Abstract
Objective
Oxygen and hemodynamic management are important for providing a sufficient adequate oxygen-containing blood to the organs for septic patients. In present study, we aimed to explore the application of sequential respiratory support (SRS) and the association of SRS with the outcome of septic patients who needed continuous renal replacement therapy (CRRT).
Methods
We extracted the medical information of septic patients who received CRRT within 24 h of intensive care unit (ICU) admission from the MIMIC-III v1.4. SRS was defined as receiving firstly oxygen therapy followed by mechanical ventilation (MV) within 24 h of admission to ICU. The propensity score matching (PSM) was performed to compare the differences in clinical characteristics and outcomes of patients with or without SRS. Finally, we developed logistic regression models to analyze the effects of SRS on hospital mortality.
Results
A total of 181 patients entered in this study, and there were 80 patients undergoing MV including SRS group (n = 61) and non-SRS group (n = 19). In the multivariate logistic regression, the value of SRS was associated with the lower risk of hospital mortality adjusted by minimum systolic BP (SBP), maximum lactate, vasopressor use, and sequential organ failure assessment (SOFA) score or Logistic Organ Dysfunction System (LODS) scores within the first 24 h of ICU stay. After PSM adjusted by SBP, maximum lactate, vasopressor use, SOFA, and LODS, there were 31 patients in SRS group with a and 18 cases in non-SRS group, displaying a significantly lower hospital mortality in SRS group than that in patients without SRS (19.4 % vs. 83.3 %, P < 0.001). In addition, age, qSOFA, necessitating the administration of vasopressor, and duration of vasopressor were significantly correlated with the hospital mortality in septic patients undergoing CRRT and SRS.
Conclusions
Receiving SRS within the first 24 h upon admission to the ICU was independently associated with the hospital mortality in patient with sepsis undergoing CRRT, and patients who were directly received MV had a high risk of death.
Keywords: Sepsis, Acute respiratory distress syndrome, Mortality, Continuous renal replacement therapy, Sequential respiratory support
1. Introduction
Sepsis is a life-threatening disease in intensive care units (ICUs) with high mortality. About 1.7 million adults hospitalize due to sepsis per year and 270,000 deaths in America [1]. In children, the pooled case-fatality rates (CRF) of severe sepsis and septic shock were 31.7% in developing countries, comparing with a CRF of 19.3% in developed countries [2]. Usually, the common site of infection is the respiratory tract [3], and sepsis-associated acute respiratory distress syndrome (ARDS) requiring mechanical ventilation (MV) would suffer the modest excess risk of mortality [4]. In pediatric patients with sepsis-associated ARDS, the hospital mortality was 24.6% and severity of hypoxemia accurately stratified the patient outcomes [5]. Either adult or children, elucidating the risk factors of prognosis should be helpful for the personalized management of patients with sepsis complicated by ARDS or respiratory failure.
Hemodynamic instability is a common problem and high risk factor in septic patient developing ARDS, which leads to tissue hypoxia either in lung or extra-pulmonary organs [6,7]. Liberal oxygen therapy increases mortality, and the recommendation about conservative administration of oxygenation is raised [8]. In regard of quality of life, lung injury, or the occurrence of sepsis-associated cardiovascular events, there are no differences between higher and lower oxygenation strategies [9]. Until now, the application of conservative oxygen therapy is still controversial, and the evidence-based research is not enough to make recommendations [10,11]. The surviving sepsis campaign (SSC) recommendations of international guideline suggest that continuous replacement renal therapy (CRRT) is used in sepsis or septic shock [12],. Recently, CRRT has become a preferred application to manage fluid overload during ARDS, leading to improved dynamic lung compliance (Cdyn) and oxygenation index (OI) in pediatric patients with ARDS [13]. Early extubation following non-invasive MV is feasible to cope with respiratory failure in children [14] or patients with difficult weaning off MV [15]. Sequential respiratory support (SRS) was defined as receiving firstly oxygen therapy followed by MV within 24 h of admission to ICU. However, there is little information about the effect of SRS in septic patients undergoing CRRT.
In the present study, septic patients received CRRT within 24 h of admission to ICU were selected from the public database of Martin Intensive Care (MIMIC)-III database v1.4 as described in our previous study [16]. The medical records about the application of respiratory support of these selected patients were collected, and we found that the application of SRS within 24 h of ICU hospitalization was significantly associated with the decreased hospital mortality in patient with sepsis undergoing CRRT.
2. Materials and methods
2.1. Study design
Referenced to our previous study [16], the patients met sepsis 3.0 and their medical recorded information were collected [17]. The application for access permission to the MIMIC-III database was authorized by Institutional Review Boards of BIDMC and Massachusetts Institute of Technology.
2.2. Study population
This is a secondary analysis of detailed information about 181 patients as reported in our previous study [16]. In the present study, SRS group and non-SRS group were included according to whether the patients received SRS during hospitalization. Hospital mortality was the primary outcome, and both length of ICU and hospital stay were the secondary outcomes.
2.3. Ethics approval
The MIMIC database is a well-known freely accessible database. In this study, 4 authors (Wang C, Zheng J, Zhao Y, and Liu T) completed the web-based course and obtained permission to access the dataset. Because all personal information was listed anonymously, informed consent and ethical approval were waived.
