Article, Emergency Medicine

ED bedside point-of-care lactate in patients with suspected sepsis is associated with reduced time to iv fluids and mortality

a b s t r a c t

Objective: Early recognition and treatment of sepsis improves outcomes. We determined the effects of bedside point-of-care (POC) lactate measurement on test turnaround time, Time to administration of IV fluids and antibiotics, mortality, and ICU admissions in adult ED patients with suspected sepsis. We hypothesized that bedside lactate POC testing would reduce time to IV fluids and antibiotics.

Methods: We compared 80 ED patients with suspected sepsis and a lactate level of 2 mmol/L or greater before and 80 similar patients after introduction of POC lactate measurements. Groups were compared with ?2 and Mann Whitney U tests. A sample size of 80 patients in each group had 85% power to detect a 30-minute difference in time to IV fluids or antibiotics.

Results: Study groups were similar in age, gender, baseline lactate levels, sepsis severity, and Sequential Organ Failure Assessment scores. Introduction of POC lactate was associated with significant reductions in test turnaround time (34 [26-55] vs. 122 [82-149] minutes; P b 0.001), time to IV fluids (55 [34-83] vs. 71 [42- 110] minutes; P = 0.03), mortality (6% vs. 19%; P = 0.02), and ICU admissions (33% vs. 51%, P = 0.02), but not time to IV antibiotics (89 [54-156] vs. 88 [60-177] minutes; P = 0.35).

Conclusions: Implementation of bedside POC lactate measurement in adult ED patients with suspected sepsis reduces time to test results and time to administration of IV fluids but not antibiotics. A significant reduction in mortality and ICU admissions was also demonstrated, which is likely due, at least in part, to POC testing.

(C) 2014

Introduction

Sepsis is defined as a systemic inflammatory response to infection. Each year there are approximately 1,000,000 cases of sepsis treated in emergency departments (ED) in the U.S. with a mortality rate ranging from 20-50% [1-3]. Early identification of sepsis allows initiation of goal-directed therapies, such as antibiotics and intravenous fluids, aimed at reducing morbidity and mortality [4]. Delays in initiating antibiotics are associated with increased mortality [5]. Based on the evidence, a number of professional societies have published consen- sus-based guidelines on the evaluation and management of sepsis [6].

? Source of funding: Abbott Point of Care, Princeton, NJ.

?? Author Contributions: AJS conceived of the study, and participated in its design

and coordination and helped to draft the manuscript. HCT performed the statistical

analysis of the data. All authors read and approved the final manuscript.

? Competing Interests: This research was supported by a research grant from Abbott

Point of Care (Princeton, NJ). In addition to the provision of funding, Abbott was consulted during the design of the study. Study conception, data analysis, interpretation and manuscript preparation was performed independently by the study investigators. The first author (AJS) is on the Speaker’s Bureau of Abbot Point of Care.

* Corresponding author. Department of Emergency Medicine, Stony Brook University, Stony Brook, NY 11794-8350. Tel.: +1 631 444 7857; fax: +1 631 444 3919.

E-mail address: [email protected] (A.J. Singer).

A number of studies have shown that lactate levels are predictive of outcomes in patients with sepsis [7-9]. As a result the guidelines recommend early measurement of lactate levels in order to identify patients with tissue hypoperfusion who are at the greatest risk of morbidity and mortality, especially in patients with cryptic shock in which hypotension is not yet apparent [6]. A major limitation with reliance on early lactate measurements is the delays associated with central laboratory testing. Point-of-care (POC) testing at the bedside with a hand held accurate device has the potential to significantly reduce test turn around times and more importantly therapeutic turn around times, which represent the time from ordering a test until the results are available to the practitioner allowing therapeutic decisions. Several observational studies of ED patients with suspected sepsis found that compared with central lab testing, POC measurement of lactate was both feasible and accurate [10,11]. However, these studies did not compare test turn around time or determine the clinical impact of POC testing.

