Sage Journals: Discover world-class research

Abstract

Background

The clinical use of transthoracic echocardiography (TTE) in patients with acute respiratory distress syndrome (ARDS) in the intensive care unit (ICU) has dramatically increased, its impact on long-term prognosis in these patients has not been studied. This study aimed to explore the effect of early-TTE on long-term mortality in patients with moderate-to-severe ARDS in ICU.

Methods

A total of 2833 patients with moderate-to-severe ARDS who had or had not received early-TTE were obtained from the Medical Information Mart for Intensive Care (MIMIC-III) database after imputing missing values by a random forest model, patients were divided into early-TTE group and non-early-TTE group according to whether they received TTE examination in ICU. A variety of statistical methods were used to balance 41 covariates and increase the reliability of this study, including propensity score matching, inverse probability of treatment weight, covariate balancing propensity score, multivariable regression, and doubly robust estimation. Chi-Square test and t-tests were used to examine the differences between groups for categorical and continuous data, respectively.

Results

There was a significant improvement in 90-day mortality in the early-TTE group compared to non-early-TTE group (odds ratio = 0.79, 95% CI: 0.64–0.98, p-value = 0.036), revealing a beneficial effect of early-TTE. Net-input was significantly decreased in the early-TTE group on the third day of ICU admission and throughout the ICU stay, compared with non-early-TTE group (838.57 vs. 1181.89 mL, p-value = 0.014; 4542.54 vs. 8025.25 mL, p-value = 0.05). There was a significant difference in the reduction of serum lactate between the two groups, revealing the beneficial effect of early-TTE (0.59 vs. 0.83, p-value = 0.009). Furthermore, the reduction in the proportion of acute kidney injury demonstrated a correlation between early-TTE and kidney protection (33% vs. 40%, p-value < 0.001).

Conclusions

Early application of TTE is beneficial to improve the long-term mortality of patients with moderate-to-severe ARDS.

Keywords

Transthoracic echocardiography ARDS critical care long-term outcomes propensity score matching

Introduction

The dyspnea as a manifestation of acute respiratory distress syndrome (ARDS) is characterized by non-cardiogenic pulmonary edema and rapidly progressive hypoxemia. Despite recent advances in medical technology, in-hospital mortality rate in ARDS patients remains high, ranging from 40% to 50%.¹ Therefore, early identification and timely intervention are important measures to improve the prognosis of ARDS.

Lung protective mechanical ventilation and fluid management are widely used to improve survival in ARDS patients by maintaining adequate oxygenation, carbon dioxide removal, and systemic perfusion.² Nonetheless, these strategies should be applied with caution, as they may also cause negative impacts on patients. Mechanical ventilation can increase the loading of right ventricle through hemodynamic and respiratory systems, further increasing the risk of cardiovascular dysfunction.^3,4 Fluid overload caused by increased alveolar-capillary barrier permeability in ARDS patients may increase hydrostatic pressure and promote the development of pulmonary edema.⁵ Therefore, optimal management with adequate oxygen supply is required to minimize the occurrence of lung edema.^6–10 Furthermore, acute cor pulmonary heart disease, characterized by severe right ventricular dysfunction, is a frequent and usually lethal complication of ARDS and has been considered as an independent risk factor for mortality in ARDS.^11–13

Transthoracic echocardiography (TTE) is a noninvasive technique that is frequently used to collect important hemodynamic data.¹⁴ TTE information can provide valuable diagnostic and prognostic information that support the strategic planning of fluid management, vasopressor, and ventilator use to avoid pulmonary edema, left ventricular failure, shock, and multiple organ dysfunction due to right ventricular dysfunction.^15–17 Therefore, early-TTE may be of great clinical benefit in ARDS patients.

A recent study showed that early-TTE can significantly reduce 28-day mortality, suggesting a clinical benefit of early-TTE on short-term survival outcomes in ARDS patients.¹⁸ However, the impact of early-TTE on the long-term outcomes of ARDS patients has not been reported. In fact, long-term outcomes may be more likely to be influenced by multiple factors, such as pulmonary dysfunction, psychological morbidity, neuromuscular weakness, and long-term cognitive impairment.^19–21 Moreover, up to 40% of ARDS survivors had at least one post-discharge hospitalization during the 1-year follow-up period,^22,23 suggesting that the risk of death in ARDS patients after discharge remains high.^24–26 In addition, recent studies have shown that there are large discrepancies between ARDS survivors at discharge and those during long-term follow-up.^27–30

Given this discrepancy between short-term and long-term mortality among ARDS survivors, the impact of early-TTE on long-term mortality in ARDS patients should be explored. Therefore, this study investigated whether early-TTE can significantly improve long-term mortality in patients with moderate-to-severe ARDS.

Materials and methods

Study design and patients

Patients from the Medical Information Mart for Intensive Care (MIMIC-III) database who met the following inclusion criteria were included in the study: (a) age ≥ 16 years old; (b) use of mechanical ventilation on the first day in the intensive care unit (ICU); (c) length of ICU stay ≥7 days; and (d) diagnosis of ARDS according to the Berlin definition. MIMIC-III database lacked chest x-ray results. Thus, to rule out cardiogenic pulmonary edema as a potential etiology of acute hypoxemic respiratory failure,³¹ we excluded patients who had heart failure as the admission diagnosis in the MIMIC-III database.

For patients with multiple ICU readmissions, only data from their first admission were obtained and used for analysis. For patients admitted to the ICU for less than 24 h, it is difficult to determine why they left the ICU and whether they developed ARDS within a short period of time. To simplify the process of personnel screening, we have added restrictions on the length of ICU stay.

Patients who underwent TTE within 24 h after admission to the ICU were divided into early-TTE group, and the rest were divided into non-early-TTE group. All ARDS patients in the study cohort were further graded into mild, moderate, and severe ARDS according to the PaO₂/FiO₂ (P/F) ratio of the Berlin definition,³² as follows: mild: 200 < P/F ≤ 300 mmHg; moderate: 100 < P/F ≤ 200 mmHg; severe: P/F ≤ 100 mmHg. Patients diagnosed with mild ARDS were excluded from this study.

Data sources

The data used in this study obtained from a public database, the MIMIC-III, which contains information on more than 15,000 patients admitted to ICU. The first measurement data of patients after admission to the ICU were extracted from the MIMIC-III database by using PostgreSQL, including age, sex, weight, Sequential organ failure assessment (SOFA) score,³³ Simplified acute physiology score (SAPS) II score,³⁴ Elixhauser score,³⁵ Acute physiology and chronic health evaluation (APACHE) III score,³⁶ concomitant diseases (chronic obstructive pulmonary disease (COPD), atrial fibrillation (AFIB), chronic kidney disease, congestive heart failure (CHF), liver disease, coronary heart disease (CAD), malignancy), vital signs (mean arterial pressure (MAP), heart rate, temperature, central venous pressure (CVP)), interventions (vasopressors (epinephrine, norepinephrine, dopamine, dobutamine, phenylephrine), and sedative drugs (midazolam, propofol, diazepam, dexmedetomidine, alprazolam)), laboratory values (white blood cell (WBC) count, platelet count, hemoglobin (HB), lactate, chloride, creatinine, creatinine kinase, sodium, N-terminal prohormone of brain natriuretic peptide (NT-proBNP), troponin, pH, partial pressure of oxygen (PO₂), P/F ratio, potassium, partial pressure of carbon dioxide (PCO₂), bicarbonate, blood urea nitrogen (BUN)), the amount of input (include intravenous fluids, blood products, oral medications, and tube feedings) and output (include urine output, chest tube drainage, wound drainage, and dialysis output) in the first, second, third day after admission to ICU, the total amount of in-hospital input and output.

The primary outcome was 90-day mortality after ICU admission. Secondary outcomes included mechanical ventilation- and vasopressor-free days within 28 days of ICU admission. Total net-input patients received on days one, two, and three in the ICU. The reductions in serum creatinine and serum lactate were calculated as follows: for the early-TTE group, the measured value of serum creatine and serum lactate after TTE minus the measured value before TTE. For the non-early-TTE group, the value measured 48 h after admission to the ICU minus the value measured for the first time in the ICU.