2.4. Data collection
All variables were collected from MIMIC III database as reported in our previous study [16]. In this study, we selected demographics, sequential organ failure assessment (SOFA), quick SOFA (qSOFA), systemic inflammatory response syndrome (SIRS), logistic organ dysfunction system (LODS), the values of clinical and laboratory indexes on the first service, the status of blood infection, the minimum values of systolic BP (SBP) and mean arterial pressure (MAP), oxyhemoglobin saturation (SpO2) and PaO2/FiO2, the minimum levels of platelet or albumin, the maximum levels of creatinine, bilirubin, blood urea nitrogen (BUN), or lactate, the maximum values of international normalized ratio (INR), and the minimum and maximum of white blood cells (WBCs). In addition, the ratio of vasopressor needed, vasopressor duration, the mode of oxygen therapy, and duration of ventilation were accessed.
2.5. Statistical analysis
Continuous variables were summarized as the mean ± standard deviation (SD) for normal distribution, and the median (interquartile range, IQR) were used for continuous variables with non-normal distribution. Accordingly, student's t-test or Mann-Whitney U test was used to compare the group differences according to data characteristics, respectively. Numbers or percentage were used to describe categorical variables, with the chi-square test for detecting the group differences. P value less than 0.05 was considered as statistically significant.
Bi-variable crude odds ratio and 95% confidence interval (CI) (P < 0.05) were used to select candidate variables for entering multivariable analysis. Then, selected variables were entered into Multivariate logistic regression to assess the association with the prognosis, and the results presented as the odd ratio (OR) and 95% confidence intervals (CI). STATA 15.0 MP (College Station, Texas, USA) were used to conduct all data analyses.
To avoid the differences in baseline characteristics, we performed propensity score matching (PSM) between patients with or without SRS. After PSM, confounders are evenly distributed, and it is more feasible to conduct the targeted comparisons of the two groups [18]. A PSM model (cal = 0.05 and k = 4) was used to minimize or reduce the confounding effect of covariate when estimating the effects of SRS on the outcome of hospital mortality. In this PSM model, minimum systolic BP, maximum lactate, vasopressor, SOFA, and LODS were considered to match. Based on the logistic regression, the propensity score for receiving oxygen therapy was assessed, and the correspondingβ coefficients and P-values for selected variables in the full propensity score model were shown in Supplemental Table S1.
3. Results
3.1. Baseline characteristics
Among 181 patients selected as described in our previous study [16], there were 101 cases without MV [MV (−)] and 80 patients received MV [MV (+)] during ICU stay. No significant differences were found in age, gender, ethnicity, the ratio of first service, and blood infection (all P > 0.05). In MV (−) subgroup, the length of both ICU and hospital stay was shorter in non-survivors compared with survivors (Table 1). In addition, the scores of SOFA, qSOFA, and LODS were significantly higher in non-survivors of MV (+) group (all P < 0.05), but not in non-survivors of MV (−) group (Table 1).
Table 1.
Baseline characteristics in patients with sepsis received CRRT.
| Parameters |
MV (−) (n = 101) |
P |
MV (+) (n = 80) |
P |
||||
|---|---|---|---|---|---|---|---|---|
| Total (n = 101) | Survivors (n = 95) | Non-survivors (n = 6) | Total (n = 80) | Survivors (n = 55) | Non-survivors (n = 25) | |||
| Demographic variables | ||||||||
| Gender male, n (%) | 63 (62.2) | 60 (63.2) | 3 (50) | 0.519 | 46 (57.5) | 29 (52.7) | 17 (68) | 0.200 |
| Age, year, mean (SD) | 62.6 (15.1) | 62.0 (15.1) | 72.5 (12.7) | 0.098 | 61.7 (13.8) | 61.0 (14.3) | 63.2 (12.8) | 0.517 |
| Ethnicity, n | 101 | 95 | 6 | 0.605 | 55 | 22 | 0.605 | |
| White | 49 | 46 | 3 | 35 | 17 | |||
| Black | 28 | 26 | 2 | 7 | 0 | |||
| Hispanic | 8 | 8 | 0 | 1 | 0 | |||
| others | 16 | 15 | 1 | 12 | 5 | |||
| Severity, median (IQR) | ||||||||
| SOFA | 6 (5, 7) | 6 （4, 7） | 7 （5, 8） | 0.359 | 10.7 (4.3) | 9.5 (3.7) | 13.3 (4.5) | <0.001 |
| qSOFA, mean (SD) | 1.7 （0.7） | 1.7 （0.7） | 2.3 （0.5） | 0.022 | 2.1 (0.7) | 1.9 (0.7) | 2.3 (0.7) | 0.021 |
| SIRS | 2 （2， 3） | 2 （2， 3） | 7 （5， 8） | 0.928 | 3 (3, 4) | 3 (3, 4) | 4 (3, 4) | 0.461 |
| LODS, mean (SD) | 4.9 （2.3） | 4.8 (2.2) | 6.7 (2.6) | 0.051 | 8.8 (3.4) | 7.9 (2.6) | 10.7 (4.1) | <0.001 |
| First service, n | 101 | 95 | 6 | 0.962 | 80 | 55 | 25 | 0.744 |
| CMED | 10 | 9 | 1 | 14 | 9 | 5 | ||
| MED | 96 | 81 | 5 | 65 | 45 | 20 | ||
| NMED | 2 | 2 | 0 | 1 | 1 | 0 | ||
| Others | 3 | 3 | 0 | 0 | 0 | 0 | ||
| Blood Infection, n (%) | 35 (34.7) | 34 (35.8) | 1 (16.7) | 0.340 | 33 (41.3) | 20 (36.4) | 13 (52) | 0.188 |
| Mechanical ventilation, n (%) | 0 (0) | 0 (0) | 0 (0) | 80 (100) | 55 (68.8) | 25 (31.2) | ||
| Length of ICU stay, days, median (IQR) | 2 (2, 3) | 2 (2, 3) | 3 (3, 7) | 0.093 | 5 （3，10） | 7 （3， 12） | 3 (2, 5) | 0.011 |
| Length of hospital stay, days, median (IQR) | 7 (4, 11) | 7 (5, 11) | 4 (2, 7) | 0.117 | 11（4， 21） | 13 (8, 22) | 3 (2, 10) | <0.001 |
SOFA: sequential organ failure assessment; qSOFA: Quick SOFA; SIRS: systemic inflammatory response syndrome; LODS: Logistic Organ Dysfunction System.
