a b s t r a c t

Background: Previous studies suggest a relationship between chloride-rich intravenous fluids and acute kidney injury in critically ill patients.

Objectives: The aim of this study was to evaluate the relationship of intravenous fluid chloride content to kidney function in patients with severe sepsis or septic shock.

Methods: A retrospective chart review was performed to determine (1) quantity and type of bolus intravenous fluids, (2) serum creatinine (Cr) at presentation and upon discharge, and (3) need for Emergent hemodialysis (HD) or Renal replacement therapy . Linear regression was used for continuous outcomes, and logistic regression was used for binary outcomes and results were controlled for initial Cr. The primary outcome was change in Cr from admission to discharge. Secondary outcomes were need for HD/RRT, length of stay (LOS), mortality, and organ dysfunction.

Results: There were 95 patients included in the final analysis; 48% (46) of patients presented with acute kidney injury, 8% (8) required first-time HD or RRT, 61% (58) were culture positive, 55% (52) were in shock, and overall mortality was 20% (19). There was no significant relationship between quantity of chloride administered in the first 24 hours with change in Cr (? = -0.0001, t = -0.86, R² = 0.92, P = .39), need for HD or RRT (odds ratio [OR] = 0.999; 95% confidence interval [CI], 0.999-1.000; P = .77), LOS N 14 days (OR = 1.000; 95% CI, 0.999-

1.000; P = .68), mortality (OR = 0.999; 95% CI, 0.999-1.000; P = .88), or any type of organ dysfunction. Conclusion: Chloride administered in the first 24 hours did not influence kidney function in this cohort with severe sepsis or septic shock.

(C) 2014

Introduction

Acute kidney injury is prevalent in patients with severe sepsis and septic shock [1-7]. Research has shown that AKI due to sepsis is associated with higher mortality, increased ventilator dependency, greater need for vasopressors, and higher length of stay (LOS) than other causes of AKI [8-10]. intravenous fluid administration is central

? Study Site: All patients were enrolled at UF Health Jacksonville, 655 West 8th Street, Jacksonville, FL 32209.

?? Author contributions: FWG, DJW, MH, and RLW conceived the study. FWG, DJW, AAB,

AS, NP, CJK, AEJ, and RLW supervised the data collection and chart reviews. RLW, SD, CJK, and AEJ provided methodological and statistical advice on study design and data analysis. AEJ and SD provided expertise on clinical trial design, and AAB and AS provided clinical ex- pertise regarding the subject matter. FWG, MH, DJW, AAB, AS, and NP drafted the manu- script, and all authors contributed substantially to its revision.

? Funding was provided by a UF Faculty Dean’s Fund Grant.

* Corresponding author at: Department of Emergency Medicine, UF Health Jacksonville, 655 West 8th Street, Jacksonville, FL 32209. Tel.:+1 904 244 2986.

E-mail address: [email protected] (F.W. Guirgis).

to sepsis management, but controversies exist regarding the effects of

fluid type and volume on kidney function [6,11,12].

Fluid optimization is necessary for adequate renal perfusion in preshock and shock states; however, several factors associated with fluids have been postulated to impair kidney function. Chloride (Cl) content in particular has been shown to have a negative effect on kidney function [13]. This is clinically relevant given the commonplace use of 0.9% saline (NS), a Cl-rich solution, in daily practice. Cl-rich fluids have been demonstrated to cause arteriolar vasoconstriction resulting in a decrease in renal blood flow and glomerular filtration rate [14,15]. A study by Yunos et al. [13] showed that eliminating Cl-rich fluids from the intensive care unit (ICU) results in a reduction in AKI according to the risk, injury, failure, loss end-stage classification, and less need for Renal replacement therapy . Others have noted Hyperchloremic metabolic acidosis, fluid retention, and decreased urine output as reasons for avoiding Cl-rich fluids in critically ill patients [16].

Separately, patients with severe sepsis or septic shock in the emer- gency department (ED) are prone to overresuscitation and fluid

http://dx.doi.org/10.1016/j.ajem.2014.12.013

0735-6757/(C) 2014

overload because reliable measures of fluid status are not always readily available [17]. fluid overload is associated with poor outcome and in- creased mortality particularly when greater than 10% [18-21]. Fluid overload is also associated with decreased survival, increased ventilator and ICU days [22], and in patients with AKI, it has been associated with lack of renal recovery at 1 year [23].