Statistical analysis

R package “missForest” was used to impute missing values through a random forest model.³⁷ Propensity score matching (PSM) was used to match characteristics of study covariates to reduce bias.³⁸ A propensity score (PS) was calculated for each individual using logistic regression based on covariates, including age, gender, service unit, weight, SOFA score, SAPS II score, Elixhauser score, APACHE III score, height, use of vasopressor, percentage of severe state, use of sedatives, CHF, AFIB, renal, liver, COPD, CAD, malignancy, temperature, CVP, MAP, heart rate, WBC, HB, platelet, pH, PO₂, PCO₂, P/F ratio, sodium, potassium, bicarbonate, chloride, BUN, creatinine, creatinine kinase, NT-proBNP, troponin, lactate, and day of lCU admission. Nearest neighbor matching was used to find a patient in the non-early-TTE group whose PS score was closest to the patient in the early-TTE group (maximum caliper = 0.2, 1:1 matched), and then the two patients were no longer involved in subsequent matching (replace = F). To assess the balance of covariate distributions between early-TTE and non-early-TTE groups, standardized mean differences (SMDs) were calculated. In this study, t-test and Chi-Square test were used for comparison analysis of continuous and categorized data, respectively. The power of the significant test was calculated using R package “pwr.”

Sensitivity analysis

Sensitivity analysis was conducted using the inverse probability of treatment weight (IPTW),^39,40 covariate balancing propensity score (CBPS),⁴¹ multivariable regression, and doubly robust estimation.⁴² IPTW is a PS-based weighting method. Each individual is given a weight of 1/PS and 1/(1-PS) according to whether it belongs to the treatment group or the control group. A new cohort is obtained according to weight sampling, which can reduce the influence of cofoundering factors. However, a very large or very low PS will cause the individual to have a large weight, thereby increasing the variability of the estimated treatment effect.⁴³ Therefore, the improved stabilized weights proposed by Cole SR et al. was used for weighting⁴³:

w = \frac{Z \Pr (Z = 1)}{P S} + \frac{(1 - Z) \Pr (Z = 0)}{1 - P S}

Z = 1 denotes that this individual belongs to the early-TTE group. Pr(Z = 1) and Pr(Z = 0) represent the marginal probability of the early-TTE and no-early-TTE group in the overall sample.

The CBPS approach refers to modeling treatment allocation while optimizing covariate balance. This was done by balancing the conditional probabilities of score and treatment assignment using the dual features of PSs as covariates. CBPS was conducted using the R package “CBPS.”⁴⁴ Doubly robust estimation is an estimation technique that prevents model misspecification using two models. One for outcomes, and another for exposure. If one of the models is correct (not necessarily both), this will yield an unbiased estimate of the exposure-outcome association. Here, we used R package “fastDR” for doubly robust estimation.⁴⁵ In addition, the multivariable regression used in this study was based on logic regression analysis using R package “stats.”

Results

Propensity score matching and other methods

After reviewing 61,532 patients admitted to the ICU from the MIMIC-III database, ARDS patients were initially identified according to the following criteria: (a) patients had a first-time ICU admission and (b) patients had a diagnosis of ARDS according to the Berlin criteria. Finally, a total of 3912 eligible ARDS patients were included in the study cohort, including 1079 mild, 1858 moderate, and 975 severe ARDS patients (Figure 1). The number of patients in the early-TTE group and non-early-TTE was 811 and 2022, respectively. Since patients with moderate and severe ARDS are more likely to develop pulmonary heart disease and may benefit from early-TTE, we hypothesized that early-TTE would be of great clinical value in these patients. Therefore, we focused on the prognostic value of early-TTE in patients with moderate and severe ARDS.

Figure 1.

Flow chart illustrating the exclusion and inclusion criteria used to screen 2833 patients as the final study cohort. Abbreviations: MIMIC: Medical Information Mart for Intensive Care; ARDS: Acute respiratory distress syndrome.

Table 1 summarizes the characteristics of patients with moderate-to-severe ARDS who received or did not receive early-TTE. Patients in the early-TTE group had significantly higher disease severity scores than those in the non-early-TTE group, including SAPS II score (23.13 ± 4.74 vs. 21.72 ± 4.64, p-value < 0.001), SOFA score (8.51 ± 3.38 vs. 7.18 ± 3.37, p-value < 0.001), APACHE III score (61.29 ± 23.33 vs. 51.82 ± 21.58, p-value <0.001), and Elixhauser score (4.27 ± 2.25 vs. 3.51 ± 2.09, p-value <0.001). A higher proportion of patients in the early-TTE groups received vasopressor therapy (68.3% vs. 51.5%, p-value <0.001) and use of sedatives (48.5% vs. 42.8%, p-value = 0.007).

Table 1.

Demographic an d characteristics of patients with moderate-to-severe ARDS stratified by early-TTE.

Variables	Early-TTE		p-value	SMD	Missing data (%)
Variables	No (2022)	Yes (811)	p-value	SMD	Missing data (%)
Age (mean (SD))	62.76 (16.97)	63.62 (16.52)	0.217	0.052	0.00%
Gender = male (%)	1164 (57.6)	483 (59.6)	0.353	0.040	0.00%
Weight (kg, mean (SD))	85.80 (26.54)	87.84 (29.19)	0.074	0.073	0.68%
Height (cm, mean (SD))	169.89 (14.70)	169.85 (16.21)	0.950	0.003	20.19%
Service unit (%)			<0.001	0.671	0.00%
CCU	154 (7.6)	201 (24.8)
CSRU	429 (21.2)	47 (5.8)
MICU	770 (38.1)	372 (45.9)
SICU	297 (14.7)	84 (10.4)
TSICU	372 (18.4)	107 (13.2)
SAPS (mean (SD))	21.72 (4.64)	23.13 (4.74)	<0.001	0.301
SOFA (mean (SD))	7.18 (3.37)	8.51 (3.38)	<0.001	0.392
Elixhauser score (mean (SD))	3.51 (2.09)	4.27 (2.25)	<0.001	0.347	0.02%
APACHE III (mean (SD))	51.82 (21.58)	61.29 (23.33)	<0.001	0.421
Vasopressor use (%)	1042 (51.5)	554 (68.3)	<0.001	0.347	0
Sedative use (%)	866 (42.8)	393 (48.5)	0.007	0.113	0
Severe (%)	631 (31.2)	344 (42.4)	<0.001	0.234	0
CHF (%)	539 (26.7)	374 (46.1)	<0.001	0.413	0
AFIB (%)	624 (30.9)	252 (31.1)	0.948	0.005	0
Rennal (%)	205 (10.1)	114 (14.1)	0.004	0.120	0
Liver (%)	208 (10.3)	79 (9.7)	0.714	0.018	0
COPD (%)	119 (5.9)	46 (5.7)	0.896	0.009	0
CAD (%)	417 (20.6)	203 (25.0)	0.012	0.105	0
Malignancy (%)	121 (6.0)	40 (4.9)	0.316	0.046	0.00%
Temperature (°C, mean (SD))	36.51 (3.35)	36.63 (2.44)	0.329	0.043	7.07%
CVP (mmHg, mean (SD))	12.00 (4.32)	13.59 (4.79)	<0.001	0.349	44.89%
MAP (mmHg, mean (SD))	81.70 (18.41)	80.14 (20.23)	0.047	0.081	0.73%
Heart rate (bpm, mean (SD))	93.32 (20.62)	97.41 (22.29)	<0.001	0.190	0.24%
WBC (10³/µL, mean (SD))	13.94 (13.30)	14.91 (9.23)	0.056	0.085	0.11%
HB (g/dL, mean (SD))	10.79 (3.07)	11.16 (2.25)	0.002	0.140	0.02%
Platelet (10³/µL, mean (SD))	209.84 (120.28)	228.45 (119.33)	<0.001	0.155	0.02%
Sodium (mEq/L, mean (SD))	138.76 (5.25)	138.44 (5.41)	0.143	0.060	0.02%
Potassium (mEq/L, mean (SD))	4.23 (1.14)	4.29 (0.98)	0.233	0.051	0.02%
Bicarbonate (mEq/L, mean (SD))	23.11 (5.19)	21.71 (5.93)	<0.001	0.251	0.02%
Chloride (mEq/L, mean (SD))	106.12 (6.51)	105.24 (6.87)	0.001	0.132
pH (mean (SD))	7.34 (0.10)	7.30 (0.12)	<0.001	0.369
PO₂ (mmHg, mean (SD))	152.68 (106.89)	129.91 (89.18)	<0.001	0.231
PCO₂ (mmHg, mean (SD))	45.10 (14.56)	45.58 (14.43)	0.420	0.034	0.02%
BUN (mg/dL, mean (SD))	26.72 (21.67)	33.26 (24.45)	<0.001	0.283	0.02%
Creatinine (mg/dL, mean (SD))	1.35 (1.37)	1.69 (1.57)	<0.001	0.232
Creatinine kinase (IU/L, mean (SD))	1150.26 (6572.09)	1699.07 (10555.29)	0.096	0.062	24.87%
NT-proBNP (mEq/L, mean (SD))	25.29 (3.54)	24.44 (3.65)	<0.001	0.237	81.93%
Troponin (ng/ml, mean (SD))	0.58 (1.22)	1.16 (2.77)	<0.001	0.269	61.43%
Lactate (mmol/L, mean (SD))	2.65 (2.16)	2.80 (2.27)	0.091	0.069	5.54%
P/F ratio (mmHg, mean (SD))	162.84 (65.83)	154.70 (68.64)	0.003	0.121	0.00%
Day of lCU admission (%)			0.674	0.084	0.00%
Sunday	280 (13.8)	127 (15.7)
Monday	278 (13.7)	112 (13.8)
Tuesday	285 (14.1)	123 (15.2)
Wednesday	282 (13.9)	115 (14.2)
Thursday	293 (14.5)	118 (14.5)
Friday	305 (15.1)	112 (13.8)
Saturday	299 (14.8)	104 (12.8)