The current study was designed to determine whether introducing bedside lactate POC testing would reduce the time to lactate measurement and thereby shorten the time to recognition of severe sepsis and initiation of IV fluids and antibiotic therapy. We hypothesized that compared with central lab testing, bedside POC lactate testing would reduce the time to intravenous fluid and antibiotic administration in ED patients with sepsis.

http://dx.doi.org/10.1016/j.ajem.2014.06.027 0735-6757/(C) 2014

Methods

Study design

We conducted a before and after study to test the study hypothesis. Consecutive patients identified by our institutional sepsis screening tool that visited the ED before introduction of bedside lactate measurements were compared with a similar sized conve- nience sample of patients with suspected sepsis that presented to the ED after introducing bedside lactate measurements. Our study was approved by the Institutional Review Board and all patients enrolled in the prospective arm of the study (or their legal guardians) gave written informed consent. Waiver of informed consent was obtained for patients enrolled in the retrospective arm of the study.

Setting

The study was conducted at a suburban, academic tertiary care medical center with an annual ED census of approximately 90,000. The study center is also the site of an emergency medicine residency training program. The study site has been part of the Surviving Sepsis Campaign for more than a decade.

Patients

Our institution maintains an ongoing sepsis registry that includes all admitted patients that screened positive for suspected infection and had at least two SIRS criteria. Using this database we conducted a structured medical record review of the first 80 consecutive patients presenting to our ED 12 months prior to introduction of bedside lactate testing who also had an initial lactate level of at least 2 mmol/L, starting from one calendar year prior to study initiation. Validation of inclusion criteria, SOFA scores, sepsis severity, lactate levels as well as the primary and secondary outcomes was verified by a second chart abstracter and any disagree- ments between the two abstractors were resolved by consensus. Chart abstraction followed the methods described by Gilbert et al. [12] Patients in the after period were prospectively enrolled by an investigator who screened all ED patients for the following: suspected infection and at least two of the clinical criteria for the Systemic Inflammatory Response Syndrome including a temperature of 38 degrees Celsius or greater, a temperature of 35 degrees Celsius or less, a heart rate greater than or equal to 90 beats per minute, a respiratory rate of 20 or greater per minute, a systolic blood pressure less than 90 mmHg, or an acute change in mental status. Screening was performed as soon as possible after the patient was triaged and assigned a bed. Patients who could not give consent or in whom consent could not be obtained from a legal guardian were excluded. Patient who received an intravenous antibiotic for suspected sepsis within the last 12 hours were also excluded.

Study interventions

After receiving informed consent a whole blood specimen was obtained from all patients in the after group who met inclusion criteria by venipuncture and placed in a heparin coated vacutainer tube. Immediately after obtaining the blood specimen, a small drop of blood was placed onto a study cassette and analyzed at the bedside for lactate levels using a portable, hand held POC device (i-STAT System, Abbott Point Of Care, Princeton, NJ). Test results were available within 2-3 minutes. If the lactate level was less than 2 mmol/L the patient was excluded from the study, regardless of clinical presentation including patients in obvious septic shock. In patients with a lactate level of 2 mmol/L or greater the bedside POC results were immediately communicated to the physician and nurse assigned to the study patient. Patients in whom the initial lactate level was 4 mmol/L or greater were transferred to the critical care area for further evaluation and management. Further treatment, including ordering of

diagnostic tests, IV fluids, and antibiotics was at the discretion of the treating attending physician. A repeat bedside POC lactate measure- ment was performed approximately two hours later and the results were immediately communicated to the treating physician and nurse. In order to determine the accuracy of the bedside POC lactate measurements, results on the POC device were compared to those from the central lab. However, all Clinical decisions were made based on the POC measurements.

