Chest Email Content Delivery
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ceriani, R.
Right arrow Articles by Parodi, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ceriani, R.
Right arrow Articles by Parodi, O.
(Chest. 2003;123:1229-1239.)
© 2003 American College of Chest Physicians

Application of the Sequential Organ Failure Assessment Score to Cardiac Surgical Patients*

Roberto Ceriani, MD; Maurizio Mazzoni, MD; Franco Bortone, MD; Sara Gandini, MD; Costantino Solinas, MD; Giuseppe Susini, MD and Oberdan Parodi, MD

* From the Department of Anesthesia and ICU (Drs. Ceriani, Mazzoni, Bortone, and Solinas), Humanitas Gavazzeni, Bergamo; the Division of Epidemiology and Biostatistics (Dr. Gandini) and Department of Anesthesia and ICU (Dr. Susini), European Institute of Oncology, Milano; and Section of Milano (Dr. Parodi), CNR Clinical Physiology Institute, Niguarda Ca’ Granda Hospital, Milano, Italy.

Correspondence to: Maurizio Mazzoni, MD, Anestesia e Terapia Intensiva, Humanitas Gavazzeni, Via Mauro Gavazzeni 21, 24125 Bergamo, Italy; e-mail: maurizio.mazzoni{at}gavazzeni.it


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Objective: To assess the applicability of the sequential organ failure assessment (SOFA) score to cardiac surgical patients.

Design: Observational cohort study.

Setting: Adult cardiac surgical ICU.

Patients: Two hundred eighteen patients requiring ICU stay > 96 h.

Measurements and results: The SOFA score was calculated daily until ICU discharge. Derived SOFA variables—total maximum SOFA (TMS), {Delta}SOFA, maximum SOFA (maxSOFA), and {Delta}maxSOFA—were considered. Length of ICU stay was 8.9 ± 6.7 days (mean ± SD). The mortality rate was 11.0% in the ICU and 15.6% in the hospital. Nonsurvivors had higher TMS, {Delta}SOFA, single-organ system, and mean total scores on day 1 (9.8 ± 2.5 vs 7.8 ± 2.3, p < 0.05) and thereafter until day 10. The total SOFA score on the first 10 days of ICU stay, time, survival status, and their interaction were all significant (p < 0.001), with higher SOFA scores for nonsurvivors, and lower scores for survivors that decreased as the number of days from operation increased. Cardiovascular score on day 1 carried the highest relative risk of mortality among other systems (risk ratio [RR], 2.12; 95% confidence interval [CI], 1.31 to 3.45; p < 0.01), as did maximum cardiovascular score (RR, 2.81; 95% CI, 1.62 to 4.85; p < 0.001). A growing number of failing organs was associated with mortality, from the first to the sixth postoperative day (p < 0.05). Total score on day 1, TMS, {Delta}SOFA, maxSOFA, and {Delta}maxSOFA were reliable predictors of mortality with area under receiver operating characteristic curve of 0.71 (SE, 0.08), 0.89 (SE, 0.05), 0.86 (SE, 0.06), 0.88 (SE, 0.05), and 0.88 (SE, 0.06), respectively. Length of hospital stay was significantly associated (p = 0.05) to TMS and {Delta}SOFA and not to other SOFA scores, age, or sex.

Conclusions: The SOFA score may be used to grade the severity of postoperative morbidity in cardiac surgical patients without specific adaptations. The model identifies patients at increased risk for postoperative mortality.

Key Words: cardiac surgical procedures • critical illness • intensive care • multiple organ failure • outcome assessment • postoperative complications • severity of illness index


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
In clinical research studies, a scoring system is necessary to evaluate severity of critically ill patients in the ICU. A severity score is needed to standardize reports in order to improve the understanding of the course of disease and to allow evaluation of the impact of new treatments on outcome. It has also become apparent that morbidity is an important end point in studies dealing with patients with multiple organ failure. Estimates of morbidity serve as a reliable indicator of intensive care performance, comparison among medical centers, cost/benefit analyses, and evaluation of new therapeutic or management modalities. For this reason, the sequential organ failure assessment (SOFA) score has been created in order to take into consideration the changing severity over time of the process of organ dysfunction/failure.1 It has been claimed that the clinical complexity of a multimodal event such as the multiorgan failure syndrome needed to be described quantitatively and as objectively as possible over time.1 Therefore, the SOFA score has been designed to report morbidity and to objectively quantify the degree of dysfunction/failure of each organ daily in critically ill patients.1

Both retrospective and prospective studies showed that high SOFA scores were associated with increased mortality, and that different patient groups may acquire different patterns of organ dysfunction.2 3 4 5 6 Cardiac surgical patients constitute a group with peculiar features in which the concomitance of preoperative cardiac lesions and comorbidities, perioperative events, and use of cardiopulmonary bypass (CPB) may contribute to the development of organ dysfunction and failure in the postoperative period. Although several systems have been devised to calculate the risk of death after cardiac operations, none of them examines the composite evolution of organ failure in postoperative cardiac patients.7 8 9 10 11 12