Abbreviation: CHF: congestive heart failure; AFIB: atrial fibrillation; COPD: chronic obstructive pulmonary disease; CAD: coronary heart disease; CVP: central venous pressure; MAP: mean arterial pressure; WBC: white blood cells; HB: hemoglobin; PO2: partial pressure of oxygen; PCO2: partial pressure of carbon dioxide; BUN: blood urea nitrogen; NT-proBNP: N-terminal prohormone of brain natriuretic peptide; P/F: PaO2/FiO2 ratio.

After PSM and adjusting for 41 covariates, baseline characteristics were well balanced, with all covariates having an SMD of less than 0.1 (Table 2). In addition, cohorts obtained by other methods also showed well-balanced covariates (Tables 3 and 4). Figure 2 shows the SMD before and after matching, indicating that these methods worked well.

Figure 2.

Balancing covariates using the PSM, IPTW, and CBPS models. All models showed that covariates were well balanced. Abbriviations: PSM: propensity score matching; IPTW: inverse probability of treatment weighting; CBPS: covariate balancing propensity score; SMD: standardized mean difference.

Table 2.

The covariates of the matched cohorts were balanced between the groups with and without early-TTE using propensity score matching.

Variables	Early-TTE		p-value	SMD	Missing data (%)
Variables	No (737)	Yes (737)	p-value	SMD	Missing data (%)
Age (mean (SD))	63.86 (17.06)	63.28 (16.65)	0.513	0.034	0.00%
Gender = male (%)	409 (55.5)	435 (59.0)	0.188	0.071	0.00%
Weight (kg, mean (SD))	87.06 (28.73)	87.79 (29.78)	0.633	0.025	0.68%
Height (cm, mean (SD))	169.20 (15.98)	169.68 (16.78)	0.574	0.029	20.19%
Service unit (%)			0.789	0.068	0.00%
CCU	135 (18.3)	151 (20.5)
CSRU	41 (5.6)	46 (6.2)
MICU	369 (50.1)	350 (47.5)
SICU	86 (11.7)	84 (11.4)
TSICU	106 (14.4)	106 (14.4)
SAPS (mean (SD))	23.02 (5.13)	22.83 (4.64)	0.45	0.039
SOFA (mean (SD))	8.34 (3.59)	8.30 (3.34)	0.804	0.013
Elixhauser score (mean (SD))	4.01 (2.25)	4.11 (2.19)	0.36	0.048	0.02%
APACHE III (mean (SD))	60.17 (23.41)	59.70 (22.48)	0.696	0.020
Vasopressor use (%)	481 (65.3)	484 (65.7)	0.913	0.009	0.00%
Sedative use (%)	354 (48.0)	360 (48.8)	0.794	0.016	0.00%
Severe (%)	309 (41.9)	306 (41.5)	0.916	0.008	0.00%
CHF (%)	298 (40.4)	319 (43.3)	0.291	0.058	0.00%
AFIB (%)	220 (29.9)	224 (30.4)	0.865	0.012	0.00%
Rennal (%)	97 (13.2)	97 (13.2)	1	<0.001	0.00%
Liver (%)	70 (9.5)	76 (10.3)	0.663	0.027	0.00%
COPD (%)	48 (6.5)	44 (6.0)	0.747	0.022	0.00%
CAD (%)	135 (18.3)	159 (21.6)	0.134	0.082	0.00%
Malignancy (%)	39 (5.3)	38 (5.2)	1	0.006	0.00%
Temperature (°C, mean (SD))	36.63 (2.19)	36.67 (2.50)	0.769	0.015	7.07%
CVP (mmHg, mean (SD))	13.31 (4.55)	13.27 (4.46)	0.868	0.009	44.89%
MAP (mmHg, mean (SD))	81.01 (19.41)	80.70 (20.22)	0.759	0.016	0.73%
Heart rate (bpm, mean (SD))	96.89 (22.00)	97.45 (22.35)	0.631	0.025	0.24%
WBC (10³/µL, mean (SD))	14.36 (9.77)	14.71 (9.10)	0.476	0.037	0.11%
HB (g/dL, mean (SD))	11.08 (4.19)	11.14 (2.24)	0.737	0.017	0.02%
Platelet (10³/µL, mean (SD))	227.36 (131.99)	225.86 (120.35)	0.819	0.012	0.02%
Sodium (mEq/L, mean (SD))	138.68 (5.98)	138.71 (5.31)	0.901	0.006	0.02%
Potassium (mEq/L, mean (SD))	4.25 (0.83)	4.25 (0.96)	0.846	0.010	0.02%
Bicarbonate (mEq/L, mean (SD))	22.17 (5.58)	22.10 (5.86)	0.834	0.011	0.02%
Chloride (mEq/L, mean (SD))	105.51 (7.48)	105.46 (6.85)	0.89	0.007
pH (mean (SD))	7.31 (0.11)	7.31 (0.12)	0.871	0.008
PO₂ (mmHg, mean (SD))	126.91 (83.52)	129.55 (89.69)	0.559	0.030
PCO₂ (mmHg, mean (SD))	46.01 (14.79)	45.71 (14.67)	0.699	0.020	0.02%
BUN (mg/dL, mean (SD))	33.49 (26.21)	32.32 (24.39)	0.374	0.046	0.02%
Creatinine (mg/dL, mean (SD))	1.66 (1.66)	1.63 (1.57)	0.701	0.020
Creatinine kinase (IU/L, mean (SD))	1495.46 (9592.39)	1162.96 (4440.09)	0.393	0.044	24.87%
NT-proBNP (mEq/L, mean (SD))	24.67 (3.81)	24.60 (3.70)	0.735	0.018	81.93%
Troponin (ng/ml, mean (SD))	0.70 (1.79)	0.81 (1.76)	0.243	0.061	61.43%
Lactate (mmol/L, mean (SD))	2.64 (2.28)	2.68 (2.14)	0.685	0.021	5.54%
P/F ratio (mmHg, mean (SD))	152.49 (66.80)	154.87 (68.25)	0.499	0.035	0.00%
Day of lCU admission (%)			0.992	0.047	0.00%
Sunday	112 (15.2)	109 (14.8)
Monday	100 (13.6)	100 (13.6)
Tuesday	113 (15.3)	112 (15.2)
Wednesday	96 (13.0)	105 (14.2)
Thursday	111 (15.1)	106 (14.4)
Friday	101 (13.7)	106 (14.4)
Saturday	104 (14.1)	99 (13.4)

Table 3.