The primary objective of this study was to evaluate the relationship of fluid Cl content to kidney function in patients undergoing a quantita- tive Resuscitation protocol for severe sepsis or septic shock in the ED and ICU. Secondarily, we also hoped to determine the effect of total fluid vol- ume in the first 24 hours on kidney function.

Materials and methods

Study design and setting

This study was a retrospective extension of a prospective cohort study in which patients with severe sepsis or septic shock were moni- tored for a 6-hour period of quantitative resuscitation. The following data were recorded at 0, 3, and 6 hours: vital signs, ETCO₂, SCVO₂, cen- tral venous pressure, urine output, lactate levels, FiO₂, ventilator set- tings, suspected source of infection, and disposition. To determine the relationship between chloride content and renal outcomes, we retro- spectively determined baseline and discharge creatinine as well as type and amount of fluid administered. The prospective study took place from June 1, 2012, to May 30, 2014, in the adult ED and ICU at Uni- versity of Florida (UF) Health Jacksonville. The UF Health Jacksonville ED is a high-acuity, academic, urban ED that treats approximately 90 000 patients per year. The medical ICU is a 28-bed closed unit. The research protocol was approved by the UF College of Medicine, Jacksonville, Insti- tutional Review Board.

Patient selection

Patients with suspected severe sepsis or septic shock presenting to the ED were assessed for the following inclusion criteria: (a) adult 18 years or older; (b) documented severe sepsis (as defined by 2 of 4 Sys- temic Inflammatory Response Syndrome (SIRS) criteria plus a suspected or confirmed source of infection with a lactate greater than 4 mg/dL or end-organ dysfunction) or septic shock (defined as hypoten- sion not responsive to 30 mL/kg intravenous fluids), and (c) candidate for treatment with early quantitative resuscitation. Patients were ex- cluded if they were incarcerated, were pregnant, required emergency surgery, were receiving treatment with noninvasive ventilation, had preexisting end-stage renal disease , or received greater than 1 L of fluids of unknown type.

Measurements

Power calculation, and da”>Early quantitative resuscitation for patients with severe sepsis or septic shock

At UF Health Jacksonville, early quantitative resuscitation involves two potential Treatment protocols: invasive or noninvasive. Patients who maintain a mean arterial pressure (MAP) greater than 65 and are without other signs of hemodynamic instability are treated with the noninvasive protocol. This includes the placement of two large bore intravenous cath- eters to achieve goals for fluid resuscitation (as indicated by inferior vena cava sonographic measurement), urine output, and Lactate clearance of 10% or greater over a 2-hour period. Patients with signs of hemodynamic instability as indicated by a MAP less than 65 or who are showing a lack of improvement or clinical deterioration after initiation of the noninvasive protocol receive an Edwards PreSep Central Venous Oximetry Catheter (Edwards Lifesciences, Irvine, California) placed in either the internal jug- ular or subclavian vein for continuous Central venous oxygen saturation (SCVO₂) and central venous pressure monitoring. For patients in need of SCVO₂ monitoring but without a PreSep catheter, SCVO₂ is measured

via venous blood gas from the distal port of the central venous catheter drawn every 2 hours. In addition, as part of early quantitative resuscita- tion, patients treated with both protocols received bolus antibiotics with- in 1 hour, MAP monitoring, Foley catheters to measure urine output and temperature, and serial lactate levels.

Chart review and primary end points

The Principal Investigator (PI) and a trained Research Assistant (RA) performed a chart review of enrolled patients using a standard method- ology [24,25] to assess the following: (1) quantity and type of bolus in- travenous fluids (NS, lactated Ringers, or any other type), (2) serum creatinine (Cr) at presentation and upon discharge, and (3) need for emergent hemodialysis (HD) or RRT during admission. Duplicate re- views were performed by the PI and a trained RA on a 10% subsample of patients to assess reliability. Chart review also included ED and inpa- tient medical records to confirm sepsis diagnosis and determine age, sex, ethnicity, patient disposition (discharged to home, nursing home, rehabilitation facility, hospice, or death), hospital LOS, comorbidities (diabetes mellitus, chronic obstructive pulmonary disease, ESRD, active malignancy, organ transplant, and HIV status), source of infection, cul- ture results, organ dysfunction, and the presence of shock. Organ dys- function was defined according to the 2012 Surviving Sepsis Guidelines and the Mortality in ED Sepsis score [6,26].