3.2. Clinical and laboratory parameters
Among 80 patients in MV (+) group, there were significantly different between survivors and non-survivors in aspect of SBP, MAP, lactate, bilirubin, or INR, platelet or albumin, and the ratio of vasopressor needed or receiving SRS (Table 2).
Table 2.
Indexes within 24 h after ICU admission in patients with sepsis received CRRT and MV.
| Parameters | Total (n = 80) | Survivors （n = 55） |
Non-survivors (n = 25) | P |
|---|---|---|---|---|
| Vital signs, median (IQR) if not otherwise specified | ||||
| Maximum heart rate (/min), mean (SD) | 110 (22) | 110 (22) | 110 (22) | 0.979 |
| Minimum systolic BP (mmHg), mean (SD) | 82 (19) | 87 (17) | 72 (19) | 0.001 |
| Systolic BP group, (mmHg) | 0.070 | |||
| Systolic BP ≥ 100, n (%) | 16 (20) | 14 (25.5) | 2 (8) | |
| Systolic BP < 100, n (%) | 64 (80) | 41 (74.5) | 23 (92) | |
| Minimum diastolic BP (mmHg) | 39 (31, 47) | 41 (34, 47) | 35 (30, 44) | 0.174 |
| Diastolic BP group, (mmHg) | 0.909 | |||
| Diastolic BP ≥ 60, n (%) | 6 (7.5) | 4 (7.3) | 4 (16) | |
| Diastolic BP < 60, n (%) | 74 (92.5) | 51 (92.7) | 23 (84) | |
| Minimum MAP (mmHg) | 51 (45, 58.5) | 53 (48, 61) | 47 (44, 51) | 0.018 |
| MAP group, (mmHg) | 0.303 | |||
| MAP ≥70, n (%) | 6 (7.5) | 3 (5.5) | 3 (12) | |
| MAP <70, n (%) | 74 (92.5) | 52 (94.5) | 22 (88) | |
| Maximum respiratory rate (/min) mean (SD) | 30 (8) | 30 (8) | 32 (7) | 0.197 |
| Respiratory rate group, (/min) | 0.782 | |||
| Respiratory rate ≤20, n (%) | 4 (5) | 3 (5.5) | 1 (4) | |
| Respiratory rate >20, n (%) | 76 (95) | 52 (94.5) | 24 (96) | |
| Maximum temperature (°C), | 37.6 (36.9, 384) | 37.7 (37.1, 38.2) | 37.2 (36.6, 38.9) | 0.306 |
| Serum laboratory variables, median (IQR) if not otherwise specified | ||||
| Maximum lactate (μmol/L) | 2.75 (1.8, 6.9) | 2.6 (1.6, 4.6) | 5.6 (2, 9.9) | 0.008 |
| Lactate group, (μmol/L) | 0.021 | |||
| Lactate <4, n (%) | 44 (55) | 35 (63.6) | 9 (36) | |
| Lactate ≥4, n (%) | 36 (45) | 20 (36.4) | 16 (64) | |
| Maximum creatinine (μmol/L) | 4.6 (3.3, 7) | 4.55 (3.3, 7.6) | 4.7 (3, 5.5) | 0.584 |
| Maximum glucose (mg/dL) | 198.5 (130.5, 305) | 193 (123, 301) | 240 (136, 331) | 0.332 |
| Maximum bilirubin (mg/dL) | 0.7 (0.4, 2.25) | 0.7 (0.4, 1.5) | 2.05 (0.6, 4.1) | 0.042 |
| Bilirubin group, (mg/dL) | 0.946 | |||
| Bilirubin <4, n (%) | 58 (72.5) | 40 (72.7) | 18 (72) | |
| Bilirubin ≥4, n (%) | 22 (27.5) | 15 (27.3) | 7 (28) | |
| Minimum platelet (×109/L) | 143 (85, 218) | 181 (105, 227) | 109 (74, 158) | 0.014 |
| Platelet group, (×109/L) | 0.091 | |||
| Platelet ≥ 100, n (%) | 58 (72.5) | 43 (78.2) | 15 (60) | |
| Platelet < 100, n (%) | 22 (27.5) | 12 (21.8) | 10 (40) | |
| Maximum INR | 1.5 (1.2, 2.2) | 1.4 (1.2, 1.8) | 2 (1.4, 2.7) | 0.017 |
| INR group | 0.115 | |||
| INR < 1.5, n (%) | 44 (55) | 27 (49.1) | 17 | |
| INR ≥ 1.5, n (%) | 36 (45) | 28 (50.9) | 8 | |
| Maximum BUN, (mmol/L) | 54 (43, 79) | 58.5 (43, 83) | 49 (44, 71) | 0.438 |
| Minimum WBC (×109/L) | 10.8 (6.9, 14.9) | 11.05 (7.9, 15.4) | 9.5 (5.4, 14) | 0.166 |
| Maximum WBC (×109/L) mean (SD) | 16.6 (8.0) | 17.1 (8.4) | 15.3 (7.2) | 0355 |
| Minimum albumin (g/dL), mean (SD) | 2.9 (0.7) | 3.0 (0.7) | 2.6 (0.5) | 0.043 |
| Albumin group, (g/dL) | 0.652 | |||
| Albumin ≥ 4, n | 26 | 17 | 9 | |
