Objectives: Excessive fluid therapy in patients with sepsis may be associated with risks that outweigh any benefit. We investigated the possible influence of early fluid balance on outcome in a large international database of ICU patients with sepsis.
Design: Observational cohort study.
Setting: Seven hundred and thirty ICUs in 84 countries.
Patients: All adult patients admitted between May 8 and May 18, 2012, except admissions for routine postoperative surveillance. For this analysis, we included only the 1,808 patients with an admission diagnosis of sepsis. Patients were stratified according to quartiles of cumulative fluid balance 24 hours and 3 days after ICU admission.
Measurements and Main Results: ICU and hospital mortality rates were 27.6% and 37.3%, respectively. The cumulative fluid balance increased from 1,217 mL (–90 to 2,783 mL) in the first 24 hours after ICU admission to 1,794 mL (–951 to 5,108 mL) on day 3 and decreased thereafter. The cumulative fluid intake was similar in survivors and nonsurvivors, but fluid balance was less positive in survivors because of higher fluid output in these patients. Fluid balances became negative after the third ICU day in survivors but remained positive in nonsurvivors. After adjustment for possible confounders in multivariable analysis, the 24-hour cumulative fluid balance was not associated with an increased hazard of 28-day in-hospital death. However, there was a stepwise increase in the hazard of death with higher quartiles of 3-day cumulative fluid balance in the whole population and after stratification according to the presence of septic shock.
Conclusions: In this large cohort of patients with sepsis, higher cumulative fluid balance at day 3 but not in the first 24 hours after ICU admission was independently associated with an increase in the hazard of death.
1Department of Anesthesiology and Intensive Care, Uniklinikum Jena, Jena, Germany.
2Department of Anaesthesia, Intensive Care and Acute Poisioning, Pomeranian Medical University, Szczecin, Poland.
3Department of Medicine, Medical College of Wisconsin, Milwaukee, WI.
4Department of Critical Care, Care Institute of Medical Sciences, Ahmedabad, India.
5Department of Intensive Care Medicine, University Medical Center, Hamburg Eppendorf, Germany.
6Center for Trauma and Critical Care, George Washington University Hospital, Washington, DC.
7Interdepartmental Division of Critical Care Medicine, Department of Surgery, University of Toronto, St. Michael’s Hospital, Toronto, ON, Canada.
8Department of Intensive Care, Erasme University Hospital, Université Libre de Bruxelles, Brussels, Belgium.
*See also p. 555.
Drs. Sakr and Rubatto Birri designed the study, extracted, and analyzed the data and drafted the article. Drs. Sakr, Kotfis, Nanchal, Shah, Kluge, Schroeder, Marshall, and Vincent participated in the original Intensive Care Over Nations study and reviewed the article for critical content. All authors read and approved the final article.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/ccmjournal).
Additional members of the Care Over Nations Investigators are listed in Appendix 1 (Supplemental Digital Content 1, http://links.lww.com/CCM/C246).
Dr. Marshall has received fees from AKPA Pharma (data safety management board) and Regeneron (consultancy), as well as grant support from the CIHR and Physicians Services Incorporated Foundation. The remaining authors have disclosed that they do not have any potential conflicts of interest.
For information regarding this article, E-mail: email@example.com
Sepsis frequently leads to death in ICU patients (1–3), with most deaths occurring as a result of cardiovascular or multiple organ failure (4). Fluid resuscitation is an essential component of treatment for patients with sepsis, but optimal hemodynamic targets and strategies are difficult to define. Preload optimization often necessitates administration of substantial amounts of fluid to compensate for the relative hypovolemia induced by generalized vasodilatation and increased capillary leak. But generous fluid administration can have deleterious effects, including tissue edema. Although positive fluid balance has been associated with a higher risk of death in septic and other populations of critically ill patients (5–11), optimal fluid balance targets beyond the initial resuscitation period remain unclear.
The aim of this study was to investigate possible associations between fluid balance and outcome in patients with sepsis using data from a large international database of ICU patients. We hypothesized that positive fluid balance after the initial 24-hour resuscitation period would be independently associated with mortality.