The primary outcome of our analysis was change in Cr (admission- discharge Cr). Secondary outcomes were need for HD/RRT, LOS, mortal- ity, organ dysfunction. To determine if total fluid volume in the first 24 hours adversely affected kidney function, we analyzed the same out- comes but substituted total fluid volume as the independent variable.

Predictor variables

The total volume of fluid administered from admission to 24 hours was determined and separated by fluid type. The quantity of Cl per liter of NS (154 mEq), LR (109 mEq), albumin (0 mEq), or any other fluid type was multiplied by liters administered in the first 24 hours after ED presentation to determine the total quantity of Cl administered in the first 24 hours. For total fluid volume, the total number of liters ad- ministered was summed.

Data collection and storage

Study data were collected and managed using REDCap (Research Electronic Data Capture) tools hosted at the UF [27]. REDCap is a secure, Web-based application designed to support data capture for research studies, providing (1) an intuitive interface for validated data entry;

(2) audit trails for tracking data manipulation and export procedures;

(3) automated export procedures for seamless data downloads to com- mon statistical packages; and (4) procedures for importing data from external sources. Graphical and statistical analyses were performed using Stata Version 12 (StataCorp LP, College Station, Texas).

Sample, power calculation, and data analysis

The original prospective study (comparison of ETCO₂ to SCVO₂ and lactate) was institutional review board-approved to enroll 115 patients. Preliminary data suggested a linear relationship; therefore, based on linear regression of our primary end points (ETCO₂ and SCVO₂), a sam- ple size of 70 patients was calculated to achieve 80% power to detect a regression coefficient of greater than or equal to 0.36 based on estimates of standard deviation from our pilot cases (SD of ETCO₂ ~ 4.00, SCVO₂ ~ 4.50) and a two-sided significance level of .05. Institutional review board approval was obtained to enroll an additional 45 patients to de- termine the relationship between ETCO₂ and lactate.

For the relationship of Cl and fluids to our primary and secondary outcomes, scatter plots were examined to inform the choice of regres- sion models. For the primary analysis, we constructed a linear regres- sion model in which change in Cr was the dependent variable and mEq of Cl was the independent (or predictor) variable. We constructed logistic regression models for each of the dichotomous secondary

Table 1

Characteristics of study subjects: demographics, comorbidities, suspected source of infec- tion, features of sepsis, and organ dysfunction

Variable	No. (%) of patients
Age, mean (SD), y	64 (13.6)
Race, white	47 (50)
Race, black	44 (46)
Race, other	4 (4)
Sex, male	51 (54)
Comorbidities
Diabetes mellitus	43 (45)
Nursing home resident	35 (37)
Chronic obstructive pulmonary disease	26 (27)
Active malignancy	8 (8)
Human immunodeficiency virus	5 (5)
Indwelling vascular line	2 (2)
Organ transplant	1 (1)
Suspected source of infection
Pulmonary	50 (53)
Urinary tract	35 (37)
Intra-abdominal	10 (11)
Skin/soft tissue	8 (8)
Unknown	4 (4)
Blood	2 (2)
Features of sepsis Culture positive	58 (61)
Shock	52 (55)
Blood culture positive	43 (45)
Gram positive	43 (45)
Gram negative	32 (34)
Mortality	19 (20%)
Organ dysfunction
Cardiovascular dysfunction	52 (55)
Neurologic dysfunction	48 (50)
Renal dysfunction	46 (48)
Renal failure requiring HD/RRT	8 (8)
Pulmonary dysfunction	41 (43)
Coagulopathy	24 (25)
Hepatic dysfunction	6 (6)

outcomes (HD and mortality) in which mEq of Cl was again the predic- tor variable. In addition, we dichotomized Cr (Cr >= 2 mg/dL vs Cr b 2 mg/ dL) based on the 2012 Surviving Sepsis Guideline criteria for sepsis- induced kidney dysfunction [6] and then analyzed using logistic regres- sion. Because LOS is strongly skewed to the right, it is not normally dis- tributed, which violates an assumption of linear regression. Therefore, LOS was dichotomized into greater than or less than 14 days and ana- lyzed using logistic regression.

