Beruflich Dokumente
Kultur Dokumente
a r t i c l e i n f o a b s t r a c t
Keywords: Purpose: The aim of this study is to investigate the incidence, severity, and outcome of postoperative
Valvular heart surgery acute respiratory distress syndrome (ARDS), according to the Berlin definition, in isolated valvular heart surgery.
Acute respiratory distress syndrome The preoperative and perioperative predisposing factors of this complication were also identified.
Methods: A retrospective chart review was conducted on 457 patients who underwent isolated valvular
heart surgery between January 2010 and December 2012. Clinical characteristics and outcomes were collected.
The primary outcome was postoperative ARDS, according to the 2012 Berlin definition for ARDS.
Results: A total of 37 patients (8.1%) developed postoperative ARDS, with a mortality rate of 29.7%. The multivar-
iate analysis identified that age (odds ratios [ORs], 1.067, P ≤ .001), liver cirrhosis (OR, 7.159; P = .001), massive
blood transfusion (OR, 2.980; P = .005), and tricuspid valve replacement (OR, 5.197; P = .012) were indepen-
dent risk factors of postoperative ARDS. Furthermore, we have determined that the increased severity stages of
ARDS were associated with decreased postoperative survival.
Conclusions: In conclusion, postoperative ARDS, according to Berlin definition, in valvular surgery, was associated
with high in-hospital mortality. The severity of ARDS was associated with patient midterm mortality. In multivar-
iate analysis, age, liver cirrhosis, massive blood transfusion, and tricuspid valve replacement were identified as
independent risk factors of ARDS.
© 2015 Elsevier Inc. All rights reserved.
⁎ Corresponding author at: Department of Cardiothoracic and Vascular Surgery, Chang 2.1. Study design and patient population
Gung Memorial Hospital, Linkou Medical Center, No. 5 Fuxing St, Guishan District,
Taoyuan City 33305, Taiwan. Tel.: +886 3 3281200; fax: +886 3 3285818.
E-mail address: josephchen0939@yahoo.com.tw (F.-C. Tsai). This post hoc analysis of a prospective collected database was
1
Shao-Wei Chen and Chih-Hsiang Chang contributed equally to this work. approved by the instructional review board of Chang Gung Memorial
http://dx.doi.org/10.1016/j.jcrc.2015.11.002
0883-9441/© 2015 Elsevier Inc. All rights reserved.
140 S.-W. Chen et al. / Journal of Critical Care 31 (2016) 139–143
Hospital. Between January 2010 and December 2012, medical records of with a nonparametric approach. Categorical data were tested by χ 2
506 consecutive patients, who received isolated valvular heart surgery test. P value less than .05 was considered statistically significant.
in a tertiary referral hospital, were reviewed. All patient records were
anonymized and de-identified prior to analysis. We excluded patients 3. Results
who received emergent operation or perioperative extracorporeal
membrane oxygenation support, or those patients who died on the 3.1. Characteristics of the study population: ARDS vs non-ARDS group
day of surgery. In order to appraise respiratory outcomes, patients
who underwent preoperative mechanical ventilator support were also A total of 457 adult patients, with a mean age of 58.0 ± 0.7 years and
excluded. The final cohort comprised a total of 457 patients, where 48.6% (235 men and 222 women) female, were investigated in this
their detailed demographic information is listed in Fig. 1. study. A total of 37 (8.1%) patients developed postoperative ARDS (10
[27.0%] mild, 18 [48.6%] moderate, and 9 [24.3%] severe) in our cohort.
All patient characteristic are listed in Table 1. Compared with non-
2.2. Data collection and definitions
ARDS patients, the ARDS group is older and have more diabetes mellitus,
more congestive heart failure (CHF) functional class (Fc) III/IV, more liver
Clinical characteristics and demographic data were extracted from a
cirrhosis, and higher preoperative risk score. The mean values of
prospectively collected database. The primary outcome was the
EuroSCORE and Society of Thoracic Surgeons score on mortality risk
incidence, severity, and risk factors of postoperative ARDS. Based
were 6.6% ± 0.4% and 4.9% ± 0.4% for the ARDS and non-ARDS group, re-
on the Berlin definition [9], ARDS was described as the following:
spectively. The observed in-hospital mortality rate of this study was 5.0%.