The covariates of the matched cohorts were balanced between the groups with and without early-TTE using inverse probability of treatment weighting.

Variables	Early-TTE		p-value	SMD	Missing data (%)
Variables	No (679.9)	Yes (683.8)	p-value	SMD	Missing data (%)
Age (mean (SD))	63.55 (16.82)	63.22 (16.80)	0.677	0.019	0.00%
Gender = male (%)	393.0 (57.8)	396.2 (57.9)	0.954	0.003	0.00%
Weight (kg, mean (SD))	86.99 (27.65)	87.23 (29.35)	0.859	0.008	0.68%
Height (cm, mean (SD))	169.61 (14.72)	169.56 (17.17)	0.947	0.003	20.19%
Service unit (%)			0.983	0.029	0.00%
CCU	119.4 (17.6)	119.4 (17.5)
CSRU	43.7 (6.4)	44.6 (6.5)
MICU	335.6 (49.4)	331.9 (48.5)
SICU	83.8 (12.3)	83.4 (12.2)
TSICU	97.3 (14.3)	104.6 (15.3)
SAPS (mean (SD))	22.86 (5.03)	22.73 (4.62)	0.573	0.027
SOFA (mean (SD))	8.28 (3.64)	8.22 (3.37)	0.723	0.017
Elixhauser score (mean (SD))	4.09 (2.20)	4.05 (2.16)	0.697	0.018	0.02%
APACHE III (mean (SD))	59.59 (23.54)	59.05 (22.26)	0.612	0.024
Vasopressor use (%)	434.5 (63.9)	436.1 (63.8)	0.950	0.003	0.00%
Sedative use (%)	328.3 (48.3)	330.6 (48.3)	0.980	0.001	0.00%
Severe (%)	277.0 (40.7)	279.8 (40.9)	0.940	0.004	0.00%
CHF (%)	278.9 (41.0)	277.6 (40.6)	0.856	0.009	0.00%
AFIB (%)	205.0 (30.2)	201.1 (29.4)	0.734	0.016	0.00%
Rennal (%)	90.7 (13.3)	88.5 (12.9)	0.806	0.012	0.00%
Liver (%)	73.4 (10.8)	74.1 (10.8)	0.981	0.001	0.00%
COPD (%)	41.3 (6.1)	42.3 (6.2)	0.918	0.005	0.00%
CAD (%)	132.0 (19.4)	133.3 (19.5)	0.972	0.002	0.00%
Malignancy (%)	37.1 (5.5)	37.0 (5.4)	0.968	0.002	0.00%
Temperature (°C, mean (SD))	36.66 (2.43)	36.66 (2.47)	0.983	0.001	7.07%
CVP (mmHg, mean (SD))	13.18 (4.58)	13.17 (4.43)	0.977	0.001	44.89%
MAP (mmHg, mean (SD))	80.62 (19.34)	80.73 (20.46)	0.906	0.006	0.73%
Heart rate (bpm, mean (SD))	97.09 (21.98)	97.15 (22.37)	0.957	0.003	0.24%
WBC (10³/µL, mean (SD))	14.43 (11.94)	14.47 (9.10)	0.927	0.004	0.11%
HB (g/dL, mean (SD))	11.05 (3.67)	11.06 (2.23)	0.901	0.006	0.02%
Platelet (10³/µL, mean (SD))	223.77 (130.56)	224.40 (120.66)	0.916	0.005	0.02%
Sodium (mEq/L, mean (SD))	138.71 (5.86)	138.73 (5.40)	0.916	0.005	0.02%
Potassium (mEq/L, mean (SD))	4.23 (0.91)	4.24 (0.97)	0.703	0.018	0.02%
Bicarbonate (mEq/L, mean (SD))	22.24 (5.61)	22.21 (5.90)	0.914	0.005	0.02%
Chloride (mEq/L, mean (SD))	105.43 (7.28)	105.58 (6.91)	0.659	0.021
pH (mean (SD))	7.31 (0.11)	7.31 (0.12)	0.764	0.014
PO₂ (mmHg, mean (SD))	128.38 (85.53)	130.75 (91.22)	0.562	0.027
PCO₂ (mmHg, mean (SD))	45.62 (14.57)	45.73 (14.73)	0.875	0.007	0.02%
BUN (mg/dL, mean (SD))	32.50 (25.39)	31.84 (24.31)	0.582	0.026	0.02%
Creatinine (mg/dL, mean (SD))	1.65 (1.59)	1.60 (1.58)	0.554	0.029
Creatinine kinase (IU/L, mean (SD))	1349.23 (7744.18)	1335.75 (7311.43)	0.966	0.002	24.87%
NT-proBNP (mEq/L, mean (SD))	24.68 (3.71)	24.68 (3.76)	0.985	0.001	81.93%
Troponin (ng/ml, mean (SD))	0.70 (1.70)	0.72 (1.49)	0.768	0.014	61.43%
Lactate (mmol/L, mean (SD))	2.69 (2.34)	2.66 (2.09)	0.795	0.012	5.54%
P/F ratio (mmHg, mean (SD))	154.75 (68.31)	154.79 (68.33)	0.990	0.001	0.00%
Day of lCU admission (%)			1.000	0.012	0.00%
Sunday	101.7 (15.0)	102.2 (14.9)
Monday	97.3 (14.3)	95.2 (13.9)
Tuesday	100.8 (14.8)	101.1 (14.8)
Wednesday	94.1 (13.8)	95.3 (13.9)
Thursday	97.4 (14.3)	99.0 (14.5)
Friday	98.8 (14.5)	99.9 (14.6)
Saturday	89.9 (13.2)	91.1 (13.3)

Table 4.

The covariates of the matched cohorts were balanced between the groups with and without early-TTE using CBPS.