| Albumin < 4, n | 54 | 38 | 16 | |
| Pulmonary parameters, median (IQR) if not otherwise specified | ||||
| SpO2 | 95 (90.5, 97) | 95 (92, 97) | 93 (86, 97) | 0.235 |
| PaO2/FiO2, mmHg | 111.5 (73.5, 191) | 122 (75, 196.7) | 90 (63, 140) | 0.151 |
| Impaired pulmonary function group | 0.272 | |||
| PaO2/FiO2 ≥ 300, n | 22 | 17 | 5 | |
| 300 < PaO2/FiO2 ≤ 200, n | 7 | 4 | 3 | |
| 200 < PaO2/FiO2 ≤ 100, n | 23 | 18 | 5 | |
| PaO2/FiO2 < 100, n | 28 | 16 | 12 | |
| Ventilation durations, hours | 65.6 (27.9, 167.5) | 82 (30.8, 189.8) | 60 (27.4, 98.2) | 0.338 |
| Vasopressor | 0.004 | |||
| No, n (%) | 24 (30) | 22 (40) | 2 (8) | |
| Yes, n (%) | 56 (70) | 33 (60) | 23 (92) | |
| Vasopressor duration, hours | 44.8 (26.2, 122.4) | 42.0 (20.5, 135.0) | 59.6 (27.6， 109.8) | 0.459 |
| SRS, n (%) | 61 (76.3) | 51 (92.7) | 10 (40) | <0.001 |
SD: standard deviation; IQR: Inter Quartile Range; BP: blood pressure; SpO2: Pulse Oxygen Saturation; PaO2/FiO2: the ratio of the partial pressure of oxygen in arterial blood (PaO2) to the inspired oxygen fraction (FiO2); INR: International Normalized Ratio; SRS, sequential respiratory support.
3.3. Risk factors of hospital mortality
The univariate logistic regression analysis was performed using the variables with statistically significant differences according to the results of Table 1, Table 2. SOFA, qSOFA, LODS, SBP, lactate, vasopressor needed, and SRS were significantly correlated with hospital mortality in 80 septic patients undergoing CRRT and MV (all P < 0.05) (Supplementary Table S2). Moreover, multivariate logistic regression model adjusted by SOFA or LODS displayed that SRS was a protective factor of hospital mortality in 80 septic patients undergoing both CRRT and MV support (Table 3).
Table 3.
Multivariate Logistic analysis of factor related to hospital mortality.
| Parameters | OR (95% CI) | P |
|---|---|---|
| Model 1 | ||
| SRS | 0.047 (0.009–0.260) | <0.001 |
| Minimum systolic BP | 0.967 (0.924–1.012) | 0.144 |
| Maximum lactate | 1.022 (0.858–1.218) | 0.803 |
| Vasopressor | 4.956 (0.390–62.910) | 0.217 |
| SOFA | 1.020 (0.831–1.252) | 0.852 |
| Model 2 | ||
| SRS | 0.052 (0.010–0.255) | <0.001 |
| Minimum systolic BP | 0.985 (0.937–1.034) | 0.538 |
| Maximum lactate | 1.004 (0.849–1.186) | 0.966 |
| Vasopressor | 7.447 (0.378–146.810) | 0.187 |
| LODS | 1.218 (0.898–1.652) | 0.204 |
OR: odds ratio; CI: confidence interval; BP: blood pressure; SOFA: sequential organ failure assessment; SRS, sequential respiratory support.
To further analyze the significant association of SRS and hospital mortality in 80 septic patients received CRRT and MV, we compared the difference of critical variables between patients with or without SRS. The severity as shown as SOFA and LODS, the values of lactate, bilirubin, INR, the ratio of vasopressor needed, the length of ICU and hospital stay, as well as hospital mortality were significantly different (Table 4). After PSM, the hospital mortality in patients without SRS was still high compared with patients received SRS (83.3 % vs. 19.4 %, P < 0.001). In addition, all these patients without SRS had a shorter length of hospital or ICU stay (Table 4).
Table 4.
Comparison of the clinical characteristics and outcomes in septic patients received CRRT with (SRS) or without SRS (non-SRS).