Results

A total of 115 patients were enrolled, and 20 were excluded from final analysis due to the following: (a) ESRD (11); (b) received un- known fluids (4); (c) nonsepsis patients (3); (d) missing study labs

(1); and (e) consent rescinded by patient (1). Complete data were ob- tained on 95 (83%) subjects that constituted our study sample.

Table 1 presents patient characteristics. There were slightly more white patients (50%) than black patients (46%), and more patients were male (54%). The most common source of infection was lung (53%), followed by urinary tract (37%) and intra-abdominal (11%). For organ dysfunction, cardiovascular (55%) and neurologic (50%) were most prevalent, followed by renal (48%), pulmonary (43%),

coagulopathy (25%), and hepatic (6%). There were 8 patients (8%) who required emergency HD or RRT. Overall mortality was 20%. Almost 55% of patients were in shock, 61% were culture positive, with 45% blood culture positive. Gram-positive infections (45%) were more com- mon than Gram-negative (34%). Mean initial lactate was 4.5 mg/dL (95% confidence interval [CI], 3.8-5.2), and mean baseline SCVO₂ was 72 mm Hg (95% CI, 68.6-76.0).

The mean quantity of Cl administered in the first 24 hours was 1012 mEq (95% CI, 851-1174). The mean volume of fluids administered was 7.3 L (95% CI, 6.1-8.5). The only types of fluid given were NS, LR, and albumin. There were 92 patients who received NS, 55 patients received LR, and 5 patients received albumin. The mean quantity of NS and LR ad- ministered in the first 24 hours were 4.8 and 2.5 L, respectively. For the primary outcome, there was no statistically or clinically significant rela- tionship between mEq of Cl administered in the first 24 hours with change in Cr from admission to discharge (Table 2). Scatter plots of the relationship between quantity of Cl and total volume of fluids versus change in Cr are represented in Figs. 1 and 2. For the secondary out- comes of the relationship between quantity of Cl and total fluid volume with need for HD/RRT, LOS N 14 days, mortality, and organ dysfunction, there was also no statistically significant relationship (Tables 2 and 3). Scatter plot representations of total Cl mEq administered in the first 24 hours by Cr are shown in Figs. 3 and 4. For patients with baseline el- evated Cr, there was no relationship between quantity of Cl (? =

-0.0001; 95% CI, -0.0006 to 0.0004; P = .81) or total fluid volume in

the first 24 hours (? = -0.001; 95% CI, -0.0001 to 0.0001; P = .73) with change in Cr from admission to discharge. We examined the same relationships for change in Cr from admission to 24 hours and also found no statistically significant relationship.

Discussion

In this retrospective extension of a prospective cohort study, we showed that the quantity of Cl and resuscitative fluids administered to patients with severe sepsis or septic shock in the first 24 hours has no significant effect on kidney function. Even after controlling for baseline kidney function, there was no difference in kidney function, need for RRT, LOS, mortality, or organ dysfunction.

Several studies have made physiologic arguments against the use of Cl-rich fluids for resuscitation due to several factors including hyperchloremic metabolic acidosis, fluid retention, and decreased urine output as compared to Cl-poor solutions [16]. Studies of critically ill patients showing clinically relevant adverse outcomes, however, have been scarce. The most compelling study regarding the negative ef- fects of Cl-rich fluids on kidney function came from Yunos et al., which showed that the elimination of Cl-rich fluids from their ICU resulted in a reduction in mean Cr, need for RRT, and overall kidney injury [13]. In contrast to our study of acutely septic patient presenting to the ED, they enrolled critically ill ICU patients. Patients admitted to the ICU are not always in the resuscitative phase of their treatment and may re- ceive a slower rate of fluids over a longer period. Our patients received large volumes of intravenous fluids in the first 24 hours, which did not prove to be detrimental to kidney function [13].