(1) hypoxemia, occurring within 1 week of a known clinical insult or
Aortic valve replacement was performed in 163 (36.8%) patients.
new or worsening respiratory symptoms; (2) chest image with bilateral
The mitral procedures such as mitral valve repair and replacement
opacities that cannot be fully explained by pleural effusions, atelectasis,
were performed in 195 (42.7%) and 112 (24.5%) patients, respectively.
or nodules; and (3) clinical outcomes that are not fully explained by
Tricuspid valve repair was also performed in 151 (33%) patients.
cardiac failure or fluid overload. Also, ARDS is divided into 3 categories
The mean cardiopulmonary bypass and aortic clamping time were
that included the following: (1) mild ARDS (200 mm Hg b PaO2/FiO2
160 ± 3.0 and 109.3 ± 2.1 minutes, respectively. The detailed surgical
≤300 mm Hg with PEEP [positive end-expiratory pressure] or continu-
information and perioperative data are listed in Table 2. Operative char-
ous positive airway pressure ≥ 5 cm H2O), (2) moderate ARDS
acteristics, including the type of surgery performed, type of valve re-
(100 mm Hg b PaO2/FiO2 ≤ 200 mm Hg with PEEP ≥ 5 cm H2O), and
placement, concomitant procedures, duration of cardiopulmonary
(3) severe ARDS (PaO2/FiO2 ≤100 mm Hg with PEEP ≥5 cm H2O). Patient
bypass time, and aortic cross-clamping time, were similar between the
diagnoses were independently confirmed by 2 physicians with who
2 groups. However, the ARDS group had more cases of tricuspid valve
have reviewed the medical records.
replacement and tissue valve tricuspid replacement. Compared with
the non-ARDS group, the ARDS group had more patients who had mas-
2.3. Statistical analysis sive blood transfusions (18.3% and 4.9%; ARDS vs non-ARDS, P b .001),
defined by the transfusion of more than 4 units of packed red blood
Continuous variables were summarized by mean and SE unless cells (PRBCs). Furthermore, the scoring system results, determined
otherwise stated. The primary end point was the comparison of postop- when patients were admitted into the intensive care unit (ICU), which
erative outcome between ARDS and non-ARDS groups. Kolmogorov- included Acute Physiology, Age and Chronic Health Evaluation III
Smirnov test was used to determine the normal distribution of each (APACHE III) and the Sequential Organ Failure Assessment (SOFA),
variable. Student t test was used to compare the means of continuous were significantly higher in the ARDS group than in the non-ARDS
variables and normally distributed data; otherwise, Mann-Whitney U group (all P b .001).
test was used. Categorical data were tested using the χ 2 test or Fisher
exact test. Furthermore, risk factors of ARDS were assessed using 3.2. Postoperative outcomes
univariate analysis. Variables that were statistically significant in the
univariate analysis were included in the multivariate analysis with There were 23 (5.0%) in-hospital mortalities, where the ARDS
logistic regression. Discrimination was assessed using the area under a group had higher in-hospital mortality than did the non-ARDS group
receiver operating characteristic curve (AUROC), which was compared (29.7% and 2.9%; ARDS vs non-ARDS, P b .001). Postoperative morbidity
analysis showed that when compared with the non-ARDS group, the
ARDS group had more patients who underwent tracheostomy (16.2%
and 0.5%; ARDS vs non-ARDS, P b .001), more acute renal failures
(24.3% and 1.2%; ARDS vs non-ARDS, P b .001), and more postoperative
cerebral vascular accidents (13.5% and 1.2%; ARDS vs non-ARDS, P =
.045). Furthermore, the patients in the ARDS group also had longer ven-
tilator time, ICU, and hospital stay durations. The detailed comparison of
the different postoperative outcomes is listed in Table 3. In the subse-
quent analysis, where the patients ARDS severities were classified ac-
cording to the Berlin definition, 24%, 49%, and 27% of the patients met
the criteria of mild, moderate, and severe stages of ARDS, respectively.