Variables	Early-TTE		p-value	SMD	Missing data (%)
Variables	No (698)	Yes (698)	p-value	SMD	Missing data (%)
Age (mean (SD))	62.96 (16.90)	63.25 (16.66)	0.748	0.017	0.00%
Gender = male (%)	399 (57.2)	400 (57.3)	1.000	0.003	0.00%
Weight (kg, mean (SD))	87.98 (27.65)	86.65 (28.27)	0.375	0.048	0.68%
Height (cm, mean (SD))	169.20 (15.98)	169.68 (16.78)	0.574	0.029	20.19%
Service unit (%)			0.609	0.088	0.00%
CCU	120 (17.2)	127 (18.2)
CSRU	41 (5.9)	44 (6.3)
MICU	362 (51.9)	337 (48.3)
SICU	86 (12.3)	84 (12.0)
TSICU	89 (12.8)	106 (15.2)
SAPS (mean (SD))	22.63 (5.03)	22.78 (4.60)	0.556	0.032
SOFA (mean (SD))	8.06 (3.67)	8.27 (3.39)	0.269	0.059
Elixhauser score (mean (SD))	4.08 (2.19)	4.04 (2.17)	0.722	0.019	0.02%
APACHE III (mean (SD))	59.40 (23.39)	59.34 (22.47)	0.963	0.002
Vasopressor use (%)	430 (61.6)	451 (64.6)	0.267	0.062	0.00%
Sedative use (%)	347 (49.7)	336 (48.1)	0.592	0.032	0.00%
Severe (%)	285 (40.8)	286 (41.0)	1.000	0.003	0.00%
CHF (%)	275 (39.4)	285 (40.8)	0.623	0.029	0.00%
AFIB (%)	205 (29.4)	206 (29.5)	1.000	0.003	0.00%
Rennal (%)	87 (12.5)	92 (13.2)	0.749	0.021	0.00%
Liver (%)	76 (10.9)	76 (10.9)	1.000	<0.001	0.00%
COPD (%)	46 (6.6)	43 (6.2)	0.827	0.018	0.00%
CAD (%)	134 (19.2)	133 (19.1)	1.000	0.004	0.00%
Malignancy (%)	38 (5.4)	38 (5.4)	1.000	<0.001	0.00%
Temperature (°C, mean (SD))	36.78 (1.78)	36.72 (2.20)	0.548	0.032	7.07%
CVP (mmHg, mean (SD))	13.07 (4.62)	13.25 (4.57)	0.469	0.039	44.89%
MAP (mmHg, mean (SD))	79.81 (18.83)	80.87 (20.37)	0.314	0.054	0.73%
Heart rate (bpm, mean (SD))	97.16 (22.26)	97.26 (22.38)	0.934	0.004	0.24%
WBC (10³/µL, mean (SD))	13.99 (8.40)	14.53 (9.17)	0.249	0.062	0.11%
HB (g/dL, mean (SD))	10.99 (2.19)	11.06 (2.23)	0.548	0.032	0.02%
Platelet (10³/µL, mean (SD))	221.40 (123.50)	225.14 (121.77)	0.569	0.030	0.02%
Sodium (mEq/L, mean (SD))	138.91 (5.81)	138.84 (5.40)	0.812	0.013	0.02%
Potassium (mEq/L, mean (SD))	4.25 (1.58)	4.23 (0.96)	0.738	0.018	0.02%
Bicarbonate (mEq/L, mean (SD))	22.30 (5.72)	22.17 (5.89)	0.681	0.022	0.02%
Chloride (mEq/L, mean (SD))	105.53 (7.25)	105.65 (6.89)	0.745	0.017
pH (mean (SD))	7.31 (0.11)	7.31 (0.12)	0.617	0.027
PO₂ (mmHg, mean (SD))	126.53 (79.84)	129.33 (90.19)	0.539	0.033
PCO₂ (mmHg, mean (SD))	46.39 (14.93)	45.55 (14.71)	0.289	0.057	0.02%
BUN (mg/dL, mean (SD))	31.81 (25.51)	31.75 (24.27)	0.965	0.002	0.02%
Creatinine (mg/dL, mean (SD))	1.60 (1.53)	1.59 (1.56)	0.926	0.005
Creatinine kinase (IU/L, mean (SD))	1649.62 (10440.78)	1271.71 (5235.45)	0.393	0.046	24.87%
NT-proBNP (mEq/L, mean (SD))	24.89 (3.79)	24.65 (3.76)	0.237	0.063	81.93%
Troponin (ng/ml, mean (SD))	0.65 (1.52)	0.70 (1.28)	0.559	0.031	61.43%
Lactate (mmol/L, mean (SD))	2.61 (2.30)	2.70 (2.14)	0.450	0.040	5.54%
P/F ratio (mmHg, mean (SD))	155.51 (68.84)	154.51 (68.29)	0.785	0.015	0.00%
Day of lCU admission (%)			0.920	0.076	0.00%
Sunday	111 (15.9)	99 (14.2)
Monday	91 (13.0)	97 (13.9)
Tuesday	108 (15.5)	103 (14.8)
Wednesday	95 (13.6)	98 (14.0)
Thursday	104 (14.9)	103 (14.8)
Friday	91 (13.0)	104 (14.9)
Saturday	98 (14.0)	94 (13.5)

Considering that patients in non-early-TTE group may also receive TTE 24 h after admission to the ICU, we compared the number of TTE used for patients in the non-early-TTE group with the total number of TTE except for the earliest one for patients in the early-TTE group. The result indicated that there was no significant difference between groups (Figure 3).

Figure 3.

There were no differences between the number of TTE used for patients in non-early-TTE group and the total number of TTE, except for the earliest one for patients in early-TTE groups.

Primary outcomes

Patients in the early-TTE group had a significantly lower 90-day mortality rate compared with those in non-early-TTE group (33.4% vs. 39.1%, p-value = 0.026). The results of Kaplan–Meier curve analysis revealed that early-TTE significantly improved the survival rate (p-value = 0.031, Figure 4). The adjusted odds ratio was 0.79 (95% CI: 0.64 − 0.98, p-value = 0.036). Sensitivity analysis based on other four different methods also showed similar results (Table 5), confirming the positive effect of early-TTE on improving the 90-day mortality in patients with moderate-to-severe ARDS.

Figure 4.

Kaplan–Meier survival analysis of propensity score-matched patients stratified by early-TTE.

Table 5.

Analysis of primary outcome based on five different models.

Method	OR	95T CI		p-value
Method	OR	2.5%	97.5%	p-value
Multivariate	0.80	0.64	0.99	0.039
PSM	0.79	0.64	0.98	0.036
IPTW	0.79	0.65	0.96	0.017
CBPS	0.79	0.64	0.99	0.039
Doubly robust estimation	0.78	0.62	0.97	0.026
Unmatched	1.18	0.99	1.40	0.058

Legend: Determination of the benefit of early-TTE after balancing covariates using different statistical methods, including ORs of Multivariate, PSM, IPTW, CBPS, and Doubly robust estimation. Abbreviation: PSM: propensity score matching; IPTW: I nverse probability of treatment weighting; CBPS: covariate balancing propensity score.

Secondary outcomes

Since the above results suggest that early-TTE can improve long-term mortality, we sought to investigate the underlying reason for this benefit. We assessed several secondary outcomes. To better illustrate the condition of the patients’ fluid in the circulation, we generated a “net-input,” which is equal to the input minus the output. As shown in Table 6, the duration of vasopressors and mechanical ventilation did not differ significantly between the two groups. However, the amount of net-input in the early-TTE group on the third day of ICU admission was significantly lower than that in the non-early-TTE group (838.57 vs. 1181.89 mL, p-value = 0.014). Similar results were also observed in the total amount of net-input (early-TTE vs. non-early-TTE: 4542.54 vs. 8025.25 mL, p-value = 0.05). Although the decrease in serum creatinine did not differ significantly between the two groups, the proportion of patients with acute kidney injury (AKI) decreased in the early-TTE group (Figure 5, 33% vs 40%, p-value <0.001), which demonstrated the correlation between early-TTE and protection of kidney. AKI were identified by assessing whether serum creatinine increased by 0.3 mg/dL during the first 3 days after ICU admission (non-early-TTE group) or after the first TTE (early-TTE group).^46,47 Serum lactate could reflect the tissue hypoxia caused by impaired gas exchange and increased oxygen demand in patients with ARDS,^48–51 and also reflect global tissue perfusion and organ dysfunction in critically ill patients.^52–55 Therefore, the decrease in serum lactate reduction was significantly higher in the early-TTE than non-early-TTE group (0.83 vs. 0.59 mmol/L, p-value = 0.009), suggesting that early-TTE is associated with improved prognosis of ARDS.

Figure 5.

Difference in the proportion of AKI between the two groups. AKI was identified by assessing a 0.3 mg/dL increase in serum creatinine during the first 3 days after ICU admission (non-early-TTE group) or after the first TTE (early-TTE group). AKI: acute kidney injury.

Table 6.

Analysis of secondary outcome in propensity score-matched patients with moderate-to-severe ARDS.