Measures and outcomes

Structured collection of demographic and clinical information was performed by trained research assistants and entered into REDCAP (Research Electronic Data CAPture) software, which is a secure web- based application used for database management. The time of patient arrival and triage, ordering and reporting of central lactate levels, IV catheter insertion, IV fluid administration, antibiotic administration, time of decision to admit, and actual time of admission to a floor or an ICU were extracted from the electronic medical records. These time points are routinely captured by the patient information system. A structured chart review by one of the principle investigators was performed to determine ultimate diagnosis and severity of sepsis (sepsis, severe sepsis, septic shock) [2], source of infection, Sequential Organ Failure Assessment scores [13], modified early warning scores (MEWS) scores [14], length of stay, and in-hospital mortality masked to study group and lactate levels. The primary outcomes were time from ED triage to IV fluids and antibiotic administration. Secondary outcomes were time from ED triage to ordering of antibiotics, total volume of IV fluids given within the ED or first 6 hours (whichever was shorter), ED length of stay, need for vasoactive agents, admission to an intensive care unit (ICU), length of stay in the ICU, total length of hospital stay, and in-hospital mortality.

Data analysis

Binary data were summarized as the percentage frequency of occurrence and compared between groups with ?2 or Fischer’s exact tests. Continuous data were presented as means and 95% confidence intervals (CI) for parametric data and medians and inter-quartile ranges (IQR) for nonparametric data and compared with t-tests and Mann Whitney U tests as appropriate.

Prior experience at our institution suggested that the mean time to administration of antibiotics in septic patients is approximately 90 +/

– 64 minutes (unpublished data). We felt that a reduction of at least 30 minutes in the time to IV fluid and antibiotic administration would be clinically relevant. Therefore we calculated that a sample size of 80 patients in each of the study periods (before and after) would have 85% power to detect a difference of 30 minutes or greater in time to IV fluids and/or antibiotics at a significance level of 0.05. The agreement between bedside POC and central lab lactates was analyzed with scatterplots, Correlation coefficients and Bland Altman analysis. Stepwise logistic regression was performed to determine the association between potential predictor variables (study group, gender, lactate level, sepsis severity, source of infection, time to antibiotics and fluids) and mortality.

Results

During the prospective portion of the study 258 patients presented to the ED when an investigator was present with suspected infection and met at least two of the physiological criteria for SIRS and consented to be in the study. Of these patients 80 (29%) had a bedside POC lactate of at least 2 mmol/L and were included in the study between January and September 2013. In the remaining 178 patients the lactate level was less than 2 mmol/L and these patients were excluded from the study. The 80 patients from the retrospective

portion of the study were consecutive patients included in our hospital sepsis registry whose lactates were at least 2 mmol/L who presented to the hospital between January and November 2011. Thus, 160 patients (80 patients in the before period and 80 patients in the after period) were included in our analysis. Their median (IQR) age was 71 (56-83) years; 42% were female, 88% were white, 6% black, and 3% Hispanic. Overall in-hospital mortality was 12.5%.

Both study groups were well matched in baseline demographic and clinical characteristics (Table 1). Median [IQR] SOFA scores in the before and after study groups were similar (2.5 [1-5] vs. 2.5 [1-4]; P = 0.63) as were MEWS scores (Table 1). Baseline lactate levels were also similar in the two groups (Table 1). The percentage of patients in septic shock in the before and after group were 19% and 20% respectively. While Hospital admission rates were similar, ICU admission rates were higher in the before group (Table 2).

Introduction of POC measurement of lactate reduced the time to

Table 2

Outcomes

Before After P value

Mortality

15 (19%)

5 (6%)

0.02

Time from arrival

122 (82-149)

71 (53-101)?

b 0.001

to SOC result, min.

Time from arrival to POC

34 (26-55)

result (after group only), min.

Time from order to test SOC results, min.

71 (53-91)

38 (26-53)

b 0.001

Time from order to POC test

3 (2-3)

results (after group only), min

Time to IV fluids, min.

71 (42-110)

55 (34-83)

0.03

Time to antibiotic order, min.

62 (26-114)

69 (34-133)

0.27

Time to antibiotics, min.