Therefore, the objective of this study was to apply the SOFA score in critically ill cardiac surgical patients, with particular attention to the severity and time course of organ dysfunction. Furthermore, the role of the derived variables2 3 4 5 6 in discriminating between survivors and nonsurvivors was evaluated.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Subjects
The study was conducted in a seven-bed ICU of a tertiary care, teaching hospital. Since this was an epidemiologic study without intervention, informed consent was not deemed necessary by the institutional review board for human studies. Because all cardiac surgery patients are transferred from the operating room to the ICU for routine surveillance and care, those discharged within 96 h were not considered for the study. Patients who had been readmitted to ICU and those with a rapidly worsening condition leading to early death were excluded.

No patients were transferred to another ICU, or to a specialist intermediate care unit within the hospital, or to a different institution. After ICU stay, all patients were transferred to the regular ward or to an intermediate care unit according to predefined criteria. Patients suitable to be cared for in the ward had hemoglobin oxygen saturation > 94% on spontaneous breathing, with fraction of inspired oxygen (FIO2) of 0.31 and a respiratory rate < 30 breaths/min; serum creatinine < 120 µmol/L, urea < 8 mmol/L, and urine output > 800 mL/24 h; postoperative increase in serum creatinine < 60 µmol/L, urea < 4 mmol/L, and urine output same as above in the presence of chronic renal failure; no infusion of inotropic agents; no invasive arterial pressure monitoring; and no newly acquired arrhythmias.

The patients discharged to the intermediate care unit evidenced one or more of the following: hemoglobin oxygen saturation > 90% on spontaneous breathing, with FIO2 of 0.40 and respiratory rate < 30 breaths/min without requirement for invasive ventilatory support; serum creatinine > 120 µmol/L, urea > 8 mmol/L, and urine output > 800 mL/24 h; postoperative increase in serum creatinine > 60 µmol/L, urea > 4 mmol/L, and urine output same as above without need for renal replacement therapy in the presence of chronic renal failure; infusion of up to one inotropic agent (either dopamine or dobutamine <= 2 µg/kg/min); invasive monitoring of arterial pressure; or need to monitor and/or to treat newly acquired arrhythmias.

SOFA Score
The SOFA score was computed daily, starting on the first postoperative day, according to the modalities described by Vincent et al1 (Table 1 ). The score was calculated for the entire duration of ICU stay for each patient. The worst value for each organ system in each 24-h period was considered.1 Organ failure was defined as a SOFA score >= 3.1 The aggregate total maximum SOFA (TMS) score was calculated by adding the worst scores for each of the organ systems, and the maximum SOFA (maxSOFA) was defined as the highest total SOFA score recorded during the observation period. We calculated the difference between the TMS score and the total SOFA score on day 1 ({Delta}SOFA),5 and the difference between the maxSOFA and total SOFA on day 1 ({Delta}maxSOFA; Table 2 ).


View this table:
[in this window]
[in a new window]

  		Table 1.. The SOFA Score*

 

View this table:
[in this window]
[in a new window]

  		Table 2.. Definitions of Derived SOFA Variables

 
Statistical Analysis
All variables were tested for normal distribution. A nonparametric rank test, the Mann Whitney U test, was used to compare means in the case of nonnormality and noncontinuous distributed data, otherwise the Student t test was adopted. Categorical data were tested using the {chi}2. Cox proportional hazards models were applied to evaluate the effects of sex, age, total SOFA on day 1, the scores of the different organ system on day 1, maximum score of each organ system, TMS, {Delta}SOFA, maxSOFA, {Delta}maxSOFA, on the risk of hospital death. Linear regression analysis with duration of hospital stay as the response variable was performed with sex, age, SOFA for each organ system on day 1, total SOFA on day 1, TMS, {Delta}SOFA, maxSOFA, and {Delta}maxSOFA as independent variables. Log transformation of hospital stay was considered for this analysis, being this variable was nonnormally distributed. Stepwise analysis was also applied on the Cox proportional hazard model and on the linear regression model.

Mortality discrimination by SOFA score on day 1, TMS, {Delta}SOFA, maxSOFA, and {Delta}maxSOFA were assessed utilizing the area under the receiver operating characteristic (AUROC) curve.13 14 Bonferroni correction was taken into account for multiple testing problems. Demographic data and SOFA scores are expressed as mean ± SD, and a difference was considered significant at p < 0.05. Data were analyzed using S+Plus software (SPLUS 2000; MathSoft; Seattle, WA).