This was a planned substudy of the Intensive Care Over Nations (ICON), a multicenter, worldwide audit. Full details of the methodology have been provided previously (3), and a list of participating ICUs is provided in Appendix 1 (Supplemental Digital Content 1, http://links.lww.com/CCM/C246). Institutional recruitment for participation was by open invitation and was voluntary, with no financial incentive. Institutional review board approval was obtained by the participating institutions according to local ethical regulations. Informed consent was not required because of the observational and anonymous nature of data collection. The ICON audit included all 10,069 adult patients (more than 16 yr old) admitted to the participating centers between May 8 and May 18, 2012. The organizational characteristics of these centers are shown in Table S1 (Supplemental Digital Content 1, http://links.lww.com/CCM/C246). For the purposes of the current analysis, only those patients with a diagnosis of sepsis at admission to the ICU were considered (n = 1,808). Patients were followed up until death, hospital discharge, or for 60 days.
Data were collected prospectively and were electronically introduced by the investigators using an internet-based website. Data collection at admission included demographic data and comorbid diseases. Clinical and laboratory data for Simplified Acute Physiology Score (SAPS) II (12) and Acute Physiology and Chronic Health Evaluation II (13) scores were recorded as the worst values within 24 hours after admission. Microbiologic and clinical infections were recorded daily, as well as the antibiotics administered. Daily evaluation of organ function according to the Sequential Organ Failure Assessment (SOFA) score (14) was performed. Data were collected for a maximum of 28 days in the ICU.
Full details of the definitions have been published elsewhere (3). Infection was defined according to the International Sepsis Forum definitions (15). Only infections requiring administration of antimicrobial agents were considered. Sepsis was defined as the presence of infection with the concomitant occurrence of at least one organ failure, defined as a SOFA score of more than 2 for the organ in question. Septic shock was defined as sepsis associated with cardiovascular failure (cardiovascular SOFA score > 2). Lactate levels were not considered in the diagnosis of septic shock as they were not expected to be available in all patients. Surgical admissions were defined as patients who had had surgery in the 4 weeks preceding admission.
Calculation of Fluid Balance
Fluid balance was calculated from the daily fluid intake (enteral or parenteral) and output (urinary output and other fluid losses, including drainage fluids and extracorporeal fluid elimination) recorded every 24 hours after ICU admission. Insensible fluid loss was not recorded and hence not considered in the calculation of fluid balance. The cumulative fluid balances in the first 24 hours, 3 days, and 7 days of the ICU stay were calculated. Patients were stratified according to quartiles of cumulative fluid balance at 24 hours and at 3 days after ICU admission to investigate the possible influence of early fluid balance on outcome in these patients. In patients who died or were discharged before 3 days in the ICU, the cumulative fluid balance during the ICU stay was considered as the 3-day fluid balance.
Data Management and Quality Control
Detailed instructions, explaining the aim of the study, data collection, and definitions for various items were available through a secure website for all participants. Validity checks were made concurrent with data entry on the electronic case record form, including plausibility checks within each variable and between variables. Data were further reviewed by the coordinating center for plausibility and availability of the outcome variables, and doubts were clarified with the corresponding ICU. We performed no other supplementary quality control measures.
Data are summarized using means and SD, medians and interquartile ranges (IQRs), or numbers and percentages. The Kolmogorov-Smirnov test was used, and histograms and normal-quantile plots were examined to verify if there were significant deviations from the normality assumption of continuous variables. Difference testing between groups was performed using analysis of variance, Kruskal-Wallis test, Student’s t test, Mann-Whitney test, chi-square test, or Fisher exact test as appropriate.
To determine the adjusted relative risk of hospital death, right censored at 28 days, according to quartiles of cumulative fluid balance at 24 and 72 hours following admission to the ICU, we developed a multivariable Cox proportional hazard model in the overall population and in patients stratified according to the presence of septic shock. The categories of fluid balance were included as categorical variables with the first quartile as the reference category. Full details of this analysis are given in Appendix 2 (Supplemental Digital Content 1, http://links.lww.com/CCM/C246).
Data were analyzed using IBM SPSS Statistics software, version 20 for Windows (IBM, Armonk, NY). All reported p values are two sided, and a p value less than 0.05 was considered to indicate statistical significance.
Of the 1,808 patients with sepsis at admission to the ICU, 1,098 (60.7%) had septic shock. The baseline characteristics of these patients are summarized in Table 1. The overall ICU and hospital mortality rates were 27.6% (n = 491) and 37.3% (n = 646), respectively, and the median ICU and hospital lengths of stay were 5 days (IQR, 2–10 d) and 13 days (IQR, 6–27 d).
Fluid intakes in the whole cohort were 3,325 (2,028–4,932), 9,399 (5,425–13,614), and 12,595 (6,065–20,673) mL at 24 hours, 3 days, and 7 days after ICU admission, respectively. The cumulative fluid balance increased from 1,217 mL (–90 to 2,783 mL) in the first 24 hours after ICU admission to 1,794 mL (–951 to 5,108 mL) mL at 3 days and decreased thereafter to reach 1,453 mL (–2,173 to 5,548 mL) at 7 days following ICU admission.