97 (55-160)

89 (63-182)

0.59

Total ED LOS, min

326 (249-436)

352 (246-457)

0.50

ICU admits, No. (%)

41 (51%)

26 (33%)

0.02

ICU length of stay, days

4 (2-6)

3 (2-6)

0.90

Hospital length of stay??, days

8 (4-13)

7 (3-13)

0.27

availability of lactate results by 88 minutes compared with central lab testing in the before period (34 [26-55] vs. 122 [82-149]; P b 0.001, Fig. 1). While time to physician ordering of antibiotics and antibiotic administration was similar in the two study groups (Table 2, Figs. 2 and 3), introduction of POC lactate measurements was associated with a significant reduction in median (IQR) time to administration of IV fluids; 55 [34-83] vs. 71 [42-110] minutes (P = 0.03, Fig. 4) in the after and before groups respectively. While time to initiation of IV fluids was faster when bedside lactate measurements were per- formed, the total volume of IV fluids administered in the ED was similar in both study periods (Table 2). Bedside measurement of POC lactates was associated with a significant reduction in in-hospital mortality (6% vs. 19% in the after and before groups respectively; P = 0.02). It was also associated with a reduction in the percentage of patients that were admitted to an ICU (33% vs. 51%, P = 0.02). There were no between group differences in ED, ICU and total hospital lengths of stay (Table 2).

Significant predictors of mortality demonstrated on stepwise logistic regression included sepsis severity, OR 5.1 (95% CI, 1.7-

Table 1

Baseline patient characteristics

Before

After

P

% Female

43

41

0.87

Median age (IQR)

73 (60-83)

70 (56-83)

0.67

Race

White

91

85

0.59

Black

5

6

Asian

0

3

Hispanic

3

4

Other

1

3

Diabetes

33

33

1.00

COPD

23

20

0.70

CHF

31

23

0.21

CAD

49

39

0.20

HIV infection

0

1

0.32

Renal disease

3

5

0.41

Active malignancy

20

10

0.08

Organ transplant

0

3

0.16

Indwelling vasc. Line

4

5

0.70

Nursing home resident

35

21

0.053

Source of infection (%)

Respiratory

50%

29%

0.01

Urinary

26%

36%

0.17

Gastrointestinal

10%

5%

0.23

Skin

6%

11%

0.26

Blood

3%

5%

0.41

Other

6%

5%

0.73

Median (IQR) SOFA

2.5 (1-5)

2.5 (1-4)

0.63

Median (IQR) MEWS

3 (2-5)

3 (2-5)

0.64

Septic shock, No. (%)

15 (19%)

16 (20%)

0.84

Central lab lactate, mmol/L

3.0 (2.4-4.1)

3.0 (2.6-4.0)

0.84

Amount of IV fluid

2500 (2000-4000)

2000(2000-3125)

0.71

After group: Time to POC vs. time to SOC (values above): P b0 .001. Time to SOC, retrospective vs. prospective (see values above): P b0 .001.

* All patients in the prospective arm receiving a POC lactate test result also had their

Serum lactate levels measured in the central laboratory. The purpose of the central laboratory testing was to assess the performance of the POC lactate assay compared with the standard of care (SOC). It is important to note that treatment was initiated based on the POC result; it was not delayed or contingent on the value or the availability of the central lab serum lactate result.

?? excludes deaths.

15.3); before study group, OR 3.8 (95% CI, 1.2-12.3); and initial lactate level, OR 1.4 per mmol/L (95% CI, 1.1-1.8).

Agreement between bedside POC and central lab lactates

The correlation between the first bedside POC lactate and central lab lactate was 0.94 [95% CI, 0.91-0.97] (Fig. 5). The mean difference between POC and central lab lactates of 0.26 +/-0.43 mmol/L. Bland Altman analysis demonstrated that most differences ranged between about -1 to 0.6 (Fig. 6). All of the prospective patients had an initial POC lactate of 2 mmol/L or greater. Of these, only 2 (2.5%, 95% CI 0.4- 9.6%) had a central lab lactate less than 2 mmol/L.

Serial lactates

Of the 80 patients in the after group, a second bedside POC lactate was measured in 68 (85%) patients within a median (IQR) of 146

Fig. 1. Boxplots of time to lactate results administration. The black line in the middle of the box represents the median and the box contains the inter-quartile range.

Fig. 2. Boxplots of time to antibiotic ordering. The black line in the middle of the box represents the median and the box contains the inter-quartile range.

