Effect of Initial Infusion Doses of Fluid Resuscitation on Prognosis in Patients with Septic Shock: A Prospective Multicentre Observational Study

OBJECTIVES The 2018 Surviving Sepsis Campaign (SSC) recommends rapid administration of 30 ml/kg crystalloid for hypotension or lactate ≥ 4 mmol/L in patients with septic shock; however, there is no credible evidence to support this recommendation. The purpose of this study was to examine the relationship between initial uid resuscitation doses and prognosis in patients with septic shock. QC: acquisition of data, analysis and interpretation of data, and critical revision of the manuscript for important intellectual content. WL: data analysis and interpretation and ensured that the accuracy and integrity of all work was appropriately maintained. FW: experimental conception and design, data acquisition, rst drafting of the manuscript, and full access to all the data in the study. HC: data acquisition, analysis and interpretation; revision of the manuscript. JS: data acquisition and analysis of data; statistical analysis and revision of the manuscript. RZ: data acquisition and interpretation of data; administrative and technical aspects. HW: data acquisition and interpretation of data; study supervision; and critical revision of the manuscript. All authors read and approved the nal manuscript.

Septic shock is the major cause of death in intensive care units (ICUs) [1]. The published rates of inhospital mortality caused by septic shock are approximately 30-50% [2][3]. Similar to other severe diseases, early identi cation and appropriate management in the initial hours after the development of septic shock can improve patient outcomes [4]. Initial uid resuscitation for septic shock can restore tissue perfusion before the onset of irreversible tissue damage and prevent organ failure and death [5][6]. Therefore, appropriate initial uid resuscitation within the rst 3 h of septic shock is strongly recommended by the Surviving Sepsis Campaign (SSC) guidelines [7] as the cornerstone of septic shock treatment [8].
The initial infusion dose for patients with septic shock is based on limited evidence. Lee et al showed that septic shock patients who received a larger volume of uid in the rst 3 h were more likely to survive [9].
Another study [10] on critical care published this year showed that an initial uid resuscitation rate of 0.25-0.50 ml/kg/min may be associated with early shock reversal and lower 28-day mortality in septic shock patients. Nonetheless, a few trials have demonstrated increased mortality with uid resuscitation [11][12].
The 2018 SSC guidelines proposed the hour-1 bundle and recommended rapid administration of 30 ml/kg crystalloid for hypotension or lactate ≥ 4 mmol/L in patients with septic shock [13]. However, the guidelines did not specify the uid resuscitation dose within the rst hour or the completion time of the initial 30 ml/kg uid resuscitation in these patients. The purpose of this study was to examine the relationship between initial uid resuscitation doses and prognosis in patients with septic shock.

Ethical approval
The study was approved by the Institutional Review Board of Northern Jiangsu People's Hospital (2017KY-021) and was registered in the Chinese Clinical Trial Registry (Registration Number: ChiCTR-OOC-17013223). Informed consent was obtained from each participant prior to his or her enrolment in this study.

Population study and design
This was a multicentre prospective observational study of adult patients with septic shock admitted to four intensive care units (ICUs) with more than 100 beds in a total in three Jiangsu province teaching hospitals, including Northern Jiangsu People's Hospital, Jiangdu People's Hospital of Yangzhou, A liated Hospital of Yangzhou University and A liated Hospital of Yangzhou University, over a 1-year span from May 8, 2018, to June 31, 2020. The patients enrolled were categorized into the following groups according to the initial infusion dose of uid resuscitation: below 20 ml/kg uid, 20-30 ml/kg uid and above 30 ml/kg uid. Patients younger than 18 years of age or with any of the following primary conditions were excluded: pregnancy, trauma, epilepsy, cardiogenic pulmonary oedema, stroke, or active bleeding. Furthermore, patients whose eventual outcome was unknown or ICU length of stay was less than 72 hours were excluded.