Variables	Early-TTE		p-value	SMD	Power
Variables	no	yes	p-value	SMD	Power
Ventilation-free days in 28 days (mean (SD))	17.96 (6.99)	18.33 (6.39)	0.292	0.055	0.184
Vasopressor-free days in 28 days (mean (SD))	24.37 (3.95)	24.29 (3.81)	0.707	0.02	0.066
IV net-input day 1 (mean (SD))	3277.48 (4084.14)	3025.32 (3990.95)	0.231	0.062	0.224
IV net-input day 2 (mean (SD))	2200.85 (3236.42)	2086.81 (3130.12)	0.492	0.036	0.106
IV net-input day 3 (mean (SD))	1181.89 (2890.30)	838.57 (2427.94)	0.014	0.129	0.694
IV net-input total (mean (SD))	8025.25 (40935.81)	4542.54 (25402.21)	0.05	0.102	0.501
Serum lactate reduction (mmol/L, mean (SD))	0.59 (1.70)	0.83 (1.85)	0.009	0.137	0.746
Serum creatinine reduction (mg/dL, mean (SD))	0.09 (0.89)	0.04 (0.38)	0.182	0.069	0.266
die in 90 days (%)	288 (39.1)	246 (33.4)	0.026	0.119	0.747

Legend: Net-input was significantly decreased in the early-TTE group on the third day of ICU admission and throughout the hospital stay. The reduction in serum creatinine was significantly different between the two groups, revealing the beneficial effect of early-TTE on renal function.

Discussion

The severity of the illness and long lengths of ICU stay are positively associated with substantial treatment costs, family burden, and the use of healthcare resources.⁵⁶ Given that there may be differences between long-term and short-term outcomes in patients with ARDS, short-term outcomes such as in-hospital mortality can not fully reflect long-term outcomes.⁵⁷ While a recent study described the benefit of early-TTE on short-term outcomes in ARDS patients, its impact on long-term outcomes remains unclear.¹⁸

To ensure the robustness of our study, we conducted multiple sensitivity analyses. In Tables 1–4, the SMD of 41 covariates was less than 0.1, demonstrating their balance. Although there were some patients from CCU, considering the exclusion of patients with heart failure and the excellent effect of balancing of baseline, we speculate that the bias was minimal. According to the results of analyses using balanced covariates by PSM, IPTW, and CBPS, early-TTE can significantly improve 90-day mortality, supporting the benefit of early-TTE on long-term survival outcomes in ARDS patients. Besides, we also found a significantly difference in infusion volume on day 3, indicating the adjustments in fluid management. Notably, the reduction of serum lactate was higher in early-TTE group compared to no-early-TTE group, which indicated a better prognosis. Also, the decrease in the percentage of AKI between two groups revealed that the early-TTE was correlated to improved renal function. Taken together, the benefit of early-TTE on long-term outcome may be attributable to three aspects: first, early-TTE can help clinicians identify cardiovascular disease and its related comorbidities, and then provide timely treatments. This is because recent studies have found that comorbidities and sequelae in ARDS patients are highly correlated with long-term outcomes.^30,58 In particular, cardiac injury is known to be associated with increased long-term morbidity and mortality.^59,60 TTE, as a non-invasive cardiac imaging, can provide valuable information about cardiac morphology and function and help clinicians identify potential comorbidities such as right ventricular dysfunction, infective endocarditis,⁶¹ cardiac amyloidosis,⁶² and left ventricular hypertrophy.⁶³ Second, in clinical practice, results of TTE could guide the adjustments to fluid management strategies. Recent studies have noted that fluid overload is highly associated with worse outcomes, such as increased 90-day mortality, prolonged duration of ventilation, and longer ICU and hospital stays,^64–66 whereas conservative fluid management is significantly associated with better outcomes, such as shorter duration of ventilation and shorter ICU stays.^7,8 In addition, inappropriate fluid management strategies may increase the risk of long-term cognitive impairment, which negatively affect the long-term prognosis of ARDS patients.⁶⁷ Recent studies have further revealed the heterogeneity of ARDS: ARDS patients with different phenotypes may respond differently to the same fluid management strategy.⁶⁸ In the study by Liu et al., for example, three distinct ARDS clinical phenotypes were identified, each with different clinical features and outcomes.⁶⁹ Fluid-conservative strategy was associated with improved mortality in phenotype II ARDS but worse outcome in phenotype III ARDS, supporting the need for customized and optimal fluid management in ARDS patients. TTE is currently widely used to assess fluid responsiveness, which helps to adjust fluid management strategies in a safe and non-invasive manner.^70,71 This is consistent with our finding that net-input was significantly different between early-TTE group and non-early-TTE group, supporting the notion that adjustments of fluid administration are associated with improved long-term outcomes in ARDS patients. Also, appropriate fluid management strategies can help improve hypoxia and tissue perfusion which can be revealed by serum lactate.^5,72 In our results, the early-TTE group had better serum lactate improvement than the non-early-TTE group, suggesting that TTE may help improve tissue hypoxia and perfusion by helping adjust fluid strategies, thereby improving prognosis. Third, early-TTE is beneficial to the improvement of kidney function in ARDS patients. Severe AKI is known to be significantly associated with short-term outcomes in critically ill patients.⁷³ AKI as a critical risk factor for the development of chronic kidney disease (CKD) affects the long-term prognoses of CKD patients, including those with ARDS.^74–76 This is also supported by the results of our study, as ARDS patients in the early-TTE group had a lower proportion of AKI than those in the non-early-TTE group. In other words, early-TTE was associated with improved kidney function, which may lead to better long-term outcomes. Nonetheless, our study was retrospective and observational, and therefore cannot establish a causal relationship between early-TTE and improved kidney function. There may be other confounding factors that influenced the outcomes of the patients, such as the severity of illness, the timing and dose of fluid administration, the use of renal replacement therapy, and the quality of care. Further prospective studies are needed to confirm our findings and elucidate the mechanisms by which early-TTE can prevent AKI in ARDS patients. One possible explanation is that adjustments to fluid management due to early-TTE further optimizes renal perfusion and thus prevents AKI.⁷⁷ Surprisingly, there were no significant differences between groups in terms of ventilation-free days and vasopressor-free days. We speculate that early-TTE may lead to changes in the parameters of ventilators rather than duration of use.^6,78

This study has some limitations. First, the analyzed data was extracted from the MIMIC-III database, which is a real-world dataset from one medical center. While having the advantage of a large sample size, it may still limited the generalizability of our findings. Second, the analyses of this study were based on the clinical data from 2001 to 2012 in the MIMIC-III database. Improvements in the treatment methods and concepts for patients with ARDS may also affect the outcomes. Third, although TTE has the advantages of non-invasiveness and convenience, the reproducibility of echocardiography may be relatively poor due to differences in the proficiency of echocardiography among clinicians. Furthermore, due to the lack of specific echocardiography parameters and outcomes in the MIMIC-III database, further analysis of the detailed mechanisms of this beneficial effect could not be performed. Fourth, the proportion of missing values in some covariates is large, and random forest imputation of missing values may cause bias.

In general, our results revealed that early-TTE can improve long-term outcome in ARDS patients. The underlying mechanism may be that early-TTE reveals patients’ comorbidities, thereby guiding strategies to adjust fluid management, which may lead directly or indirectly to improved serum lactate level and kidney function. Considering the limitations of our study, more large-scale prospective studies are needed to confirm the beneficial effect of early-TTE on the short-term and long-term outcomes of ARDS patients, and to explore the detailed mechanism of this benefit.

Conclusion

Early-TTE is beneficial to improve the long-term mortality of patients with moderate-to-severe ARDS. Although the detailed mechanism remains to be explored, the benefits of early-TTE may achieved through optimization of subsequent treatments, such as fluid management, decrease serum lactate, and improved kidney function.

Footnotes

Acknowledgments

We acknowledge the MIMIC-III database for providing their platforms and contributors for uploading their meaningful datasets. This manuscript has been previously posted as a preprint at Research Square on September 16, 2022.

Authors’ contributions

QL and NW conceptualized and designed this study, performed analysis and literature review. QL and SN extracted data from the MIMIC-III database. YT, WT, ZT, and NW participated in data analysis and interpretation. QL and HW draft the manuscript and revised the manuscript based on comments from other authors. All authors read and approved the final manuscript.