| Parameters |
Before PSM |
After PSM |
||||||
|---|---|---|---|---|---|---|---|---|
| Total (n = 80) | SRS (n = 61) | non-SRS (n = 19) | P | Total (n = 49) | SRS (n = 31) | non- SRS (n = 18) | P | |
| Gender male, n (%) | 46 (57.5) | 33 (54.1) | 13 (68.4) | 0.270 | 32 (65.3) | 19 (61.3) | 13 (72.2) | 0.438 |
| Age, year, mean (SD) | 61.7 (13.8) | 62.9 (14.3) | 57.7 (11.5) | 0.150 | 60.5 (13.7) | 62.7 (14.7) | 56.8 (11.1) | 0.144 |
| SOFA, mean (SD) | 10.7 (4.3) | 9.7 (3.7) | 14.2 (4.5) | <0.001 | 13.1 (3.3) | 12.2 (2.4) | 14.7 (4.0) | 0.031 |
| qSOFA, mean (SD) | 2.1 (0.7) | 2.0 (0.7) | 2.1 (0.8) | 0.700 | 2.2 (0.6) | 2.3 (0.4) | 2.2 (0.8) | 0.604 |
| SIRS, median (IQR) | 3 (3, 4) | 4 (3, 4) | 3 (3, 4) | 0.230 | 3.4 (0.7) | 3.4 (0.7) | 3.5 (0.7) | 0.553 |
| LODS, mean (SD) | 8.8 (3.4) | 8.1 (2.7) | 10.8 (4.5) | 0.002 | 10.2 (3.2) | 9.7 (1.9) | 11.1 (4.5) | 0.136 |
| Maximum lactate (μmol/L) | 2.75 (1.8, 6.9) | 2.6 (1.7, 5.05) | 5.75 (2, 10.6) | 0.017 | 4.2 (2, 8.4) | 3.4 (2, 7.1) | 5.75 (2, 10.6) | 0.202 |
| Maximum creatinine (μmol/L) | 4.6 (3.3, 7.0) | 4.65 (3.3, 7.6) | 4.6 (3, 5.7) | 0.419 | 4.6 (3.2, 5.7) | 4.5 (3.2, 6.7) | 4.65 (3, 5.7) | 0.885 |
| Maximum glucose (mg/dL) | 198.5 (130.5, 305) | 193 (136, 284) | 279 (121, 408) | 0.237 | 241 (142, 319) | 213 (143, 296) | 284.5 (132, 408) | 0.246 |
| Maximum bilirubin (mg/dL) | 0.7 (0.4, 2.25) | 0.7 (0.4, 1.6) | 2.15 (0.7, 4.5) | 0.022 | 1.1 (0.5, 2.95) | 0.7 (0.4, 1.8) | 2.15 (0.7, 4.5) | 0.052 |
| Minimum platelet (×109/L) | 143 (85, 218) | 158 (101.5, 223.5) | 102 (80, 172) | 0.067 | 139 (80, 193) | 158 (78, 246) | 101.5 (80, 168) | 0.097 |
| Maximum INR | 1.5 (1.2, 2.2) | 1.4 (1.2, 1.8) | 2.1 (1.7, 3.7) | 0.002 | 1.8 (1.3, 2.4) | 1.5 (1.3, 2.2) | 2.05 (1.7, 3.1) | 0.059 |
| Maximum BUN, (mmol/L) | 54 (43, 79) | 58 (44, 92) | 47 (37, 66) | 0.179 | 49 (39, 66) | 56 (41, 92) | 46 (37, 63) | 0.299 |
| Minimum WBC (×109/L) | 10.8 (6.9, 14.9) | 10.7 (7.35, 15.45) | 11.1 (4.3, 13.8) | 0.273 | 10.8 (6.9, 14.8) | 10.4 (7.5, 15.9) | 11.2 (4.3, 13.8) | 0.481 |
| Maximum WBC (×109/L), mean (SD) | 16.6 (8.0) | 16.8 (8.1) | 15.9 (7.9) | 0.667 | 17.1 (8.2) | 17.6 (8.5) | 16.3 (7.9) | 0.618 |
| Minimum albumin (g/dL), mean (SD) | 2.9 (0.7) | 3.0 (0.7) | 2.6 (0.6) | 0.117 | 2.8 (0.7) | 2.9 (0.8) | 2.4 (0.6) | 0.231 |
| SpO2 | 95 (90.5, 97) | 95 (92, 97) | 94 (86, 96) | 0.185 | 93 (87.5, 96) | 93 (89, 95) | 94 (86, 96) | 0.856 |
| PaO2/FiO2, mmHg | 111.5 (73.5, 191) | 112.9 (75, 190) | 101.5 (63, 192) | 0.407 | 94 (63, 152) | 94 (73, 133.3) | 101.5 (63, 192) | 0.654 |
| Vasopressor | 0.034 | 0.691 | ||||||
| No, n (%) | 24 | 22 | 2 | 2 | 1 | 1 | ||
| Yes, n (%) | 56 | 39 | 17 | 47 | 30 | 17 | ||
| Vasopressor duration, hours | 44.8 (26.2, 122.4) | 47.0 (20.5, 178.6) | 40.8 (29.8, 59.6) | 0.539 | 47.0 (26.5, 135.0) | 58.8 (20.5, 178.6) | 40.8 (29.8, 59.6) | 0.319 |
| Ventilation durations, hours | 65.6 (27.9, 167.5) | 93.9 (27.3, 236.3) | 37.7 (28.3, 82) | 0.115 | 86.1 (36.3, 228.2) | 149.9 (60.6, 288)) | 44.3 (28.3, 82) | 0.319 |
| Length of ICU stay, days, median (IQR) | 5 (3, 10) | 7 (4, 14) | 2 (1, 4) | <0.001 | 5.7 (3.1, 12.0) | 9.2 (5.5, 19.4) | 2.2 (1.4, 3.9) | <0.001 |
| Length of hospital stay, days, median (IQR) | 11 (4, 21) | 13 (8, 23) | 2 (1, 4) | <0.001 | 11.2 (3.2, 20.7) | 18.6 (11.2, 24.3) | 2.5 (1.5, 4.2) | <0.001 |
| Hospital mortality, n (%) | 25 (31.3) | 10 (16.4) | 15 (78.9) | <0.001 | 21 (42.9) | 6 (19.4) | 15 (83.3) | <0.001 |
SOFA: sequential organ failure assessment; qSOFA: Quick SOFA; SIRS: systemic inflammatory response syndrome; LODS: Logistic Organ Dysfunction System.
After comparing the outcomes and the clinical characteristics of patients received SRS, the age of non-survivors was significantly larger than that of survivors (P = 0.008), and the qSOFA score was relatively higher in non-survivors than survivors, but without statistical significance (P = 0.062). The ratio of vasopressor support in non-survivors was 90% comparing 58.8% in survivors (P = 0.060). The levels of lactate, platelet, INR, and vasopressor duration displayed slight differences but without statistical significance (all P > 0.05) (Table 5). Due to the small size of non-survivors (n = 10), if we defined P value < 0.1 as statistically significant, the variables of age, qSOFA, vasopressor need, and vasopressor duration were associated with the hospital mortality of septic patients received CRRT with SRS (Supplementary Table S3).