Three randomized trials comparing NS to lower Cl solutions in the operative setting have been performed. The first study, by Young et al., was a randomized, double-blind trial comparing NS to Plasma-Lyte A in adult trauma patients requiring emergency surgery [28]. This small

Table 2

Results of linear regression displaying slope (?), t, R², F, and P values for mEq of Cl infused over the first 24 hours and for total volume of fluids administered in the first 24 hours

Slope (?) for baseline ETCO₂ t R² F P

Quantity (mEq) of Cl first 24 h

Delta Cr from admission to discharge -0.0001 -0.86 0.92 492 .39 Volume of fluids (mL) first 24 h

Delta Cr from admission to discharge -0.0001 -1.03 0.92 494.15 .31

Fig. 1. Scatter plot of mEq of chloride in the first 24 hours vs admission-discharge creatinine.

study included 46 patients, and found an improvement in base excess, arterial pH, and serum chloride at 24 hours with Plasma-Lyte A as com- pared to LR. However, there was no difference in 24-hour urine output, resource utilization, or mortality. Another randomized, double-blinded study compared NS to LR in 66 patients (33 in each arm) undergoing surgery for Abdominal aortic aneurysms and found that patients devel- oped hyperchloremic acidosis requiring increased bicarbonate therapy (this was administered for Base deficit N 5) and that patients required a statistically significant increase in amount of blood products [29]. There was no difference in duration of mechanical ventilation, length of ICU stay, hospital stay, or complications. Finally, a study by O’Malley et al. compared NS to LR in patients undergoing renal transplant [30]. They found no difference in the primary outcome, which was Cr concen- tration at day 3, but the study was stopped early because 5 patients in the NS group experienced hyperkalemia greater than 6 mEq/L requiring treatment, versus 0 in the LR group. There were a total of 50 patients in this study, 25 patients in each group. In summary, these randomized tri- als all had small sample sizes and were performed on different patient populations than ours, but essentially found no differences in important clinical outcomes with the exception of blood products in one study and hyperkalemia in another. However, given the few patients who experi- enced these outcomes, no conclusions should be made from such a small number of patients. Although these studies are frequently quoted

Fig. 2. Scatter plot of total fluid volume in the first 24 hours vs admission-discharge creatinine.

Table 3 Results of logistic regression displaying for mEq of Cl infused over the first 24 hours and for total volume of fluids administered in the first 24 hours

Variable	Odds ratio	95% CI	P
Quantity (mEq) of Cl first 24 h
HD or RRT	0.999	0.999-1.0001	.77
LOS N 14 d	1.000	0.999-1.000	.68
Death	0.999	0.999-1.000	.88
Hepatic dysfunction	1.000	0.999-1.001	.20
Neurologic dysfunction	1.000	0.999-1.000	.66
Coagulopathy	1.000	0.999-1.000	.71
Cardiovascular dysfunction	1.000	0.999-1.000	.25
Renal dysfunction	1.000	0.999-1.000	.92
Pulmonary dysfunction	1.000	0.999-1.000	.92
Volume of fluids (mL) first 24 h HD or RRT	0.999	0.999-1.000	.69
L N 14 d	1.000	0.999-1.000	.69
Death	0.999	0.999-1.000	.84
Hepatic dysfunction	1.000	0.999-1.000	.14
Neurologic dysfunction	1.000	0.999-1.000	.50
Coagulopathy	1.000	0.999-1.000	.87
Cardiovascular dysfunction	1.000	0.999-1.000	.19
Renal dysfunction	1.000	0.999-1.000	.98
Pulmonary dysfunction	1.000	0.999-1.000	.97

to argue against the use of NS, we feel that they underscore the results of our study, which found no relationship between Cl content and kidney function despite the large volumes of NS our patients received.

Fluid overload has been shown in several studies to worsen kidney function [18-23]. However, most studies of kidney function have taken place in the ICU setting, where patients are administered intrave- nous fluids over much longer periods. Although patients in our study re- ceived a mean of 7.2 L of intravenous fluids, we again found no relationship between quantity of fluids during early resuscitation with kidney function or several other important outcomes.

To our knowledge, there are no other studies evaluating the associ- ation between quantity of Cl or total fluid volume administered during the first 24 hours with kidney function in patients with severe sepsis or septic shock. Knowing whether the Cl content of bolus intravenous fluids influences kidney function is vital to managing severely septic pa- tients because these patients necessitate a large volume of fluids to re- store tissue perfusion. From our study, the quantity of Cl-rich fluids administered during the first 24 hours of resuscitation was not found to be associated with adverse outcomes in patients with sepsis.