Availability of data and materials

The datasets presented in the current study are available in the MIMIC-III database (). The corresponding author has passed the online training and obtained access to the database (researchers certification number 9659050, record id 39674244).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Chongqing Yongchuan District Natural Science Fund, Chongqing medical scientific research project, (grant number 2020nb00503, 2020FYYX006)

Ethics approval and consent to participate

We used data from the MIMIC-III database, which had approval from the Massachusetts Institute of Technology's institutional review board (no. 0403000206). The study did not affect clinical care and de-identified all protected health information, so individual patient consent was not required.

ORCID iD

Qiuyu Liu

Author Biographies

Qiuyu Liu is a postgraduate in Critical Care Medicine. Her area of research is the management of ARDS.

Yingkui Tang is a PhD student in Biomedical Engineering. His area of research is Biotechnology and Engineering.

Wu Tao is a Professor in Critical Care Medicine. His area of research is the management of ARDS and sepsis.

Ze Tang is a Professor in Critical Care Medicine. His area of research is the management of ARDS and sepsis.

Hongjin Wang holds a Master of Medicine degree. Her area of research is critical disease focusing on ARDS.

Shiyu Nie holds a Master of Medicine degree. Her area of research is critical disease focusing on sepsis.

Nian Wang is a PhD student in Critical Care Medicine. His area of research is the management of critical diseases.

References

Bellani

Laffey

Pham

, et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries. JAMA 2016; 315: 788–800.

Meyer

Gattinoni

Calfee

. Acute respiratory distress syndrome. Lancet 2021; 398: 622–637.

Gattinoni

Carlesso

Cadringher

, et al. Physical and biological triggers of ventilator-induced lung injury and its prevention. Eur Respir J Suppl 2003; 47: 15s–25s.

Magder

Guerard

. Heart-lung interactions and pulmonary buffering: lessons from a computational modeling study. Respir Physiol Neurobiol 2012; 182: 60–70.

Casey

Semler

Rice

. Fluid management in acute respiratory distress syndrome. Semin Respir Crit Care Med 2019; 40: 57–65.

Saguil

Fargo

. Acute respiratory distress syndrome: diagnosis and management. Am Fam Physician 2020; 101: 730–738.

Wiedemann

Wheeler

Bernard

, et al. Comparison of two fluid-management strategies in acute lung injury. N Engl J Med 2006; 354: 2564–2575.

Silversides

Major

Ferguson

, et al. Conservative fluid management or deresuscitation for patients with sepsis or acute respiratory distress syndrome following the resuscitation phase of critical illness: a systematic review and meta-analysis. Intensive Care Med 2017; 43: 155–170.

Vignon

Evrard

Asfar

, et al.

Fluid administration and monitoring in ARDS: which management?

Intensive Care Med 2020; 46: 2252–2264.

10.

Gorman

O'Kane

McAuley

. Acute respiratory distress syndrome in adults: diagnosis, outcomes, long-term sequelae, and management. Lancet 2022; 400: 1157–1170.

11.

Bull

Clark

McFann

, et al. Pulmonary vascular dysfunction is associated with poor outcomes in patients with acute lung injury. Am J Respir Crit Care Med 2010; 182: 1123–1128.

12.

Mekontso Dessap

Boissier

Charron

, et al. Acute cor pulmonale during protective ventilation for acute respiratory distress syndrome: prevalence, predictors, and clinical impact. Intensive Care Med 2016; 42: 862–870.

13.

Amato

MBP

Meade

Slutsky

, et al. Driving pressure and survival in the acute respiratory distress syndrome. N Engl J Med 2015; 372: 747–755.

14.

Teboul

Saugel

Cecconi

, et al. Less invasive hemodynamic monitoring in critically ill patients. Intensive Care Med 2016; 42: 1350–1359.

15.

Vignon

. What is new in critical care echocardiography? Crit Care 2018; 22: 40.

16.

Mercado

Maizel

Beyls

, et al. Transthoracic echocardiography: an accurate and precise method for estimating cardiac output in the critically ill patient. Crit Care 2017; 21: 136.

17.

Zhang

Liu

, et al. Feasibility of focused transthoracic echocardiography in intensive care unit performed by intensivists. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue 2012; 24: 739–741.

18.

Dong

Wang

, et al. Effects of transthoracic echocardiography on the prognosis of patients with acute respiratory distress syndrome: a propensity score matched analysis of the MIMIC-III database. BMC Pulm Med 2022; 22: 247.

19.

Davidson

Rubenfeld

Caldwell

, et al. The effect of acute respiratory distress syndrome on long-term survival. Am J Respir Crit Care Med 1999; 160: 1838–1842.

20.

Chen

Hao

, et al. Long term outcomes in survivors of epidemic influenza A (H7N9) virus infection. Sci Rep 2017; 7: 17275.

21.

Mart

Ware

. The long-lasting effects of the acute respiratory distress syndrome. Expert Rev Respir Med 2020; 14: 577–586.

22.

Hashem

Hopkins

Colantuoni

, et al. Six-month and 12-month patient outcomes based on inflammatory subphenotypes in sepsis-associated ARDS: secondary analysis of SAILS-ALTOS trial. Thorax 2022; 77: 22–30.

23.

Ruhl

Huang

Colantuoni

, et al. Healthcare utilization and costs in ARDS survivors: a 1-year longitudinal national US multicenter study. Intensive Care Med 2017; 43: 980–991.

24.

Herridge

Cheung

Tansey

, et al. One-year outcomes in survivors of the acute respiratory distress syndrome. N Engl J Med 2003; 348: 683–693.

25.

Herridge

Tansey

Matté

, et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med 2011; 364: 1293–1304.

26.

Dinglas

Aronson Friedman

Colantuoni

, et al. Muscle weakness and 5-year survival in acute respiratory distress syndrome survivors. Crit Care Med 2017; 45: 446–453.

27.

Clermont

Kong

Weissfeld

, et al. The effect of pulmonary artery catheter use on costs and long-term outcomes of acute lung injury. PloS One 2011; 6: e22512.

28.

Needham

Colantuoni

Mendez-Tellez

, et al. Lung protective mechanical ventilation and two year survival in patients with acute lung injury: prospective cohort study. Br Med J 2012; 344: e2124.

29.

Sigurdsson

Sigvaldason

Gunnarsson

, et al.

Acute respiratory distress syndrome: nationwide changes in incidence, treatment and mortality over 23 years

. Acta Anaesthesiol Scand 2013; 57: 37–45.

30.

Blanchard

Godet

Pons

, et al. One-year patient outcomes based on lung morphology in acute respiratory distress syndrome: secondary analysis of LIVE trial. Crit Care 2022; 26: 159.

31.

Stokes

Roberts-Thomson

. The role of cardiac imaging in clinical practice. Aust Prescr 2017; 40: 151–155.

32.

Sarkar

Hochedlinger

. The sox family of transcription factors: versatile regulators of stem and progenitor cell fate. Cell Stem Cell 2013; 12: 15–30.

33.

Lambden

Laterre

Levy

, et al. The SOFA score-development, utility and challenges of accurate assessment in clinical trials. Crit Care 2019; 23: 374.

34.

Le Gall

Lemeshow

Saulnier

. A new simplified acute physiology score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270: 2957–2963.

35.

Elixhauser

Steiner

Harris

, et al. Comorbidity measures for use with administrative data. Med Care 1998; 36: 8–27.

36.

Wagner

Draper

Knaus

. APACHE III study design: analytic plan for evaluation of severity and outcome in intensive care unit patients. Development of APACHE III. Crit Care Med 1989; 17: S199–S203.

37.

Stekhoven

Bühlmann

. Missforest–non-parametric missing value imputation for mixed-type data. Bioinformatics 2012; 28: 112–118.

38.

Rosenbaum

Rubin

. The central role of the propensity score in observational studies for causal effects. Biometrika 1983; 70: 41–55.

