microsoft® Corp.,edmond, WA).

ata Analysis

escriptive statistics are reported with interquartileanges (IQR). Differences in paired time data wereested using the Wilcoxon sign-rank test. Stata 9.0StataCorp LP, College Station, TX) was used for thetatistical analysis.

RESULTS

here were 238 patients eligible for enrollment duringhe 3-month study period. One hundred fifty-four (65%)atients consented; 5 patients left without treatment afteronsenting. Forty-four (30%) patients received both POCnd whole blood lactate level measurements, creating ourample. Characteristics of the 44 patients are shown inable 1. Median time from triage to POC lactate resultas 21 min (IQR 14–25). Median time from triage tohole blood lactate result was 172 min (IQR 119–256).edian “triage to test result” time difference between

OC and whole blood lactate results was 151 min (IQR01–247). Eleven of the 44 patients in this sample had angertip lactate � 3, all of whom were admitted.

DISCUSSION

his study found that use of a POC lactate device atriage decreased test result times by 2½ h. In the context

f national ED crowding, a reduction in time to a poten-

tppmmm

P3lbfpr

ltbdo

ti(sdptenrvdliwo

dnrobficwhPP(esPpBPut

Thchat

T

NPPNNP

AMT

S

T

P

MMD

*dLsmcI

580 M. Goyal et al.

ially management-changing test may have a large im-act on triage classifications and assist in identifying illatients upon arrival. Given these results, POC lactateeasurement at triage may fulfill recommendations toeasure lactate levels in patients with sepsis “withininutes of arrival” (4).In a previous study, we demonstrated that fingertip

OC testing saves 65 min (95% confidence interval0–103) compared to simultaneously drawn whole bloodactate (6). The added difference in time to test resultetween the two testing strategies in this study derivesrom a combination of factors, including time to roomlacement, physician evaluation, test request, and test

able 1. Patient Characteristics and Disposition

o. patients screened 238atients refused, n (%) 84 (35)atients enrolled at triage, n (%) 154 (65)o. patients LWBS 5o. patients with POC lactate measured 149atients with whole blood lactate measured,

n (%)44 (30%)

ge, years 48.8 � 17.1ale, % 55

riage SIRS criteria, n (%)Temperature � 36°C 2 (5)Temperature � 38°C 38 (86)Heart rate � 90 beats/min 42 (95)Respiratory rate � 20 breaths/min 9 (20)

ource of infection known at triage, n (%)Known 14 (32)Unknown 30 (68)

riage complaint*, n (%)Fever 20 (45)Infection 4 (9)Shortness of breath 4 (9)Abdominal pain 4 (9)Neutropenia 3 (7)Chest pain 3 (7)Other 6 (14)

ast medical history, n (%)Immunocompromised 17 (39)Hypertension 13 (30)Diabetes 8 (18)HIV 4 (9)Taking anti-retroviral medication 3 (7)Dialysis 3 (7)Cirrhosis 1 (2)CHF 1 (2)ean fingertip lactate (mmol/L), (95% CI) 2.15 (1.81–2.49)ean whole blood lactate (mmol/L), (95% CI) 1.93 (1.49–2.37)isposition, n (%)Treated and discharged 6 (14)Admitted 38 (86)

In-patient ward 30 (79)IMC 6 (16)ICU 2 (5)

Triage complaint was chosen from a limited number of pre-efined categories by the triage nurse.WBS � left without being seen; POC � point of care; SIRS �ystemic inflammatory response syndrome; HIV � human im-unodeficiency virus; CHF � congestive heart failure; CI �

onfidence interval; IMC � intermediate intensive care unit;CU � intensive care unit.

equest recognition. In a system with minimal triage o

aboratory testing such as ours, it is much more feasibleo incorporate fingertip POC testing as opposed to wholelood testing at triage. Our study reflects the real worldifference triage POC testing may have in systems with-ut triage laboratory testing.