The mortality rate was determined to correspond with increased sever-
ity stages of ARDS from mild (11%), moderate (27.8%), to severe (50%).
Table 1
Preoperative demographic data and clinical characteristics between the ARDS and non-ARDS groups (expressed as mean ± SE)
ALT indicates alanine transaminase; BMI, body mass index; COPD, chronic obstructive pulmonary disorder; hs-CRP, high-sensitivity C-reactive protein; NS, not significant; STS, Society of
Thoracic Surgeons.
ARDS in the univariate analysis, only age (odds ratios [ORs], 1.067; P ≤ had adequate discriminatory power for predicting ARDS in valvular
.001), liver cirrhosis (OR, 6.641; P = .005), massive blood transfusion heart surgery.
(OR, 2.980; P = .005), and tricuspid valve replacement (OR, 5.197;
P = .012) were shown to be independently associated with postopera-
tive ARDS in the multivariate analysis (Table 4). 3.5. Midterm outcome based on severity of ARDS
3.4. Scoring systems and ARDS prediction A follow-up was conducted for 380 (87.6%) surviving patients, with
a mean follow-up duration of 35.7 ± 0.8 months. Fig. 3 illustrates the
Using Hosmer-Lemeshow analysis, the SOFA and APACHE III scores, cumulative survival rate of 457 isolated valvular heart surgery patients,
determined during ICU admission, were tested for their discriminatory stratified by ARDS severity. Patient who experienced ARDS exhibited
power for predicting postoperative ARDS (Fig. 2). For ARDS risk predic- significantly lower survival rate as expected (log rank, P b .001), and
tion, AUROC analysis verified that both the APACHE III (AUROC: 0.779 ± the increasing severity stages of ARDS were associated with decreased
0.036, P b .001) and the SOFA (AUROC: 0.767 ± 0.033, P b .001) scores postoperative survival.
Table 2
Operative and perioperative data between the ARDS and non-ARDS groups (expressed as mean ± SE)
Surgical procedure
Aortic valve replacement, n (%) 163 (36.8) 156 (37.1) 12 (32.4) NS (.569)
Mechanical valve, n (%) 32 (7.0) 30 (7.1) 2 (5.4) NS (.619)
Tissue valve, n (%) 136 (29.8) 126 (30.0) 10 (27.0) NS (.705)
Aortic valve repair, n (%) 14 (3.1) 14 (3.3) 0 (0) NS (.259)
Mitral valve replacement, n (%) 112 (24.5) 100 (23.8) 12 (32.4) NS (.242)
Mechanical valve, n (%) 27 (5.9) 25 (6.0) 2 (5.4) NS (.892)
Tissue valve, n (%) 85 (18.6) 75 (17.9) 10 (27) NS (.169)
Mitral valve repair, n (%) 195 (42.7) 180 (42.9) 15 (40.5) NS (.785)
Tricuspid valve replacement, n (%) 21 (4.6) 16 (3.8) 5 (13.5) .007
Mechanical valve, n (%) 4 (0.9) 4 (1.0) 0 (0) NS (.551)
Tissue valve, n (%) 17 (3.7) 12 (2.9) 5 (13.5) .001
Tricuspid valve repair, n (%) 151 (33) 135 (32.1) 16 (43.2) NS (.169)
Concomitant procedures
Maze procedure, n (%) 134 (29.3) 122 (29.0) 12 (32.4) NS (.665)
Cardiopulmonary bypass time (min) 160.5 ± 3.0 159.0 ± 3.1 177.0 ± 11.5 NS (.104)
Aortic cross-clamping time (min) 109.3 ± 2.1 109.1 ± 2.2 112.1 ± 7.9 NS (.78)
Redo operation, n (%) 89 (19.5) 79 (8.8) 10 (27.0) NS (.226)
Postoperative massive transfusion (red blood cell transfusion N4 units), n (%) 37 (8.1) 17 (4.9) 20 (18.3) b.001
Postoperative IABP support, n (%) 8 (1.8) 6 (1.4) 2 (5.4) NS (.077)
Postoperative score
APACHE III 37.9 ± 0.7 36.6 ± 0.7 52.3 ± 2.5 b.001
SOFA 6.8 ± 0.1 6.7 ± 0.1 8.4 ± 0.3 b.001
CT indicates chest tube; IABP, intra-aortic balloon pump; NS, not significant.
