Ann Thorac Surg 2002;73:714-718
© 2002 The Society of Thoracic Surgeons
Original article: cardiovascular
Factors influencing intensive care unit length of stay after surgery for acute aortic dissection type A
Daniel Hoefer, MDa,
Elfriede Ruttmann, MDa,
Markus Riha, MDa,
Wolfgang Schobersberger, MDb,
Andreas Mayr, MDb,
Guenther Laufer, MDa,
Johannes Bonatti, MD*a
a Department of Cardiac Surgery, Innsbruck University Hospital Innsbruck, Austria
b Division for General and Surgical Intensive Care Medicine, Innsbruck University Hospital, Innsbruck, Austria
Accepted for publication November 19, 2001.
* Address reprint requests to Dr Bonatti, Department of Cardiac Surgery, Innsbruck University Hospital, Anichstrasse 35, A-6020 Innsbruck, Austria
e-mail: johannes.o.bonatti{at}uibk.ac.at
 |
Abstract
|
|---|
Background. Operative mortality after acute aortic dissection type A is still high, and prolonged stay at the intensive care unit is common. Little has been documented about factors influencing the intensive care unit length of stay. The aim of this study was to determine such variables.
Methods. During a 10-year period, 67 patients (47 male, 20 female) were operated on for acute aortic dissection type A. In 42 patients (63%), an ascending aortic replacement was performed, 23 patients (34%) underwent a Bentall procedure, and 2 patients (3%) received a valve-sparing David type of operation. In 14 of these cases (20%), an additional partial or total arch replacement was performed.
Results. Hospital mortality was 9 of 67 (14%). Median postoperative intensive care unit length of stay was 5 days (range, 1 to 72 days). Intensive care unit stay was in univariate analysis significantly influenced by the following factors: age (p = 0.008), body mass index (p = 0.039), cardiopulmonary bypass time (p = 0.018), aortic cross-clamp time (p = 0.031), postoperative low cardiac output syndrome (p < 0.001), and postoperative lactate levels (p = 0.01). By multivariate analysis, age (p = 0.012), cardiopulmonary bypass time (p = 0.037), and the presence of a postoperative low cardiac output syndrome (p < 0.001) significantly influenced intensive care unit stay.
Conclusions. Stay in the intensive care unit after operation for acute aortic dissection type A seems to be determined by age, cardiopulmonary bypass time, and the postoperative presence of a low cardiac output syndrome.
|
Introduction
|
|---|
Acute aortic dissection type A is a life-threatening situation with a spontaneous mortality of 1% to 3% per hour within the first 48 hours [1]. The incidence is about 10 per 100,000 per year [2–4]; aortic dissection usually appears between 40 and 60 years of age. Common risk factors are hypertension, arteriosclerosis, connective tissue disease (eg, Marfan syndrome, Ehlers-Danlos syndrome) and iatrogenic factors (eg, cardiac operation, cardiac catheterization) [5, 6]. Patients usually experience initial chest pain similar to symptoms of myocardial infarction. Further symptoms are caused by end-organ hypoperfusion causing myocardial, cerebral, renal, intestinal, or peripheral ischemia. The most common cause of death is pericardial tamponade caused by hemorrhagic pericardial effusion or in rare cases free rupture of the ascending aorta [5, 6]. As the mortality rate is very high, immediate surgical treatment is indicated.
Owing to the severity of the disease and the extended operation, most patients have a prolonged stay in the intensive care unit (ICU) in comparison to patients undergoing routine cardiac or aortic operation [7, 8]. Quite in contrast to many investigations performed to evaluate factors influencing operative mortality after acute aortic dissection, little is known about variables influencing the ICU length of stay. It was the aim of this study to determine such factors in patients operated on for acute dissection type A at our department during the last decade.
|
Patients and methods
|
|---|
From January 1991 to August 2001, 67 patients were treated for acute aortic dissection type A in our department. Median age of patients was 56 years (range, 18 to 85 years); 47 were male and 20 female (Table 1).
Hypertension was present in 48 cases. Retrosternal chest pain and back pain were the most common presenting symptoms on admission. Stroke or transient ischemic attack occurred in 10 patients; other organ ischemia (myocardium, small bowel, upper or lower extremity) was observed in 7. Preoperative diagnosis and symptoms are shown in Table 2.
Computed tomographic scan or transesophageal echocardiography were the modalities of definitive diagnosis.