De nitions
According to the Sepsis-3 categories [14][15], sepsis and septic shock are de ned as follows. Sepsis is identi ed as an acute change in the total sequential organ failure (SOFA) assessment score of more than 2 points caused by infection. The baseline SOFA score is assumed to be zero in patients not known to have pre-existing organ dysfunction. Septic shock is de ned as a clinical construct of sepsis if the patient has persistent hypotension that requires vasopressors to maintain a mean arterial blood pressure (MAP) greater than 65 mmHg and if the serum lactate level is greater than 2 mmol/L despite initial uid resuscitation.
The initial infusion dose of uid resuscitation (ml/kg) was calculated as the total uid resuscitation within the rst hour divided by the actual body weight on admission. As mentioned above, the patients were categorized into three groups based on the initial infusion dose of uid resuscitation: below 20 ml/kg uid, 20-30 ml/kg uid and above 30 ml/kg uid. SSCs and the literature support of the use of 30 ml/kg of crystalloid for initial volume resuscitation among septic shock patients, though the time to complete the 30 m/kg uid resuscitation is unclear. Therefore, we further performed subgroup analysis according to the time to complete 30 ml/kg of crystalloid uid resuscitation. The difference in sequential organ failure assessment between the third day after admission and the rst day of admission (∆SOFA) was de ned as the SOFA score change from day 1 to day 3.

Data collection
We provided the principals of each research centre with important study information through emails and online training, including the protocol and answers to questions. Two data collectors collected prospective data, including age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II score, rst day and third day total SOFA scores, principal diagnosis, infection source, and initial lactate level.
The uid infusion volume per hour was recorded within 12 hours. Laboratory examinations were obtained from the electronic health database. The primary outcome of the study was 28-d mortality. Secondary outcomes included ICU length of stay, incidence of acute respiratory distress syndrome (ARDS) or acute kidney injury (AKI), respiratory support treatment radio, ventilator-free days and renal replacement therapy (RRT)-free days.

Statistical analysis
Continuous data with a normal distribution are expressed as the mean and standard deviation (±SD), and differences between groups were analysed by one-way analysis of variance. Continuous data with skewed distributions are expressed as the median and interquartile range (IQR), and differences between groups were analysed by the Kruskal-Wallis test. Dichotomous variables are reported as n (%), and differences between groups were compared using the Chi-square test (or Fisher exact test when appropriate).
A binary logistic regression model was applied to adjust for potential confounding factors in uencing a poor prognosis in patients with septic shock. In addition, curve tting was adopted to compare the relationship between the initial uid resuscitation dose and ∆SOFA score. Data analysis was performed using SPSS 22.0 statistical software; P<0.05 was considered statistically signi cant.

Results
Among 850 ICU patients with a suspected infection, 528 met the sepsis criteria according to the Sepsis-3 de nitions. In total, 316 of these patients met the septic shock criteria; 163 were excluded. The remaining 153 septic shock patients were entered into the registry. Of these patients, 39 completed a uid resuscitation dose below 20 ml/kg within the rst hour, and 68 completed a uid resuscitation dose between 20 ml/kg and 30 ml/kg; the remaining 46 patients had a uid resuscitation dose above 30 ml/kg ( Figure 1). Table 1 (Table 1).

Patient outcomes
A total of 153 septic shock patients had a 28-day mortality of 37.3%. Among them, 28-day mortality was highest in the above 30 ml/kg uid group (47.8%) and lowest in the 20-30 ml/kg uid group (26.5%, P<0.05, Figure 2). The incidence of ARDS was highest in the 30 ml/kg uid group (80.4%) and lowest in the below 20 ml/kg uid group (40.6%). Patients in the above 30 ml/kg uid group had the highest rate of invasive mechanical ventilation (67.4%) and the lowest number of mechanical ventilation-free days [24(1-28)], while patients in the 20-30 ml/kg uid group had the lowest rate of invasive mechanical ventilation (35.3%, P<0.01) and the highest number of mechanical ventilation-free days [28(12-28), P<0.05] ( Table 2).