Table 5.
Comparison of the clinical characteristics in septic patients received CRRT with SRS.
| Parameters | Total (n = 61) | Survivors (n = 51) | Non-survivors (n = 10) | P |
|---|---|---|---|---|
| Gender male, n (%) | 33 (54.1) | 27 (52.9) | 6 (60) | 0.682 |
| Age, year, mean (SD) | 62.9 (14.3) | 60.8 (14.5) | 73.8 (7.0) | 0.008 |
| SOFA, mean (SD) | 9.7 (3.7) | 9.5 (3.6) | 10.5 (4.1) | 0.519 |
| qSOFA, mean (SD) | 2.0 (0.7) | 2.0 (0.7) | 2.4 (0.5) | 0.062 |
| SIRS, median (SD) | 3.2 (0.8) | 3.2 (0.8) | 3.0 (0.8) | 0.369 |
| LODS, mean (SD) | 8.1 (2.7) | 7.9 (2.6) | 8.9 (2.7) | 0.301 |
| Maximum lactate (μmol/L) | 2.6 (1.7, 5.1) | 2.6 (1.6, 4.6) | 5 (2, 7.1) | 0.129 |
| Maximum creatinine (μmol/L) | 4.7 (3.3, 7.6) | 4.6 (3.3, 7.8) | 5.1 (3.8, 6.7) | 0.923 |
| Maximum glucose (mg/dL) | 193 (136, 284) | 196 (140, 309) | 160 (122, 240) | 0.231 |
| Maximum bilirubin (mg/dL) | 0.7 (0.4, 1.6) | 0.7 (0.4, 1.5) | 1.4 (0.4, 2.9) | 0.381 |
| Minimum platelet (×109/L) | 158 (10.1.5, 223.5) | 188.5 (105, 227) | 128 (74, 158) | 0.110 |
| Maximum INR | 1.4 (1.2, 1.8) | 1.4 (1.2, 1.6) | 1.8 (1.4, 2.7) | 0.170 |
| Maximum BUN, (mmol/L) | 58 (44, 92) | 58.5 (41, 83) | 54 (46, 98) | 0.552 |
| Minimum WBC (×109/L) | 10.7 (7.4, 15.5) | 10.9 (7.9, 15.9) | 9.7 (6.7, 14.6) | 0.558 |
| Maximum WBC (×109/L), mean (SD) | 16.8 (8.1) | 17.0 (8.5) | 15.6 (6.2) | 0.601 |
| Minimum albumin (g/dL), mean (SD) | 3.0 (0.7) | 3.0 (0.7) | 2.8 (0.6) | 0.658 |
| SpO2 | 95 (92, 97) | 95 (91, 97) | 95 (93, 98) | 0.826 |
| PaO2/FiO2, mmHg | 113 (75, 190) | 119 (75, 193) | 101 (84, 140) | 0.744 |
| Vasopressor, n (%) | 39 (63.9) | 30 (58.8) | 9 (90) | 0.060 |
| Vasopressor duration, hours | 47.0 (20.5, 178.6) | 41.8 (15.5, 139.4) | 109.8 (68.0, 218.0) | 0.134 |
| Ventilation durations, hours | 93.9 (27.3, 236.3) | 93.9 (27.9, 226.5) | 95.3 (25.5, 408) | 0.671 |
| Length of ICU stay, days, median (IQR) | 6.7 (3.6, 13.7) | 6.7 (3.6, 13.2) | 8.0 (3.4, 24.1) | 0.726 |
| Length of hospital stay, days, median (IQR) | 13.0 (7.8, 23.0) | 13.7 (9.7, 23.0) | 8.1 (3.2, 24.0) | 0.205 |
SOFA: sequential organ failure assessment; qSOFA: Quick SOFA; SIRS: systemic inflammatory response syndrome; LODS: Logistic Organ Dysfunction System.
4. Discussion
Patients with sepsis with respiratory support receive non-invasive oxygen therapy or MV when complicated with respiratory failure. However, until now, there is still little information variable of detailed criteria for respiratory support [10,[19], [20], [21]]. Though oxygen support significantly increases survival in animal sepsis model [22], but it is ambiguous for the clinical benefits in different mode of respiratory support in patients [10,20]. In the present study, we firstly get evidence that SRS could get clinical benefits in septic patients received CRRT.
Acute respiratory failure is risk factor for the worse outcome of sepsis. The patients with sepsis induced by pulmonary infection displayed a lower in-hospital mortality than that in nonpulmonary source of sepsis [23]. There were metabolically different between sepsis-induced direct and indirect ARDS [24], and there are two phenotypes identified as hyperinflammatory phenotype with worse clinical outcomes and hypoinflammatory phenotype with a longer duration of MV [25]. In our present study, the total mortality of 80 septic patients undergoing both CRRT and MV support was 31.3% (16.4% in the SRS group vs.78.9% in the non-SRS group). The differences of hospital mortality could be mainly related to the subtype of sepsis because the results after PSM still showed the worse outcome of patients with sepsis receiving CRRT and MV (42.9% in the SRS group vs. 83.3% in the non-SRS group). High-flow nasal cannula, one of kind of non-invasive ventilation (NIV), significantly decreased the respiratory drive and effort in septic patients of nonpulmonary origin [26], and NIV failure was associated with the severity of sepsis [27]. Considering that the advantage of CRRT in improving hemodynamics in patients with septic shock, SRS could be favorable for early oxygen delivery to improve the microcirculation and organ function. It was possible that the critical emergency led to direct MV. However, after PSM, SRS was still significantly associated with the better outcome of septic patients receiving CRRT. It may be worth noting that the importance of oxygen supplement for extrapulmonary requirement.