Fig. 4. Boxplots of time to intravenous (IV) fluid administration. The black line in the middle of the box represents the median and the box contains the inter-quartile range.

(126-165) minutes. The median (IQR) second POC lactate was significantly lower than the first lactate (1.7 [1.0-2.4] mmol/L vs. 2.7 [2.3-3.6] mmol/L; P b 0.001 respectively). The lactate levels increased in 5 (7%) patients and dropped by at least 10% in 59 (87%) patients. The second lactate normalized, dropping below 2 mmol/L, in 43 of the 68 (63%) patients who had serial POC bedside lactates. Mortality rates among patients whose second lactate normalized and those who did not were 2% and 12% respectively (P = 0.10).

Discussion

In our study introduction of bedside POC measurements of lactate was associated with a significant reduction in time to test results, time to administration of intravenous fluids, ICU admission rates and in- hospital mortality in ED patients with suspected sepsis. However, measurement of bedside lactate was not associated with a reduction in time to administration of IV antibiotics. The fact that IV fluids are readily available at the patient bedside while antibiotics are not always readily available may help explain why IV fluids were administered faster while there was no impact on antibiotic administration. Whether or not the reduction in ICU admission rates and mortality were a direct result of earlier measurement of lactate is unclear. Based on SOFA scores, MEWS

Fig. 3. Boxplots of time to antibiotic administration. The black line in the middle of the box represents the median and the box contains the inter-quartile range.

scores, sepsis severity, and initial lactate levels, clinical severity was similar in the two study groups and cannot account for the large difference in mortality. It is likely that other unmeasured or unidentified confounding variables may explain some of the reduction in mortality seen in the after group, especially since the mortality was considerably lower than in previous reports. Since the physicians and nurses caring for patients in the after group were aware of POC testing, a Hawthorne effect leading to more aggressive therapy may have biased the study in favor of the group that had bedside lactate measurements. It seems that introduction of bedside testing tends to emphasize the importance of rapid diagnosis and treatment encouraging the entire system to become more efficient. This possibility is further supported by the fact that even the time to standard central laboratory lactate results was also shorter in the after group. Our institution has participated in the Surviving Sepsis Campaign for more than a decade and we are unaware of any other major changes in our clinical practice during the study period. Most likely the improvement in mortality noted in the after period was a result of active and aggressive screening of severe sepsis patients resulting in better and faster treatment. However, this was only made possible by the addition of beside point of care lactate testing.

While the reduction in time to lactate results was significantly shortened by introducing bedside POC testing, it still took a relatively

Fig. 5. Scatterplot of point-of-care and central lab lactates (r = 0.94).

Fig. 6. Bland Altman plot of point-of-care and central lab lactates. Mean difference between lactates is -.26 (SD .43).

long time between patient arrival and lactate results. Some of the time delay from patient arrival to lactate measurement was an artificial byproduct of the study design that required full informed consent prior to performing POC testing, which may have required 10-20 minutes. The reduction in time to lactate results in the real word scenario is likely to be even greater than in our clinical study environment. Indeed, since completion of the current study we have introduced routine POC testing in all critical care patients and have noted considerable shortening of times to completion of POC lactate measurements, in many cases within 10 minutes of patient arrival (unpublished data).

It is not clear why earlier lactate measurement was associated with a significant reduction in time to IV fluids but not antibiotic administration. While there is clear evidence that early initiation of goal directed therapy (which includes administration of IV fluids and antibiotics) reduces mortality in septic patients, it unclear which of the individual interven- tions in the overall Sepsis Bundle is most important. Our study also supports the feasibility of serial bedside POC measurement of lactate. Since studies now suggest that Lactate clearance is associated with outcome and can help guide management [9], the ability to measure more than one lactate at the bedside is more important than ever. While mortality was lower in patients whose lactate normalized, due to the relatively small sample size, this difference was not statistically significant. Finally, our study demonstrates that bedside measurement of POC lactate using the study device is rapid and highly accurate as evidenced by the excellent correlation and minimal differences in lactate values between the bedside and central lab lactate levels. The misclassification rate of roughly 2.5% when POC and central laboratory methods were compared seems quite reasonable to us and the clinical

relevance of this difference is unclear.