142 S.-W. Chen et al. / Journal of Critical Care 31 (2016) 139–143
Table 3
Postoperative outcomes between the ARDS and non-ARDS groups (expressed as mean ± SE)
Table 4
Multivariate logistic regression analysis of risk factors for ARDS
Fig. 2. Receiver operating characteristic curves that compares the ICU scoring systems of Fig. 3. Midterm survival of the non-ARDS and ARDS groups, stratified by different severity
for postoperative ARDS prediction. stages.
because the occurrences of combined cardiogenic pulmonary edema [4] Milot J, Perron J, Lacasse Y, Letourneau L, Cartier PC, Maltais F. Incidence and predic-
tors of ARDS after cardiac surgery. Chest 2001;119:884–8.
and ARDS were not uncommon. Finally, this study was conducted at a [5] Gajic O, Dabbagh O, Park PK, Adesanya A, Chang SY, Hou P, et al. Early identification
single tertiary care medical center in Asia. Regional and ethnic differ- of patients at risk of acute lung injury: evaluation of lung injury prediction score in a
ences should be considered, and the results may not be directly extrap- multicenter cohort study. Am J Respir Crit Care Med 2011;183:462–70.
[6] Kor DJ, Lingineni RK, Gajic O, Park PK, Blum JM, Hou PC, et al. Predicting risk of post-
olated to other patient populations. Because the etiology of ARDS after operative lung injury in high-risk surgical patients: a multicenter cohort study. An-
cardiac surgery is often multifactorial, the exclusion of intraoperative esthesiology 2014;120:1168–81.
and postsurgical care factors in this study can result in the inaccurate [7] Herridge MS, Tansey CM, Matte A, Tomlinson G, Diaz-Granados N, Cooper A, et al.
Functional disability 5 years after acute respiratory distress syndrome. N Engl J
prediction of ARDS occurrence.
Med 2011;364:1293–304.
[8] Dimick JB, Chen SL, Taheri PA, Henderson WG, Khuri SF, Campbell Jr DA. Hospital
costs associated with surgical complications: a report from the private-sector Na-
6. Conclusions tional Surgical Quality Improvement Program. J Am Coll Surg 2004;199:531–7.
[9] Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, et al. Acute
To the best of our knowledge, this is the first assessment of postop- respiratory distress syndrome: the Berlin Definition. JAMA 2012;307:2526–33.
[10] Kor DJ, Warner DO, Alsara A, Fernandez-Perez ER, Malinchoc M, Kashyap R, et al.
erative ARDS, according to the Berlin definition, in isolated valvular sur- Derivation and diagnostic accuracy of the surgical lung injury prediction model. An-
gery. The principle findings of current study are as follows. (1) the esthesiology 2011;115:117–28.
incidence rate of ARDS was 8.1%, with a mortality rate that approaches [11] Koh Y. Update in acute respiratory distress syndrome. J Intensive Care 2014;2:2.