The cumulative fluid intake was similar in survivors and nonsurvivors. However, fluid output was significantly less in nonsurvivors leading to a more positive fluid balance in these patients. Nonsurvivors were more likely to have received synthetic colloid solutions, including hydroxyethyl starch and gelatin, than survivors and received more vasopressor/inotropic support (Table S2, Supplemental Digital Content 1, http://links.lww.com/CCM/C246). Fluid balance remained positive over time in the nonsurvivors but became negative in the survivors after the third day of the ICU stay (Fig. 1). Patients with septic shock had greater fluid intake during the first 4 days in the ICU and more positive fluid balance during the first 3 days compared with those without shock (Fig. S1, Supplemental Digital Content 1, http://links.lww.com/CCM/C246; and Fig. S2, Supplemental Digital Content 1, http://links.lww.com/CCM/C246).
Characteristics and Outcome According to the Cumulative Fluid Balance at 24 Hours and 3 Days After ICU Admission
The characteristics of patients at ICU admission, stratified according to quartiles of 24- and 72-hour cumulative fluid balance, are presented in Table 1 and Table S3 (Supplemental Digital Content 1, http://links.lww.com/CCM/C246). SAPS II and SOFA scores at admission to the ICU, the degree of organ failure during the ICU stay as assessed by the SOFA maximum and SOFA mean, and the need for renal replacement therapy and vasopressors and inotropes during the ICU stay increased stepwise with increasing quartiles of 24-hour cumulative fluid balance (Table S4, Supplemental Digital Content 1, http://links.lww.com/CCM/C246). Although ICU and hospital lengths of stay were similar among the quartiles of 24-hour fluid balance, there was a stepwise increase in ICU and hospital mortality rates within increasing quartiles (Table S4, Supplemental Digital Content 1, http://links.lww.com/CCM/C246; and Fig. S3, Supplemental Digital Content 1, http://links.lww.com/CCM/C246).
Similar patterns were observed with the 3-day quartiles (Table 2), with ICU and hospital mortality rates increasing more than two-fold from the lowest to the highest quartile. The hospital length of stay was shorter in the higher quartiles than in the lowest quartile (Table 2).
After adjustment for possible confounders in a multivariable Cox proportional hazard analysis, cumulative 24-hour fluid balance was not associated with an increased hazard of death in the whole cohort or in patients with septic shock (Table 3; Fig. 2; and Fig. S4, Supplemental Digital Content 1, http://links.lww.com/CCM/C246). However, there was a stepwise increase in the hazard of death with increasing 3-day cumulative fluid balance quartile.
The main findings of our study are as follows: 1) fluid balance was more positive in nonsurvivors than in survivors despite similar fluid intakes; 2) these differences were more marked on the third day following admission; and 3) after adjustment for possible confounders, there was a stepwise increase in the hazard of death with increasing 3-day cumulative fluid balance quartile but not with 24-hour quartiles.
Optimal targets for fluid resuscitation in patients with sepsis are not fully established and probably require the integration of multiple variables. In our study, differences in fluid balance were primarily the result of lower fluid output in nonsurvivors than in survivors; hence, the inability to excrete excess fluid may be an important factor. Some experimental data have suggested that high-volume resuscitation in septic animals may increase mortality (16). Several retrospective clinical studies support this hypothesis (1, 8, 17). Shum et al (17) showed that in critically ill patients who stayed in the ICU for 3 days or more, fluid balance on the second plus third ICU days and total fluid balance during the ICU stay were positively associated with hospital death, whereas a positive fluid balance on the first ICU day was negatively associated with hospital mortality. This study, however, included a relatively small number of patients and was not confined to patients with sepsis. In a large database of critically ill patients admitted to ICUs in 24 European countries, we previously reported that cumulative fluid balance within the first 72 hours following the onset of sepsis was an independent risk factor for ICU death (1). More recently, a post hoc analysis of the Vasopressin in Septic Shock Trial showed that a more positive fluid balance, both early in resuscitation and cumulatively over 4 days, was associated with an increased risk of mortality in septic shock (8). Acheampong and Vincent (6) also recently showed, in a prospective observational study, that positive fluid balance was independently associated with higher ICU mortality.