Our study had several limitations. First, this was a retrospective study. Although patients were enrolled prospectively for another study, the quantity and type of fluids administered were obtained via chart review. It is possible that during fluid resuscitation, additional un- documented fluids were given of which we were unaware. We expect this error to be minimal considering the strict documentation required by UF Health electronic medical record systems. Second, we did not doc- ument maintenance fluids because we sought to evaluate the effect of resuscitative fluids during the early period. We reasoned that the slow rate at which these fluids are given would not result in clinically mean- ingful increases in Cl or fluid volume. Third, we also did not take into ac- count other factors that could negatively affect kidney function in patients with severe sepsis. For example, the use of nephrotoxic drugs (particularly antibiotics) was not reviewed for the purpose of this study. Although this may have played a minor role, we felt it outside of the scope of this study to evaluate every potentially nephrotoxic agent that was administered. Fourth, this study focused only on patients with severe sepsis or septic shock. Although it may be tempting to ex- trapolate our data to patients with other types of shock, we would cau- tion the reader from doing so because sepsis is the most common cause of AKI, and almost half of our patients (48%) began the study with kid- ney dysfunction. In addition, the pathophysiology of AKI due to sepsis

Fig. 3. Scatter plot of mEq of chloride in the first 24 hours for patients with creatinine N 2 mg/dL.

may be different from other causes of AKI. We believe, however, that the prevalence of AKI in our patients strengthens our findings because de- spite the fact that many of our patients began the study with abnormal kidney function, we still found no relationship. Finally, our study was also limited by its small sample size.

Conclusion

On the basis of the documented study results, no association was found between quantity of Cl or fluids administered in the first 24 hours with kidney function in patients with severe sepsis or septic shock. We also found no effect on several clinical outcomes including need for HD or RRT, LOS, and mortality. We conclude that the use of chloride-rich solutions in the first 24 hours is not an unsafe practice and may be useful for restoring tissue perfusion in critically ill septic pa- tients. However, further prospective research is needed in this area to definitively answer the question.

Fig. 4. Scatter plot of mEq of chloride in the first 24 hours for patients with creati- nine b2 mg/dL.