[12] Ferguson ND, Fan E, Camporota L, Antonelli M, Anzueto A, Beale R, et al. The Berlin
29.7%, in this study group. (2) The multivariate analysis identified age,
definition of ARDS: an expanded rationale, justification, and supplementary materi-
liver cirrhosis, massive blood transfusion (PRBC N4 units), and tricuspid al. Intensive Care Med 2012;38:1573–82.
valve replacement as independent risk factors for postoperative ARDS. [13] Thille AW, Esteban A, Fernandez-Segoviano P, Rodriguez JM, Aramburu JA, Penuelas
O, et al. Comparison of the Berlin definition for acute respiratory distress syndrome
(3) Risk models such as APACHE III and SOFA score can be used for
with autopsy. Am J Respir Crit Care Med 2013;187:761–7.
predicting postoperative ARDS in isolated valvular surgery patients. [14] Fowler AA, Hamman RF, Good JT, Benson KN, Baird M, Eberle DJ, et al. Adult respira-
(4) We have determined that the increase severity stage of ARDS is tory distress syndrome: risk with common predispositions. Ann Intern Med 1983;
associated with decreased postoperative survival. 98:593–7.
[15] Messent M, Sullivan K, Keogh BF, Morgan CJ, Evans TW. Adult respiratory distress
In conclusion, postoperative ARDS, based on Berlin definition, syndrome following cardiopulmonary bypass: incidence and prediction. Anaesthe-
were still exhibiting high mortality rates after isolated valvular sia 1992;47:267–8.
surgery. Age, liver cirrhosis, massive blood transfusion, and tricus- [16] Asimakopoulos G, Taylor KM, Smith PL, Ratnatunga CP. Prevalence of acute respira-
tory distress syndrome after cardiac surgery. J Thorac Cardiovasc Surg 1999;117:
pid valve replacement were identified as independent risk factors 620–1.
in the multivariate analysis. Future studies can be performed to [17] Matuschak GM, Henry KA, Johanns CA, Lechner AJ. Liver-lung interactions following
emphasize on the development of preoperative risk model that is Escherichia coli bacteremic sepsis and secondary hepatic ischemia/reperfusion inju-
ry. Am J Respir Crit Care Med 2001;163:1002–9.
specific for cardiac surgery to guide early ARDS diagnosis and prompt [18] Vlaar AP, Cornet AD, Hofstra JJ, Porcelijn L, Beishuizen A, Kulik W, et al. The effect of
clinical management. blood transfusion on pulmonary permeability in cardiac surgery patients: a prospec-
tive multicenter cohort study. Transfusion 2012;52:82–90.
[19] Guenther T, Noebauer C, Mazzitelli D, Busch R, Tassani-Prell P, Lange R. Tricuspid
References valve surgery: a thirty-year assessment of early and late outcome. Eur J Cardiothorac
Surg 2008;34:402–9 [discussion 9].
[1] Fernandez-Perez ER, Sprung J, Afessa B, Warner DO, Vachon CM, Schroeder DR, et al. [20] Kilic A, Saha-Chaudhuri P, Rankin JS, Conte JV. Trends and outcomes of tricuspid
Intraoperative ventilator settings and acute lung injury after elective surgery: a valve surgery in North America: an analysis of more than 50,000 patients from the
nested case control study. Thorax 2009;64:121–7. Society of Thoracic Surgeons database. Ann Thorac Surg 2013;96:1546–52 [discus-
[2] Kogan A, Preisman S, Levin S, Raanani E, Sternik L. Adult respiratory distress syn- sion 52].
drome following cardiac surgery. J Card Surg 2014;29:41–6. [21] Chen SW, Tsai FC, Tsai FC, Chao YK, Huang YK, Chu JJ, et al. Surgical risk and outcome
[3] Stephens RS, Shah AS, Whitman GJ. Lung injury and acute respiratory distress syn- of repair versus replacement for late tricuspid regurgitation in redo operation. Ann
drome after cardiac surgery. Ann Thorac Surg 2013;95:1122–9. Thorac Surg 2012;93:770–5.