From our data, we cannot determine the mechanisms underlying the observed association between fluid balance and outcome in patients with septic shock. The impact of renal function on the results was considered in the multivariable analysis by adjustment for the SOFA renal subscores, so that renal function may not be the only factor explaining the association between fluid balance and outcome in these patients. We can speculate that excess fluid administration may lead to increased tissue edema and worsening of organ function. These effects may be more pronounced in encapsulated organs, such as the kidney and liver, which may not be able to accommodate excess volume without an increase in interstitial pressure, thereby compromising organ blood flow (18). In patients with acute lung injury, the Fluids and Catheters Treatment Trial study compared restrictive and liberal fluid management strategies (19) and showed no significant differences in mortality, but the conservative strategy of fluid management was associated with improved lung function and reduced duration of mechanical ventilation without increasing nonpulmonary organ failure. The inability to remove excess fluids may, therefore, play an important role in determining outcomes in critically ill patients, including those with sepsis, lung injury, and acute renal failure.
In our study, higher quartiles of cumulative fluid balance at 3 days after ICU admission but not at 24 hours were associated with a stepwise increase in the hazard of death. This finding suggests that accumulation of excess fluid beyond the initial resuscitation stage may be mechanistically linked to worse outcomes. It is possible that a cutoff value of fluid balance exists beyond which worse outcomes become apparent as evidenced by similar hazards of death in our 24-hour quartiles when positive fluid balance, even in the highest quartile, was less marked than on day 3 or 7. Boyd et al (8) reported that lower fluid balance as early as 12 hours following septic shock was independently associated with a lower 28-day hazard of survival. As our data were collected only every 24 hours, it is not possible to directly compare our results with those of Boyd et al (8) in this regard. Micek et al (20) in a retrospective analysis of 325 patients with septic shock found that nonsurvivors had a larger net fluid balance at 24 hours and 8 days after the onset of septic shock, and when using different quartiles, the highest quartile of net fluid balance on day 8 was associated with greater mortality in the multivariable analysis. Sadaka et al (21) retrospectively collected data from 350 patients with septic shock in 56 medical-surgical ICUs. They divided patients into groups based on the degree of excess fluid balance at 24 hours and observed that in-hospital mortality increased significantly with a higher degree of positive fluid balance. Taken together, these results support the four-phase salvage, optimization, stabilization, and deescalation approach to fluid administration (22, 23), suggesting that a more restrictive approach to fluid administration may be safe after initial resuscitation of septic patients.
The strengths of the current study are the large number of patients and variables included, allowing adjustment for a large number of factors. We also acknowledge some limitations. First, although we adjusted for severity of illness, organ failures, and other variables, we cannot discount the possible influence of unmeasured factors. Second, data collection was restricted to every 24 hours, so we could not control for the influence of patterns of fluid therapy during the early hours after sepsis onset, when they are perhaps most intense. We also did not collect data on fluid administration prior to ICU admission. Increased awareness of sepsis may have led to earlier fluid resuscitation prior to referral to the ICU and may explain the relatively low amount of fluid given to patients with sepsis in our cohort. Third, although we controlled for the use of colloid solutions, we do not have information on the choice of crystalloid solutions for resuscitation. The use of hyperchloremic crystalloid solutions can affect outcomes (24) and may have confounded our results. Fourth, data regarding hemodynamic targets for resuscitation, fluid responsiveness, adequacy of tissue perfusion, reasons for accumulation of fluid, diuretic use, and attempts to reduce fluid accumulation are unavailable. The use of vasopressor agents may also have influenced fluid resuscitation and balance. Nonetheless, the degree of cardiovascular dysfunction and vasopressor support, as assessed by the SOFA score, was considered in the multivariable adjustment. Finally, despite our demonstration of increasing hazard of death with progressive cumulative positive fluid balance, we are unable to discern whether there was a specific point at which the excess hazard developed or the optimal volume of fluid resuscitation. These are important questions that should be the target of future investigations.
In this large cohort of patients with severe sepsis, although the cumulative fluid intake was similar at 24 hours, 3 days, and 7 days after admission to the ICU, the corresponding cumulative fluid balance was lower in survivors than in nonsurvivors because of higher fluid output. After adjustment for possible confounders, higher quartiles of cumulative fluid balance at 3 days after ICU admission but not at 24 hours were associated with an increase in the hazard of death in the whole population and after stratification according to the presence of septic shock. Because of the retrospective nature of the analysis, we cannot elaborate on whether limiting fluid intake or enforcing fluid output would be the most effective approach to decrease fluid balance, probably a combination of both approaches is needed. These data are hypothesis generating and may provide a good basis for randomized controlled trials to investigate the possible influence of negative fluid balance and a restrictive approach to fluid therapy after initial resuscitation on the outcome of these patients.