A recent meta-analysis indicated that both SOFA score and pulmonary sepsis were associated with ARDS [28]. In our present study, it could be related to the limitation of included septic patients with CRRT, and there were no significantly association of MBP, maximum lactate, vasopressor, SOFA, or LODS. We speculated that the support of CRRT could contribute the outcome of these patients. Moreover, the higher non-respiratory pediatric logistic organ dysfunction (PELOD-2) score and PaO2/FiO2 < 100 at diagnosis were independently associated with NIV failure in pediatric ARDS, and use of NIV at diagnosis of mild to moderate hypoxemia resulted in shorter exposure to MV in children [29]. It is interesting that the hospital mortality of 61 patients with SRS was significantly lower than that of patients without SRS. Consistently, it has been proved that conservative oxygen therapy decreased the MV duration, new organ failure, and the risk of RRT application during ICU hospitalization [30]. So, we propose that SRS support within 24 h of admission to ICU could be of great advantage in septic patients undergoing CRRT support. From another view, conventional therapy would leave high ratio of patients exposed to hyperoxemia compared to conservative oxygen group (44.5% vs. 11.4%) [31]. To avoid the harmful effect of hyperbaric oxygenation, conservative oxygen supplement should be suggested.
This study has several limitations. First, it is a database study with a lack of detailed information about SRS, and the clinical features including sepsis-causing origin organ, pathogen strain, etc. Second, the sample size was small due to the limited septic patients receiving the support of CRRT. The protective roles of SRS in patients under CRRT were firstly reported in this study; however, it is more important to conduct a prospective study in well-designed clinical research.
5. Conclusions
It is still ambiguous to clearly demonstrate the clinical benefits of different mode of respiratory support in septic patients undergoing CRRT. In this study, this is the first report to get evidence that septic patients received CRRT could obtain clinical benefits from SRS within the first 24 h of ICU stay, which is an independently protective factor of hospital mortality in septic patient undergoing CRRT.
Data availability statement
Datasheet used in this study were submitted as supplementary materials, which can be made available.
CRediT authorship contribution statement
Chunxia Wang: Writing – original draft, Supervision, Investigation, Data curation, Conceptualization. Jianli Zheng: Software, Investigation, Funding acquisition, Formal analysis, Data curation. Yilin Zhao: Methodology, Investigation, Data curation. Tiantian Liu: Methodology, Investigation, Conceptualization. Yucai Zhang: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study was supported by the discipline building project from Institute of Pediatric Infection, Immune, and Critical Care Medicine, Shanghai Children's Hospital, and Shanghai Municipal Education Commission–Gaofeng Clinical Medicine Grant (20171928).
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.heliyon.2024.e27563.
Contributor Information
Chunxia Wang, Email: karencx0465@163.com.
Jianli Zheng, Email: jlzheng@usst.edu.cn.
Yilin Zhao, Email: zhaoyilinleo0629@163.com.
Tiantian Liu, Email: m17853137921@163.com.
Yucai Zhang, Email: zyucai2018@163.com.
Appendix A. Supplementary data
The following is/are the supplementary data to this article:
References
- 1.Rhee C., Dantes R., Epstein L., et al. Incidence and trends of sepsis in US hospitals using clinical vs claims data, 2009-2014. JAMA. 2017;318:1241–1249. doi: 10.1001/jama.2017.13836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tan B., Wong J.J., Sultana R., et al. Global case-fatality rates in pediatric severe sepsis and septic shock: a systematic review and meta-analysis. JAMA Pediatr. 2019;173:352–362. doi: 10.1001/jamapediatrics.2018.4839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wang S., Yin F., Zhang Y., et al. Epidemiology and clinical characteristics of pediatric sepsis in PICUs of China: a national cross-sectional study. MedComm. 2020;2023(4):e211. doi: 10.1002/mco2.211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Keim G., Percy A.G., Himebauch A.S., et al. Acute respiratory failure-related excess mortality in pediatric sepsis. Thorax. 2023;78:1135–1137. doi: 10.1136/thorax-2022-219961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhou L., Li S., Tang T., et al. A single-center PICU present status survey of pediatric sepsis-related acute respiratory distress syndrome. Pediatr. Pulmonol. 2022;57:2003–2011. doi: 10.1002/ppul.25943. [DOI] [PubMed] [Google Scholar]
- 6.Vignon P., Evrard B., Asfar P., et al. Fluid administration and monitoring in ARDS: which management? Intensive Care Med. 2020;46:2252–2264. doi: 10.1007/s00134-020-06310-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Vieillard-Baron A., Matthay M., Teboul J.L., et al. Experts' opinion on management of hemodynamics in ARDS patients: focus on the effects of mechanical ventilation. Intensive Care Med. 2016;42:739–749. doi: 10.1007/s00134-016-4326-3. [DOI] [PubMed] [Google Scholar]
- 8.Chu D.K., Kim L.H., Young P.J., et al. Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. Lancet. 2018;391:1693–1705. doi: 10.1016/S0140-6736(18)30479-3. [DOI] [PubMed] [Google Scholar]
- 9.Barbateskovic M., Schjorring O.L., Krauss S.R., et al. Higher vs lower oxygenation strategies in acutely ill adults: a systematic review with meta-analysis and trial sequential analysis. Chest. 2021;159:154–173. doi: 10.1016/j.chest.2020.07.015. [DOI] [PubMed] [Google Scholar]