These results mirror recent trends toward early iden-ification and treatment of patients with time-sensitivellnesses such as stroke and acute myocardial infarction7). Patients with stroke typically manifest obviousymptoms, allowing resources to be quickly deployed toetermine eligibility for thrombolysis. The majority ofatients with myocardial infarction can be identified byhe early use of the electrocardiogram (ECG) to differ-ntiate patients who have acute ST-segment elevation orew left bundle branch block and require immediateeperfusion therapy. By contrast, patients with an ele-ated lactate in the setting of sepsis may not manifestramatic signs and symptoms. The employment of POCactate measurement provides a quick-sort method, sim-lar to the ECG at triage, to determine which patientsith sepsis have a time-sensitive illness and are in needf immediate intervention.

The use of POC lactate measurement has been vali-ated in the care of critically ill trauma patients but hasot been studied in ED sepsis patients (8). Although weeport mean lactate values in Table 1, correlation testingf these values is limited due to the temporal separationetween fingertip and whole blood sampling. The use ofngertip capillary blood for lactate testing was recentlyautioned by Boldt et al. due to suboptimal accuracyhen compared to laboratory analyzed arterial blood (9);owever, our results from a pilot study using the Lactatero™ demonstrated good agreement between fingertipOC and reference method whole blood lactate levelsintra-class correlation coefficient � 0.90, mean differ-nce � �0.26) with POC lactate measurements biasedlightly higher (6). Although Boldt et al. used a differentOC device, the difference may be attributed to theatient population tested. We focused on ED patients;oldt et al. studied intensive care unit (ICU) patients.atients in ICU settings often have received large vol-me resuscitation resulting in diffuse tissue edema, po-entially altering the fingertip sample.

LIMITATIONS

he results of this study may not be generalizable to allospitals in that turnaround times and levels of EDrowding can vary considerably. Because we did notave paired values for all patients enrolled, there may bespectrum effect where sicker patients were more likely

o have both tests because whole blood lactates were

rdered at the discretion of the treating clinicians. As

potdaan

Taisim

1

2

3

4

5

6

7

8

9

Point-of-Care Lactate at Triage 581

reviously mentioned, fingertip POC lactate levels werebtained before whole blood lactate levels. During thisime, patients’ volume status may have slightly improvedue to early fluid therapy or worsened due to delays indequate resuscitation. Thus, correlation between POCnd whole blood lactate measurements in this study wereot performed.

CONCLUSION

his pilot study shows that the use of a POC lactate devicet ED triage may represent a useful method for the earlydentification of lactate levels in septic patients. Largertudies are needed to determine the consequences of detect-ng elevated lactate levels among septic patients in such aanner.

REFERENCES

. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J,Pinsky MR. Epidemiology of severe sepsis in the United States:

analysis of incidence, outcome, and associated costs of care. CritCare Med 2001;29:1303–10.

. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy inthe treatment of severe sepsis and septic shock. N Engl J Med2001;345:1368–77.

. Shapiro NI, Howell MD, Talmor D, et al. Serum lactate as apredictor of mortality in emergency department patients with infec-tion. Ann Emerg Med 2005;45:524–8.

. Institute for Healthcare Improvement. Implement the sepsis resus-citation bundle. serum lactate measured. Available at: http://www.ihi.org/IHI/Topics/CriticalCare/Sepsis/Changes/IndividualChanges.Accessed November 30, 2005.

. Medbo JI, Mamen A, Holt Olsen O, Evertsen F. Examination of fourdifferent instruments for measuring blood lactate concentration.Scand J Clin Lab Invest 2000;60:367–80.

. Goyal M, Fuchs BD, Shofer FS, Drumheller BC, Zogby K, GaieskiDF. Validation of point-of-care fingertip lactate measurement in theemergency department. Ann Emerg Med 2006;48(4S):2.

. Menown IB, Allen J, Anderson JM, Adgey AA. ST depression onlyon the initial 12-lead ECG: early diagnosis of acute myocardialinfarction. Eur Heart J 2001;22:218–27.

. Slomovitz BM, Lavery RF, Tortella BJ, Siegel JH, Bachl BL,Ciccone A. Validation of a hand-held lactate device in determina-tion of blood lactate in critically injured patients. Crit Care Med1998;26:1523–8.

. Boldt J, Kumle B, Suttner S, Haisch G. Point-of-care (POC) testing

of lactate in the intensive care patient. Acta Anaes Scand 2001;45:194–9.