Operative technique
Operations performed were ascending aortic replacement (n = 42; 63%), Bentall procedure (n = 23; 34%), and David procedure (n = 2; 3%). Additional partial or total replacement of the aortic arch was necessary in 14 cases (20%; additional to supravalvular replacement in 9 cases and additional to Bentall procedure in 5 cases). All operations were performed through a median sternotomy. Femoral arterial cannulation was used in 54 cases; in the last 2 years axillary arterial cannulation of the right axillary artery was established (n = 13). Retrograde cerebral perfusion was used in 3 cases. Cardioplegia was applied in an antegrade or retrograde fashion at the surgeon’s preference every 20 minutes; in addition topical ice slush was used for better myocardial protection. All operations were performed using deep hypothermic cardiocirculatory arrest (median lowest rectal temperature, 18°C) for the open distal anastomosis. Biologic glue was used for approximation of the dissection membrane at the distal anastomosis in all patients and if necessary at the proximal anastomosis.
Further intraoperative data are shown in Table 3.
After the operation patients were admitted to an ICU of the local anesthesiology department with a specialized staff.
Definitions
Postoperative low cardiac output syndrome (LCO) was defined as a cardiac index lower than 2.0 L · min-1 · m-2 and the need for positive inotropic agents. Renal failure was defined as prolonged oliguria or anuria with elevated blood urea nitrogen and creatinine levels requiring forced diuresis or hemofiltration. Neurologic complications were diagnosed clinically and confirmed by cerebral computed tomography. Tracheostomy was indicated if weaning from ventilation was impossible until the fifth postoperative day. Multiorgan dysfunction syndrome was defined as severe dysfunction of two or more organ systems.
Statistical analysis
Data collection and statistical analysis were performed with MS Excel for Windows (Microsoft Corp, Redmond, WA) and SPSS for Windows (SPSS, Inc, Chicago, IL). Categorical variables are given as absolute numbers and percentages; continuous variables are shown as median and range.
To ensure Gaussian distribution ICU time was converted to logarithmic values. Pearson’s correlation coefficients were calculated to assess the univariate association between continuous variables and ICU length of stay. Univariate effects of categorical variables on ICU time were calculated using the Mann-Whitney U test. All univariately significant variables were included as independent variables into a linear regression model to assess the multivariate association with ICU length of stay. A p value less than 0.05 was assumed to be statistically significant.
|
Results
|
|---|
Thirty-day mortality was 13.5% (9 patients). An LCO occurred postoperatively in 25 patients (37%). Fifteen patients (22%) had to be reoperated on for bleeding within the first 48 hours. Ventilation was necessary for 3 days (range, 1 to 71 days), and in 12 patients a tracheostomy was performed. Fourteen patients (21%) experienced multiorgan failure. Fifteen patients (22%) needed hemofiltration in the ICU; any patient surviving recovered stable renal and pulmonary function. Postoperative stroke was diagnosed in 5 cases (7%). Four of them were milder forms of cerebral ischemia (eg, paresis of one upper or lower limb) and partially reversible; in 1 patient a preexisting cerebral ischemia was aggravated and led to death after the operation. Ischemia of an extremity was present preoperatively in 2 patients and resolved in both.
Median ICU length of stay was 5 days (range, 0.3 to 72 days). All univariately tested variables for the influence on ICU length of stay are shown in Table 4.
Age (p = 0.008), body mass index (p = 0.039), aortic cross-clamp time (p = 0.031), cardiopulmonary bypass time (CPB; p = 0.018), presence of a postoperative LCO (p < 0.001), and immediately postoperative lactate levels (p = 0.01) showed a significant univariate influence (Figs 1–3)).
View larger version (23K):
[in this window]
[in a new window]
|
Fig 1. Intensive care unit (ICU) length of stay in patients with and without postoperative low cardiac output syndrome (LCO). Intensive care unit stay was significantly longer in patients with low cardiac output syndrome. (* = extreme values; O = outlayers.)
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig 2. Intensive care unit (ICU) length of stay shown as a function of patient age. In older patients, intensive care unit stay was significantly prolonged.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig 3. Intensive care unit (ICU) length of stay shown as a function of cardiopulmonary bypass (CPB) times. In patients with longer cardiopulmonary bypass times, intensive care unit stay was significantly prolonged.
|
|
Correlations between the significant factors were calculated: There was a strong correlation between CPB and aortic cross-clamp time (Pearson’s correlation coefficient = 0.862; p < 0.001), between LCO and immediately postoperative lactate levels (Pearson’s correlation coefficient = 0.429; p = 0.001), and between cardiopulmonary bypass (CPB) time and postoperative lactate levels (Pearson’s correlation coefficient = 0.294; p = 0.023). Any other possible correlation between these factors did not reach significance.
In the following multivariate linear regression model three factors were found to have significant influence on ICU stay: age (p = 0.012), CPB time (p = 0.037), and LCO (p < 0.001). Mortality did not have a significant effect and was therefore excluded from the model.