Logistic regression analysis
Logistic regression showed that an initial liquid dose of 20-30 ml/kg was an independent protective factor with a signi cant OR for decreased mortality (OR, 0.393; 95% CI, 0.178-0.866; P<0.05, Table 3).
Initial uid resuscitation and mortality First, we explored the relationship between the initial uid resuscitation dose within 1 hour and the 28-day mortality rate in patients with septic shock. The 28-day mortality rate (100%) was highest in those who were given initial uid resuscitation below 20 ml/kg within 1 hour; however, the 28-day mortality rate was lowest in septic shock patients with 20-30 ml/kg initial uid resuscitation within 1 hour (26.5%, P<0.05) ( Figure 2). We further investigated the relationship between the time to complete 30 ml/kg initial uid resuscitation and the 28-day mortality rate. Our results showed that patients who completed 30 ml/kg initial uid resuscitation within 1-2 hours had the lowest 28-day mortality rate (25.9%) and that those who completed more than 4 hours had the highest mortality rate (100%, P<0.05) (Figure 2).

Initial uid resuscitation and ∆SOFA
In addition, we examined the relationship between initial uid resuscitation dose and ∆SOFA score in patients with septic shock, and a parabolic relationship between liquid dose within 1 hour or time to complete 30 ml/kg liquid and ∆SOFA was detected. The ∆SOFA score was the highest for initial uid resuscitation with 25.7 ml/kg within 1 hour, reaching 5.807. When the initial uid resuscitation of 30 ml/kg was completed in 2.18 hours, the ∆SOFA score reached the maximum value of 5.56 ( Figure 3).

Discussion
This study showed that an initial uid resuscitation rate of 20-30 ml/kg within the rst 1 h was associated with lower 28-day mortality and faster organ function recovery in patients with septic shock.
Moreover, septic shock patients who completed 30 ml/kg initial uid resuscitation between the rst 1-2 h had a lower 28-day mortality rate and faster organ function recovery. This nding is consistent with a recent retrospective study, which found that an initial uid resuscitation rate of 0.25-0.50 ml/kg/min may be associated with early shock reversal and lower 28-day mortality compared with slower rates of infusion [16]. In addition, this study showed that insu cient initial uid resuscitation (below 20 ml/kg within the rst 1 h) may increase 28-day mortality in these patients.
Fluid resuscitation is the cornerstone of septic shock treatment [17]. In septic shock, blood vessel dilation and vascular permeability increase, leading to relative and absolute blood volume de ciency [18]. The goal of initial uid resuscitation in septic shock is to restore blood volume, thereby increasing cardiac output and oxygen delivery [19]. This nding is consistent with previous reports that insu cient initial uid resuscitation is associated with higher 28-day mortality. Overall, a faster initial uid resuscitation rate might improve the microcirculation and tissue perfusion, resulting in improved outcomes, including SOFA score, duration of hospital, and mortality rates [20]. Therefore, 30 ml/kg initial uid resuscitation within the rst 3 h of septic shock is strongly recommended by the SSC guidelines [7]. Furthermore, the 2018 SSC guidelines recommend faster start-up and completion of 30 ml/kg initial uid resuscitation in patients with septic shock [13].
Although uid resuscitation is very important in the early treatment of septic shock, there is often insu cient initial uid resuscitation in clinical practice [21][22]. What factors may in uence the clinician to start initial uid resuscitation? First, it depends on the level of awareness and compliance of medical staff with SSC treatment guidelines. The 2018 SSC guidelines recommend rapid administration of 30 ml/kg crystalloid for hypotension or lactate ≥4 mmol/L in patients with septic shock [13], and awareness and compliance with these guidelines may affect the speed of early uid resuscitation. In the present study, patients in the below 20 ml/kg uid group had the longest time from diagnosis to ICU admission. The reason for the lack of initial uid resuscitation in these patients may be the delay in entering the ICU. Medical staff outside the ICU have insu cient awareness of the importance of initial uid resuscitation, which leads to insu cient initial uid resuscitation. In addition, patients with a higher BMI or obesity receive relatively lower uid volumes than patients without obesity [23]. This research also showed that the weight and BMI values in the low-dose liquid group were signi cantly higher than those in the highdose liquid group. Thus, the initial uid dosing strategy for septic shock should follow guidelines recommending weight-based uid administration; however, the guidelines do not clearly specify whether actual, ideal, or adjusted body weight should be used to calculate total uid volumes.
On the other hand, very fast initial uid resuscitation may increase glycocalyx shedding and negatively impact its barrier function [24]. Fluid overload causes pulmonary oedema, pulmonary interstitial oedema, and oedema of other tissues and organs, which is not conducive to oxygen diffusion, aggravates hypoxia and is closely related to poor prognosis [25]. Our study showed that 28-day mortality was highest in patients with septic shock who received greater than 30 ml/kg initial uid resuscitation within the rst hour or completion the initial 30 ml/kg uid resuscitation in less than 1 hour, which suggested that too much or too fast initial uid resuscitation may lead to poor prognosis in patients with septic shock. The study further showed that an initial uid resuscitation rate of 20-30 ml/kg within the rst 1 h or completion of the initial 30 ml/kg uid resuscitation between the rst 1-2 h may be associated with faster organ function recovery and lower 28-day mortality in patients with septic shock. This nding is consistent with a recent retrospective study [10], which found that an initial uid resuscitation rate of 0.25-0.50 ml/kg/min may be associated with early septic shock reversal and lower 28-day mortality compared with slower rates of infusion. Thus, using an appropriate initial uid resuscitation rate may improve the prognosis of patients with septic shock. This study has some limitations. First, this study had a prospective observational design with a small sample size. We were not able to detect any causal relationship, and a large sample randomized controlled trial study is needed to con rm the results. Second, Stephanie P T showed that using adjusted body weight to calculate initial uid resuscitation volume for patients with obesity and suspected shock may improve outcomes compared to other weight-based dosing strategies [23]. This study followed actual weight-based uid administration; however, it remains unclear whether it is possible to obtain different results if ideal or adjusted body weight is used to calculate initial total uid volumes.