References

1. Poukkanen M, Vaara ST, Pettila V, Kaukonen KM, Korhonen AM, Hovilehto S, et al. Acute kidney injury in patients with severe sepsis in Finnish intensive care units. Acta Anaesthesiol Scand 2013;57:863-72.
2. Lopes JA, Jorge S, Resina C, Santos C, Pereira A, Neves J, et al. Acute kidney injury in patients with sepsis: a contemporary analysis. Int J Infect Dis 2009;13:176-81.
3. Bagshaw SM, George C, Bellomo R. ANZICS Database Management Committee. Early acute kidney injury and sepsis: a multicentre evaluation. Crit Care 2008;12:R47 [1-9].
4. Plataki M, Kashani K, Cabello-Garza J, Maldonado F, Kashyap R, Kor DJ, et al. Predic- tors of acute kidney injury in septic shock patients: an observational cohort study. Clin J Am Soc Nephrol 2011;6:1744-51.
5. Yegenaga I, Tuglular S, Ari E, Etiler N, Baykara N, Torlak S, et al. Evaluation of sepsis/sys- temic inflammatory response syndrome, acute kidney injury, and RIFLE criteria in two tertiary hospital intensive care units in Turkey. Nephron Clin Pract 2010;115:c276-82.
6. 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. Intensive Care Med 2013;39:165-228.
7. Cohen SD, Kimmel PL. Long-term sequelae of acute kidney injury in the ICU. Curr Opin Crit Care 2012;18:623-8.
8. Zarjou A, Agarwal A. Sepsis and acute kidney injury. J Am Soc Nephrol 2011;22:999-1006.
9. Molitoris BA, Okusa MD, Palevsky PM, Chawla LS, Kaufman JS, Devarajan P, et al. De- sign of clinical trials in AKI: a report from an NIDDK workshop. Trials of patients with sepsis and in selected hospital settings. Clin J Am Soc Nephrol 2012;7:856-60.
10. Bagshaw SM, George C, Bellomo R. ANZICS Database Management Committe. A comparison of the RIFLE and AKIN criteria for acute kidney injury in critically ill pa- tients. Nephrol Dial Transplant 2008;23:1569-74.
11. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early Goal- Directed Therapy Collaborative Group. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368-77.
12. 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;303:739-46.
13. Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA 2012;308:1566-72.
14. Wilcox CS. Regulation of renal blood flow by plasma chloride. J Clin Invest 1983;71: 726-35.
15. Chowdhury AH, Cox EF, Francis ST, Lobo DN. A randomized, controlled, double-blind crossover study on the effects of 2-L infusions of 0.9% saline and plasma-lyte(R) 148 on renal blood flow velocity and renal cortical tissue perfusion in healthy volunteers. Ann Surg 2012;256:18-24.
16. Lobo DN, Awad S. Should chloride-rich crystalloids remain the mainstay of fluid resusci- tation to prevent ‘pre-renal’ acute kidney injury?: con. Kidney Int 2014;86(6):1096-105.
17. Shujaat A, Bajwa AA. Optimization of preload in severe sepsis and septic shock. Crit Care Res Pract 2012;2012:761051 [14].
18. Boyd JH, Forbes J, Nakada TA, Walley KR, Russell JA. Fluid resuscitation in septic shock: a positive fluid balance and elevated central venous pressure are associated with increased mortality. Crit Care Med 2011;39:259-65.
19. Sutherland SM, Zappitelli M, Alexander SR, Chua AN, Brophy PD, Bunchman TE, et al. Fluid overload and mortality in children receiving continuous renal replacement therapy: the prospective pediatric Continuous renal replacement therapy registry. Am J Kidney Dis 2010;55:316-25.
20. Grams ME, Estrella MM, Coresh J, Brower RG, Liu KD. National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome Network. Fluid balance, diuret- ic use, and mortality in acute kidney injury. Clin J Am Soc Nephrol 2011;6:966-73.
21. Bouchard J, Soroko SB, Chertow GM, Himmelfarb J, Ikizler TA, Paganini EP, et al. Pro- gram to Improve Care in Acute Renal Disease (PICARD) Study Group. Fluid accumu- lation, survival and recovery of kidney function in critically ill patients with acute kidney injury. Kidney Int 2009;76:422-7.
22. National Heart, Lung, Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network, Wiedemann HP, Wheeler AP, Bernard GR, Thompson BT, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006;354:2564-75.
23. Heung M, Wolfgram DF, Kommareddi M, Hu Y, Song PX, Ojo AO. Fluid overload at initiation of renal replacement therapy is associated with lack of renal recovery in patients with acute kidney injury. Nephrol Dial Transplant 2012;27:956-61.
24. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emer- gency medicine research: where are the methods? Ann Emerg Med 1996;27:305-8.
25. Lowenstein SR. Medical record reviews in emergency medicine: the blessing and the curse. Ann Emerg Med 2005;45:452-5.
26. Shapiro NI, Wolfe RE, Moore RB, Smith E, Burdick E, Bates DW. Mortality in Emer- gency Department Sepsis (MEDS) score: a prospectively derived and validated clin- ical prediction rule. Crit Care Med 2003;31:670-5.
27. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)-a metadata-driven methodology and workflow process for pro- viding translational research informatics support. J Biomed Inform 2009;42:377-81.
28. Young JB, Utter GH, Schermer CR, Galante JM, Phan HH, Yang Y, et al. Saline versus Plasma-Lyte A in Initial resuscitation of trauma patients: a randomized trial. Ann Surg 2014;259:255-62.
29. Waters JH, Gottlieb A, Schoenwald P, Popovich MJ, Sprung J, Nelson DR. Normal saline ver- sus lactated Ringer’s solution for intraoperative fluid management in patients undergoing abdominal aortic aneurysm repair: an outcome study. Anesth Analg 2001;93:817-22.
30. O’Malley CM, Frumento RJ, Hardy MA, Benvenisty AI, Brentjens TE, Mercer JS, et al. A randomized, double-blind comparison of lactated Ringer’s solution and 0.9% NaCl during renal transplantation. Anesth Analg 2005;100:1518-24 [table of contents].