To examine possible reasons for the high number of patients suffering from LCO (n = 25), correlations between LCO possible causative factors were calculated; none showed significant influence: preoperative tamponade (p = 0.6), time from the first symptom to operation (p = 0.75), aortic valve incompetence (p = 0.18), aortic cross-clamp time (p = 0.491), circulatory arrest time (p = 0.19), lowest rectal temperature (p = 0.23), CPB time (p = 0.137), total operation time (p = 0.09), and need for operative revision (p = 0.182).
|
Comment
|
|---|
In this study age, CPB time, and presence of a postoperative LCO proved to be significant predictors for prolonged stay on ICU after operation for acute aortic dissection type A.
Many reports in the literature are available that determine risk factors for mortality after operation for acute aortic dissection type A. Age, renal failure, preoperative hemodynamic instability, cardiopulmonary resuscitation, tamponade, lack of retrograde cerebral perfusion during operation, and prolonged hypothermic circulatory arrest time have all been found to be significant predictors of mortality in univariate and multivariate analysis [7–12].
At present little is published on factors influencing ICU length of stay. Variables reported to determine ICU stay in routine cardiac operation are type of operation, urgency of operation, reoperation, age, sex, left ventricular function, CPB time, aortic cross-clamp time, postoperative LCO, and number of blood transfusions [13–17].
In accordance with these reports, in our study age had a significant influence on ICU length of stay. It is a known fact that older patients undergoing cardiac operation have increased mortality rates and often need a prolonged stay in the ICU. Keeping the extent of the operation in mind we had expected age to have an influence on ICU stay in the present series, which was definitely demonstrated.
As shown by Michalopoulos and coworkers [13], CPB time can significantly influence ICU stay after cardiac operation. In accordance with these findings, our analysis determined a significant influence of CPB time on ICU length of stay. Patients with a long CPB time also had a long aortic cross-clamp time; there was a significant correlation between these two variables. In the multivariate analysis aortic cross-clamp time dropped out, and CPB time appears to have the more important impact on ICU length of stay.
Presence of a postoperative LCO, which was present in more than one third of the patients, also showed significant influence on ICU stay. We had speculated that reasons for the high incidence of this syndrome could have been preoperative tamponade, aortic valve incompetence, myocardial hypoperfusion because of retrograde dissection into the coronary ostia, long aortic cross-clamp times, CPB times, circulatory arrest times, and the need for surgical revision owing to bleeding. All these factors were examined for their influence on the incidence of LCO and none showed a significant effect. Other possible reasons could have been preexisting coronary artery disease, which was not specifically diagnosed by coronary angiography owing to the immediate need for operation in the present series. Another possible reason for LCO is perioperative myocardial ischemia. In our patients cardioplegia was applied every 20 minutes; in addition topical ice slush was used. With this protocol the myocardial damage should be kept to a minimum.
We therefore think that in our series LCO was caused by multiple factors, and it was not possible to detect one single variable responsible for this postoperative problem.
In univariate analysis immediate postoperative lactate levels showed a significant influence on ICU stay. There was a strong correlation between lactate levels, the presence of LCO, and CPB times. Organ ischemia could have been another reason for high lactate levels, but only a few patients showed clinical manifestations of organ ischemia. In multivariate analysis lactate levels dropped out, and we therefore conclude that elevated lactate levels were primarily caused by long CPB times and by the presence of LCO.
Obesity is a weak predictor of outcome in cardiac operation [16–18]. It has been shown to have some influence on the role of postoperative deep sternal wound infections, but prolonged ICU length of stay has so far not been described in obese cardiac surgical patients. In the present study the body mass index had a significant influence in univariate analysis, but not in the multivariate regression model. It may be speculated that obesity has led to intraoperative technical difficulties, but there was no significant correlation with aortic cross-clamp times or CPB times. However, ventilation time was on average 2 days longer in patients with body mass index more than 25 kg/m2, which could be an explanation for a prolonged ICU stay.
We had expected that the need for surgical revision owing to bleeding, and postoperative stroke had an influence on the length of stay in the ICU, but surprisingly none of these factors reached significance.