- 10.Li Y., Zhao L., Yu Y., et al. Conservative oxygen therapy in critically ill and perioperative period of patients with sepsis-associated encephalopathy. Front. Immunol. 2022;13 doi: 10.3389/fimmu.2022.1035298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Young P., Mackle D., Bellomo R., et al. Conservative oxygen therapy for mechanically ventilated adults with sepsis: a post hoc analysis of data from the intensive care unit randomized trial comparing two approaches to oxygen therapy (ICU-ROX) Intensive Care Med. 2020;46:17–26. doi: 10.1007/s00134-019-05857-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Evans L., Rhodes A., Alhazzani W., et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47:1181–1247. doi: 10.1007/s00134-021-06506-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wang F., Wang C., Shi J., et al. Lung ultrasound score assessing the pulmonary edema in pediatric acute respiratory distress syndrome received continuous hemofiltration therapy: a prospective observational study. BMC Pulm. Med. 2021;21:40. doi: 10.1186/s12890-021-01394-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li X.Q., Zhao W.L., Li D.Y., et al. [Clinical study on early extubation and sequential non-invasive respiratory support for children with acute respiratory failure receiving invasive mechanical ventilation] Sichuan Da Xue Xue Bao Yi Xue Ban. 2022;53:321–326. doi: 10.12182/20220360206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang S.Y., Liang H.W., Lu G.S., et al. Effect of sequential high-flow nasal cannula oxygen therapy and non-invasive positive-pressure ventilation in patients with difficult weaning from mechanical ventilation after extubation on respiratory mechanics. Ann. Transl. Med. 2021;9:1251. doi: 10.21037/atm-21-3408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wang C., Zheng J., Wang J., et al. Cox-LASSO analysis for hospital mortality in patients with sepsis received continuous renal replacement therapy: a MIMIC-III database study. Front. Med. 2021;8 doi: 10.3389/fmed.2021.778536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Johnson A.E., Pollard T.J., Shen L., et al. MIMIC-III, a freely accessible critical care database. Sci. Data. 2016;3 doi: 10.1038/sdata.2016.35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kane L.T., Fang T., Galetta M.S., et al. Propensity score matching: a statistical method. Clin. Spine Surg. 2020;33(3):120–122. doi: 10.1097/BSD.0000000000000932. [DOI] [PubMed] [Google Scholar]
- 19.Roy K.L., Fisk A., Forbes P., et al. Inadequate oxygen delivery dose and major adverse events in critically ill children with sepsis. Am. J. Crit. Care. 2022;31:220–228. doi: 10.4037/ajcc2022125. [DOI] [PubMed] [Google Scholar]
- 20.Minasyan H. Oxygen therapy for sepsis and prevention of complications. Acute Crit. Care. 2022;37:137–150. doi: 10.4266/acc.2021.01200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lin M., Stewart M.T., Zefi S., et al. Dual effects of supplemental oxygen on pulmonary infection, inflammatory lung injury, and neuromodulation in aging and COVID-19. Free Radic. Biol. Med. 2022;190:247–263. doi: 10.1016/j.freeradbiomed.2022.08.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Halbach J.L., Prieto J.M., Wang A.W., et al. Early hyperbaric oxygen therapy improves survival in a model of severe sepsis. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2019;317:R160–R168. doi: 10.1152/ajpregu.00083.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sevransky J.E., Martin G.S., Mendez-Tellez P., et al. Pulmonary vs nonpulmonary sepsis and mortality in acute lung injury. Chest. 2008;134:534–538. doi: 10.1378/chest.08-0309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Chang Y., Yoo H.J., Kim S.J., et al. A targeted metabolomics approach for sepsis-induced ARDS and its subphenotypes. Crit. Care. 2023;27:263. doi: 10.1186/s13054-023-04552-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Dahmer M.K., Yang G., Zhang M., et al. Identification of phenotypes in paediatric patients with acute respiratory distress syndrome: a latent class analysis. Lancet Respir. Med. 2022;10:289–297. doi: 10.1016/S2213-2600(21)00382-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mauri T., Spinelli E., Pavlovsky B., et al. Respiratory drive in patients with sepsis and septic shock: modulation by high-flow nasal cannula. Anesthesiology. 2021;135:1066–1075. doi: 10.1097/ALN.0000000000004010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Duan J., Chen L., Liang G., et al. Noninvasive ventilation failure in patients with hypoxemic respiratory failure: the role of sepsis and septic shock. Ther. Adv. Respir. Dis. 2019;13 doi: 10.1177/1753466619888124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mayow A.H., Ahmad F., Afzal M.S., et al. A systematic review and meta-analysis of independent predictors for acute respiratory distress syndrome in patients presenting with sepsis. Cureus. 2023;15 doi: 10.7759/cureus.37055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Emeriaud G., Pons-Odena M., Bhalla A.K., et al. Noninvasive ventilation for pediatric acute respiratory distress syndrome: experience from the 2016/2017 pediatric acute respiratory distress syndrome incidence and epidemiology prospective cohort study. Pediatr. Crit. Care Med. 2023;24:715–726. doi: 10.1097/PCC.0000000000003281. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ni Y.N., Wang T., Liang B.M., et al. The effect of conservative oxygen therapy in reducing mortality in critical care patients: a meta-analysis and trial sequential analysis. Front. Med. 2021;8 doi: 10.3389/fmed.2021.738418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Eastwood G.M., Chan M.J., Peck L., et al. Conservative versus conventional oxygen therapy for cardiac surgical patients: a before-and-after study. Anaesth. Intensive Care. 2019;47:175–182. doi: 10.1177/0310057X19838753. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Datasheet used in this study were submitted as supplementary materials, which can be made available.