Limitations

Our study has several important limitations. The observational nature of the before and after study design allows us to identify associations between implementation of bedside POC lactate mea- surement and outcomes but does not prove causality. Also as noted above, there may have been other differences between the study groups or systematic changes that may have occurred over the study period that were not captured in our data. Since patients and health care practitioners were aware of the ongoing study, we cannot exclude a Hawthorne effect that may have biased the results in favor of the after group. This same Hawthorne effect may explain why the time to standard of care (central laboratory) lactate results was also improved in the after group.

Our study represents a convenience sample of patients that were identified when our Research staff is available (Mondays through Fridays from 8 am to 8 pm). Thus we cannot exclude a selection bias that may explain some of the differences between the two study groups. Because lactate levels were often obtained earlier in the patient’s presentation in the after study group, it is possible that some of the less severely ill septic patients in the before group may have been excluded from study inclusion since by the time lactates were obtained, patients may have started to receive IV fluids causing the lactate level to fall below 2 mmol/L. This would have biased the study in favor of the after group. As always, we also cannot exclude a type I error. Finally, our data is limited to a single institution and may not generalize to all settings.

Conclusions

Implementation of bedside POC lactate measurement in adult ED patients with suspected sepsis reduces time to test results and time to administration of IV fluids but not antibiotics. A significant reduction in mortality and ICU admissions was also demonstrated, which is likely due, at least in part, to POC testing.

Key Messages

  • Bedside point-of-care lactate measurement reduces time to test results
  • Bedside point-of-care lactate measurement reduces time to intra- venous fluids
  • Bedside point-of-care lactate measurement does not reduce time to Intravenous antibiotics
  • Bedside point-of-care lactate measurement is associated with reduced rates of ICU admission
  • Bedside point-of-care lactate measurement is associated with reduced mortality

References

  1. Martin GS. Sepsis, severe sepsis and septic shock: changes in incidence, pathogens and outcomes. Expert Rev Anti Infect Ther 2012;10:701-6.
  2. Martin GS, Mannino DM, Eaton S, et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003;348:1546-54.
  3. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiology of severe sepsis in the United States: Analysis of incidence, outcome, and Associated costs of care. Crit Care Med 2001;29:1303-10.
  4. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368-77.
  5. Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006;34:1589-96.
  6. Dellinger RP, Levy MM, Rhodes A, the Surviving Sepsis Campaign Guidelines Committee including the Pediatric Subgroup. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013;41:580-637.
  7. Shapiro N, Howell M, Talmor D, et al. Serum lactate as a predictor of mortality in emergency department patients with infection. Ann Emerg Med 2005;45:524-8.
  8. Mikkelson ME, Miltiades AN, Gaieski DF, et al. Serum lactate is associated with mortality in severe sepsis independent of organ failure and shock. Crit Care Med 2009;37:1670-7.
  9. 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 randomized clinical trial. JAMA 2010;303:739-46.
  10. Shapiro N, Fischer C, Suttner S, et al. The feasibility and accuracy of point-of-care lactate measurement in emergency department patients with suspected infection. J Emerg Med 2010;39:89-94.
  11. Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996;27:305-8.
  12. Gaieski DF, Drumheller BC, Goyal M, Fuchs BD, Shofer FS, Zogby K. Accuracy of handheld point-of-care fingertip lactate measurement in the emergency depart- ment. West J Emerg Med 2013;14:58-62.
  13. Innocenti F, Bianchi S, Guerrini E, Vicidomini S, Conti A, Zanobetti M, et al. Prognostic scores for early stratification of septic patients admitted to an emergency department-high dependency unit. Eur J Emerg Med 2014;21:254-9.
  14. Bulut M, Cebicci H, Sigirli D, Sak A, Durmus O, Top AA, et al. The comparison of modified early warning score with rapid emergency medicine score: a prospective multicentre observational cohort study on medical and surgical patients presenting to emergency department. Emerg Med J 2014;31:476-81.

Leave a Reply

Your email address will not be published. Required fields are marked *