In summary, of 32 common preoperative, intraoperative, and postoperative variables in patients operated on for acute aortic dissection type A, three had a significant influence on ICU length of stay. It is concluded that the major determinants of ICU stay are age, CPB time, and postoperative LCO.
|
Acknowledgments
|
|---|
The authors thank Hanno Ulmer, MD, from the Institute for Medical Biostatistics of Innsbruck University for his assistance with statistical analysis.
|
References
|
|---|
-
Anagnostopoulos C.E., Prabhakar M.J., Kittle C.F. Aortic dissections and dissecting aneurysms. Am J Cardiol 1972;30:263-273.[Medline]
-
Svensson L.G., Crawford E.S. Aortic dissection and aortic aneurysm surgery: clinical observations, experimental investigations, and statistical analyses. Part II. Curr Probl Surg 1992;29:913-1057.[Medline]
-
Svensson L.G., Crawford E.S. Aortic dissection and aortic aneurysm surgery: clinical observations, experimental investigations, and statistical analyses. Part III. Curr Probl Surg 1993;30:1-163.[Medline]
-
Svensson L.G., Crawford E.S. Aortic dissection and aortic aneurysm surgery: clinical observations, experimental investigations, and statistical analyses. Part I. Curr Probl Surg 1992;29:817-911.[Medline]
-
Bachet J.E., Termignon J.L., Dreyfus G., et al. Aortic dissection. Prevalence, cause, and results of late reoperations. J Thorac Cardiovasc Surg 1994;108:199-205.[Abstract/Free Full Text]
-
Guilmet D., Bachet J., Goudot B., Dreyfus G., Martinelli G.L. Aortic dissection: anatomic types and surgical approaches. J Cardiovasc Surg (Torino) 1993;34:23-32.[Medline]
-
Ehrlich M., Fang W.C., Grabenwoger M., Cartes-Zumelzu F., Wolner E., Havel M. Perioperative risk factors for mortality in patients with acute type A aortic dissection. Circulation 1998;98(Suppl 2):II294-II298.
-
Miller D.C., Mitchell R.S., Oyer P.E., Stinson E.B., Jamieson S.W., Shumway N.E. Independent determinants of operative mortality for patients with aortic dissections. Circulation 1984;70(Suppl 1):I153-I164.
-
Ehrlich M.P., Ergin M.A., McCullough J.N., et al. Predictors of adverse outcome and transient neurological dysfunction after ascending aorta/hemiarch replacement. Ann Thorac Surg 2000;69:1755-1763.[Abstract/Free Full Text]
-
Kawahito K., Adachi H., Yamaguchi A., Ino T. Early and late surgical outcomes of acute type A aortic dissection in patients aged 75 years and older. Ann Thorac Surg 2000;70:1455-1459.[Abstract/Free Full Text]
-
Goossens D., Schepens M., Hamerlijnck R., et al. Predictors of hospital mortality in type A aortic dissections: a retrospective analysis of 148 consecutive surgical patients. Cardiovasc Surg 1998;6:76-80.[Medline]
-
Santini F., Pessotto R., Restivo S., Montalbano G., Gatti G., Mazzucco A. Factors influencing outcome after emergency surgical repair of acute type A aortic dissection. G Ital Cardiol 1999;29:1015-1019.[Medline]
-
Michalopoulos A., Tzelepis G., Pavlides G., Kriaras J., Dafni U., Geroulanos S. Determinants of duration of ICU stay after coronary artery bypass graft surgery. Br J Anaesth 1996;77:208-212.[Abstract/Free Full Text]
-
Doering L.V., Esmailian F., Imperial-Perez F., Monsein S. Determinants of intensive care unit length of stay after coronary artery bypass graft surgery. Heart Lung 2001;30:9-17.[Medline]
-
Tu J.V., Mazer C.D., Levinton C., Armstrong P.W., Naylor C.D. A predictive index for length of stay in the intensive care unit following cardiac surgery. Can Med Assoc J 1994;151:177-185.[Abstract]
-
Tu J.V., Jaglal S.B., Naylor C.D. Multicenter validation of a risk index for mortality, intensive care unit stay, and overall hospital length of stay after cardiac surgery. Steering Committee of the Provincial Adult Cardiac Care Network of Ontario. Circulation 1995;91:677-684.[Abstract/Free Full Text]
-
Roques F., Nashef S.A., Michel P., et al. Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19,030 patients. Eur J Cardiothorac Surg 1999;15:816-822.[Abstract/Free Full Text]
-
Stahle E., Bergstrom R., Holmberg L., Nystrom S.O., Hansson H.E. Risk factors for operative mortality and morbidity in patients undergoing coronary artery bypass surgery for stable angina pectoris. Eur Heart J 1991;12:162-168.[Abstract/Free Full Text]
Related Article
-
Invited commentary
- Marc A.A.M. Schepens
Ann. Thorac. Surg. 2002 73: 718-719.
[Extract]
[Full Text]
[PDF]
This article has been cited by other articles:
|
|
|
|
 
R. Tominaga, K. Kurisu, Y. Ochiai, A. Nakashima, M. Masuda, S. Morita, and H. Yasui
Total aortic arch replacement through the L-incision approach
Ann. Thorac. Surg.,
January 1, 2003;
75(1):
121 - 125.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