39.

Robins

Hernán

Brumback

. Marginal structural models and causal inference in epidemiology. Epidemiology 2000; 11: 550–560.

40.

Hirano

Imbens

GW.

Estimation of causal effects using propensity score weighting: an application to data on right heart catheterization. Health Services and Outcomes Research Methodology 2001; 2: 259-278.

41.

Imai

Ratkovic

. Covariate balancing propensity score. J R Stat Soc Series B: Stat Methodol 2014; 76: 243–263.

42.

Feng

McSparron

Kien

, et al. Transthoracic echocardiography and mortality in sepsis: analysis of the MIMIC-III database. Intensive Care Med 2018; 44: 884–892.

43.

Cole

Hernán

. Constructing inverse probability weights for marginal structural models. Am J Epidemiol 2008; 168: 656–664.

44.

Fan

Imai

Lee

, et al. Optimal covariate balancing conditions in propensity score estimation. J Bus Econ Stat 2021. ; 41: 97–110.

45.

Ridgeway

. fastDR: Fast doubly robust estimation. R package version 1.0-1. 2021. https://github.com/gregridgeway/fastDR.

46.

Darmon

Clec'h

Adrie

, et al. Acute respiratory distress syndrome and risk of AKI among critically ill patients. Clin J Am Soc Nephrol 2014; 9: 1347–1353.

47.

Kellum

Lameire

. Diagnosis, evaluation, and management of acute kidney injury: a KDIGO summary (part 1). Crit Care 2013; 17: 204.

48.

Yang

Zhang

, et al. Lactate metabolism in human health and disease. Signal Transduct Target Ther 2022; 7: 305.

49.

Kamo

Tasaka

Suzuki

, et al. Prognostic values of the Berlin definition criteria, blood lactate level, and fibroproliferative changes on high-resolution computed tomography in ARDS patients. BMC Pulm Med 2019; 19: 37.

50.

De Backer

Creteur

Zhang

, et al. Lactate production by the lungs in acute lung injury. Am J Respir Crit Care Med 1997; 156: 1099–1104.

51.

Zhang

Zheng

, et al. Risk stratification of patients with acute respiratory distress syndrome complicated with sepsis using lactate trajectories. BMC Pulm Med 2022; 22: 339.

52.

Song

Jeon

. Peri-operative serum lactate level and postoperative 90-day mortality in a surgical ICU: a retrospective association study. Eur J Anaesthesiol 2020; 37: 31–37.

53.

Vincent

Quintairos

Silva

, et al. The value of blood lactate kinetics in critically ill patients: a systematic review. Crit Care 2016; 20: 257.

54.

Nguyen

. Lactate in the critically ill patients: an outcome marker with the times. Crit Care 2011; 15: 1016.

55.

Rhodes

Evans

Alhazzani

, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock: 2016. Intensive Care Med 2017; 43: 304–377.

56.

Bice

Carson

. Acute respiratory distress syndrome: cost (early and long-term). Semin Respir Crit Care Med 2019; 40: 137–144.

57.

Angus

Carlet

. Surviving intensive care: a report from the 2002 Brussels roundtable. Intensive Care Med 2003; 29: 368–377.

58.

Williams

Dobb

Finn

, et al. Determinants of long-term survival after intensive care. Crit Care Med 2008; 36: 1523–1530.

59.

Metkus

Guallar

Sokoll

, et al. Prevalence and prognostic association of circulating troponin in the acute respiratory distress syndrome. Crit Care Med 2017; 45: 1709–1717.

60.

Gayat

Cariou

Deye

, et al.

Determinants of long-term outcome in ICU survivors: results from the FROG-ICU study

. Crit Care 2018; 22: 8.

61.

Davido

Moussiegt

Dinh

, et al. Contribution of echocardiography in the diagnosis of definitive infective endocarditis: the infectious disease specialist's point of view. Eur J Clin Microbiol Infect Dis 2017; 36: 2329–2334.

62.

Kyrouac

Schiffer

Lennep

, et al. Echocardiographic and clinical predictors of cardiac amyloidosis: limitations of apical sparing. ESC Heart Fail 2022; 9: 385–397.

63.

Pallesen

Wagner

Lambrou

, et al. Association of hypertensive intracerebral hemorrhage with left ventricular hypertrophy on transthoracic echocardiography. J Clin Med 2020; 9: 2148.

64.

Vaara

Korhonen

Kaukonen

, et al. Fluid overload is associated with an increased risk for 90-day mortality in critically ill patients with renal replacement therapy: data from the prospective FINNAKI study. Crit Care 2012; 16: R197.

65.

van Mourik

Metske

Hofstra

, et al. Cumulative fluid balance predicts mortality and increases time on mechanical ventilation in ARDS patients: an observational cohort study. PloS One 2019; 14: e0224563.

66.

Rosenberg

Dechert

Park

, et al. Review of a large clinical series: association of cumulative fluid balance on outcome in acute lung injury: a retrospective review of the ARDSnet tidal volume study cohort. J Intensive Care Med 2009; 24: 35–46.

67.

Mikkelsen

Christie

Lanken

, et al. The adult respiratory distress syndrome cognitive outcomes study: long-term neuropsychological function in survivors of acute lung injury. Am J Respir Crit Care Med 2012; 185: 1307–1315.

68.

Famous

Delucchi

Ware

, et al. Acute respiratory distress syndrome subphenotypes respond differently to randomized fluid management strategy. Am J Respir Crit Care Med 2017; 195: 331–338.

69.

Liu

Jiang

Jia

, et al. Identification of distinct clinical phenotypes of acute respiratory distress syndrome with differential responses to treatment. Crit Care 2021; 25: 320.

70.

Boyd

Sirounis

Maizel

, et al. Echocardiography as a guide for fluid management. Crit Care 2016; 20: 274.

71.

Orme

RMLE

Oram

McKinstry

. Impact of echocardiography on patient management in the intensive care unit: an audit of district general hospital practice. Br J Anaesth 2009; 102: 340–344.

72.

Hashim

Lin

Wang

, et al. Quantifying the volume of fluid resuscitation required to normalize lactate in septic patients (cohort study). Int J Surg Open 2020; 27: 162–165.

73.

Hoste

EAJ

Bagshaw

Bellomo

, et al. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med 2015; 41: 1411–1423.

74.

Liu

Glidden

Eisner

, et al. Predictive and pathogenetic value of plasma biomarkers for acute kidney injury in patients with acute lung injury. Crit Care Med 2007; 35: 2755–2761.

75.

Chanouzas

Fallouh

, et al. Short and long-term outcome of patients with severe acute kidney injury requiring renal replacement therapy. QJM 2012; 105: 33–39.

76.

McNicholas

Rezoagli

Pham

, et al. Impact of early acute kidney injury on management and outcome in patients with acute respiratory distress syndrome: a secondary analysis of a multicenter observational study. Crit Care Med 2019; 47: 1216–1225.

77.

Joannidis

Druml

Forni

, et al.

Prevention of acute kidney injury and protection of renal function in the intensive care unit: update 2017: expert opinion of the working group on prevention, AKI section, European society of intensive care medicine

. Intensive Care Med 2017; 43: 730–749.

78.

Gupta

Hossen

Grigsby

, et al.

Risk factors for the development of acute respiratory distress syndrome in mechanically ventilated adults in Peru: a multicenter observational study

. Crit Care 2019; 23: 398.

Early transthoracic echocardiography and long-term mortality in moderate- to-severe acute respiratory distress syndrome: An analysis of the Medical Information Mart for Intensive Care database

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Materials and methods

Study design and patients

Data sources

Statistical analysis

Sensitivity analysis

Results

Propensity score matching and other methods

Primary outcomes

Secondary outcomes

Discussion

Conclusion

Footnotes

Acknowledgments

Authors’ contributions

Availability of data and materials

Declaration of conflicting interests

Funding

Ethics approval and consent to participate

ORCID iD

Author Biographies

References