We thank Hassane Njimi, PhD, for his help with the statistical analyses.
1. Vincent JL, Sakr Y, Sprung CL, et al; Sepsis Occurrence in Acutely Ill Patients Investigators. Sepsis in European intensive care units: Results of the SOAP study. Crit Care Med. 2006;34:344–353.
2. Vincent JL, Rello J, Marshall J, et al; EPIC II Group of Investigators. International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009;302:2323–2329.
3. Vincent JL, Marshall JC, Namendys-Silva SA, et al; ICON Investigators. Assessment of the worldwide burden of critical illness: The Intensive Care Over Nations (ICON) audit. Lancet Respir Med. 2014;2:380–386.
4. Sakr Y, Lobo SM, Moreno RP, et al; SOAP Investigators. Patterns and early evolution of organ failure in the intensive care unit and their relation to outcome. Crit Care. 2012;16:R222.
5. Alsous F, Khamiees M, DeGirolamo A, et al. Negative fluid balance predicts survival in patients with septic shock: A retrospective pilot study. Chest. 2000;117:1749–1754.
6. Acheampong A, Vincent JL. A positive fluid balance is an independent prognostic factor in patients with sepsis. Crit Care. 2015;19:251.
7. de Oliveira FS, Freitas FG, Ferreira EM, et al. Positive fluid balance as a prognostic factor for mortality and acute kidney injury in severe sepsis and septic shock. J Crit Care. 2015;30:97–101.
8. Boyd JH, Forbes J, Nakada TA, et al. Fluid resuscitation in septic shock: A positive fluid balance and elevated central venous pressure are associated with increased mortality. Crit Care Med. 2011;39:259–265.
9. Payen D, de Pont AC, Sakr Y, et al; Sepsis Occurrence in Acutely Ill Patients (SOAP) Investigators. A positive fluid balance is associated with a worse outcome in patients with acute renal failure. Crit Care. 2008;12:R74.
10. Bouchard J, Soroko SB, Chertow GM, et al; Program to Improve Care in Acute Renal Disease (PICARD) Study Group. Fluid accumulation, survival and recovery of kidney function in critically ill patients with acute kidney injury. Kidney Int. 2009;76:422–427.
11. Sakr Y, Vincent JL, Reinhart K, et al; Sepsis Occurrence in Acutely Ill Patients Investigators. High tidal volume and positive fluid balance are associated with worse outcome in acute lung injury. Chest. 2005;128:3098–3108.
12. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA. 1993;270:2957–2963.
13. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: A severity of disease classification system. Crit Care Med. 1985;13:818–829.
14. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med. 1996;22:707–710.
15. Calandra T, Cohen J. The international sepsis forum consensus conference on definitions of infection in the intensive care unit. Crit Care Med. 2005;33:1538–1548.
16. Brandt S, Regueira T, Bracht H, et al. Effect of fluid resuscitation on mortality and organ function in experimental sepsis models. Crit Care. 2009;13:R186.
17. Shum HP, Lee FM, Chan KC, et al. Interaction between fluid balance and disease severity on patient outcome in the critically ill. J Crit Care. 2011;26:613–619.
18. Prowle JR, Echeverri JE, Ligabo EV, et al. Fluid balance and acute kidney injury. Nat Rev Nephrol. 2010;6:107–115.
19. Wiedemann HP, Wheeler AP, Bernard GR, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med. 2006;354:2564–2575.
20. Micek ST, McEvoy C, McKenzie M, et al. Fluid balance and cardiac function in septic shock as predictors of hospital mortality. Crit Care. 2013;17:R246.
21. Sadaka F, Juarez M, Naydenov S, et al. Fluid resuscitation in septic shock: The effect of increasing fluid balance on mortality. J Intensive Care Med. 2014;29:213–217.
22. Hoste EA, Maitland K, Brudney CS, et al; ADQI XII Investigators Group. Four phases of intravenous fluid therapy: A conceptual model. Br J Anaesth. 2014;113:740–747.
23. Vincent JL, De Backer D. Circulatory shock. N Engl J Med. 2013;369:1726–1734.
24. Shaw AD, Schermer CR, Lobo DN, et al. Impact of intravenous fluid composition on outcomes in patients with systemic inflammatory response syndrome. Crit Care. 2015;19:334.
fluid administration; fluid output; outcome; septic shock
Supplemental Digital Content
Copyright © by 2017 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.