Conclusion
Initial uid resuscitation is currently viewed as the cornerstone of the treatment of septic shock [26][27]. The 2018 SSC recommends rapid administration of 30 ml/kg crystalloid for hypotension or lactate ≥4 mmol/L in patients with septic shock; however, there is no credible evidence to support this recommendation. This study demonstrates that insu cient initial uid resuscitation (below 20 ml/kg within the rst 1 h) or too much resuscitation (above 30 ml/kg within the rst 1 h) may increase 28-day mortality in patients with septic shock. The initial uid resuscitation rate of 20-30 ml/kg within the rst 1 h was associated with lower 28-day mortality and faster organ function recovery. Furthermore, septic shock patients who completed 30 ml/kg initial uid resuscitation between the rst 1-2 h had a lower 28day mortality rate and faster organ function recovery.
Abbreviations ICU, intensive care unit; SSC, Surviving Sepsis Campaign; SOFA, Sequential Organ Failure Assessment; ∆SOFA, Sequential Organ Failure Assessment change value; MAP, mean arterial blood pressure; APACHE II, Acute Physiology and Chronic Health Evaluation II; ARDS, acute respiratory distress syndrome; AKI, acute kidney injury; RRT, renal replacement therapy; IQR, interquartile range; BMI, body mass index.

Declarations Ethical Approval and Consent to participate
This study was approved by the Subei People's Hospital Institutional Review Board (2017KY-022).
Informed consent was obtained from each participant prior to his or her enrolment in this study.

Consent for publication
All authors agree to publish the article.

Availability of supporting data
The data are available from the corresponding author upon reasonable request.

Competing interests
All authors declare no competing interests. Author contributions QC: acquisition of data, analysis and interpretation of data, and critical revision of the manuscript for important intellectual content. WL: data analysis and interpretation and ensured that the accuracy and integrity of all work was appropriately maintained. FW: experimental conception and design, data acquisition, rst drafting of the manuscript, and full access to all the data in the study. HC: data acquisition, analysis and interpretation; revision of the manuscript. JS: data acquisition and analysis of data; statistical analysis and revision of the manuscript. RZ: data acquisition and interpretation of data; administrative and technical aspects. HW: data acquisition and interpretation of data; study supervision; and critical revision of the manuscript. All authors read and approved the nal manuscript.      Effect of the initial uid dose within the rst hour and the completion of 30 ml/kg liquid on the 28-day mortality rate in patients with septic shock. Abbreviations: ANOVA, analysis of variance Effect of the initial uid dose within the rst hour and the completion of 30 ml/kg liquid on ∆SOFA in patients with septic shock. Abbreviations: ∆SOFA, the difference of Sequential Organ Failure Assessment between the third day of admission and the rst day of admission.