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Patient Demographic Data
A total of 3,282 adult cardiac surgical patients were admitted to the ICU for routine postoperative care from January 1, 1997, to August 31, 2001. The SOFA score was calculated in 218 of these patients (6.6%) who fulfilled the above-mentioned criteria of complicated surgery. Demographic data, type of operation, and mean ICU and hospital stays are reported in Table 3 . None of the patients included in the study had received a heart transplant or a mechanical ventricular assist device. The ICU mortality rate for this group of critically ill cardiac surgical patients was 11.0%, and hospital mortality was 15.6%. Age and duration of stay in the ICU and in the hospital were not statistically different between survivors and nonsurvivors.


View this table:
[in this window]
[in a new window]

  		Table 3.. Demographic Data of the Study Population*

 
SOFA Score Analysis
The analysis of the scores included only the first 10 days, since only 27.9% of patients remained in the ICU for a longer period. The mean total SOFA score was higher in nonsurvivors than in survivors on day 1 (7.8 ± 2.3 in survivors and 9.8 ± 2.5 in nonsurvivors) and for all following days until postoperative day 10 (5.8 ± 2.9 in survivors and 9.8 ± 3.3 in nonsurvivors; Fig 1 ). Mean total SOFA score for survivors showed a steady decline over time that was not observed in nonsurvivors. Looking at the same data with a repeated-measure analysis on mixed-effect model, with mean total SOFA score on the first 10 days as response variable, time, survival status, and their interaction were all significant (p < 0.001), with higher scores for nonsurvivors and lower scores for nonsurvivors as the number of days from operation increased.



View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.. SOFA scores (mean ± SD) on days 1 to 10 in survivors and nonsurvivors. *p < 0.05 between the two groups.

 
The SOFA score for each organ system on days 1 to 10 for survivors and nonsurvivors is depicted in Figure 2 . No difference was found between the respiratory scores of survivors and nonsurvivors on any postoperative day during the observation period. The coagulation scores were higher in nonsurvivors on days 4, 5, and 7, and decreased for both groups after peaking on day 2. In nonsurvivors, higher liver scores were measured for the first six postoperative days but were similar to those of survivors thereafter. Cardiovascular scores differed significantly for almost all days, and peaked for survivors on the first postoperative day and declined during ICU stay. There were few neurologic complications in the patient population as a whole, namely four patients with neurologic scores >= 3 (all nonsurvivors), and the CNS score contributed negligibly to the total SOFA score, particularly in the group of survivors. Renal system scores were higher in nonsurvivors with a significant difference from the third postoperative day to the tenth day. Renal scores had a tendency to decline over time but only in survivors. An examination of the number of organ failures of nonsurvivors (zero to four organs failed) evidenced that mortality rates increased as the number of failing organs increased (day 1, p = 0.15, degrees of freedom [df] = 1; day 2, p = 0.007, df = 1; day 3, p < 0.001, df = 1; day 4, p = 0.001, df = 1; day 5, p < 0.001, df = 1; day 6, p = 0.02, df = 1; Fig 3 ).



View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2.. SOFA scores (mean ± SD) on days 1 to 10 in survivors and nonsurvivors for each organ system. *p < 0.05 between the two groups.

 


View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3.. Mortality rate (%) and number of failing organs on days 1 to 6 with number of subjects in each group. *p < 0.05 test for trend with Bonferroni correction. N = number of.

 
Observation of the first 5 postoperative days revealed that an increasing SOFA score for any single organ was associated with rising mortality over time (Table 4 ). Cardiovascular, neurologic, and hepatic scores were significant (p < 0.05), starting from the first operative day during which cardiovascular failure occurred in 59 patients (27%), neurologic in only 2 patients (0.9%), and hepatic failure in 5 patients (2.3%). The presence of organ failure (SOFA score >= 3) and its pattern over time for each of the six components of the system was not the same in survivors and nonsurvivors (Fig 4 ). The rates of coagulation failures showed a significant difference between survivors and nonsurvivors beginning from the second postoperative day, until day 5. The rates of liver, cardiovascular, and neurologic failures were higher in nonsurvivors beginning on the first postoperative day and persisted over time. Differences in the rates of renal failure between survivors and nonsurvivors appeared later. We observed an opposite trend with respect to liver and cardiovascular failures: the percentage of nonsurvivors with liver failure increased over time, whereas the percentage of survivors and nonsurvivors with cardiovascular failure generally decreased over time. As noted, the difference between the rate of cardiovascular failures in survivors vs nonsurvivors decreased over time, and at the sixth day was no longer significant. The highest rate of nonsurvivors was in patients with respiratory and cardiovascular systems failures.


View this table:
[in this window]
[in a new window]

  		Table 4.. {chi}2 Tests and p Values for Mortality Rate vs SOFA Score, for Each Organ System on Day 1 to Day 5

 


View larger version (37K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4.. Percentages of organ failures (SOFA score >= 3) in each group and in all patients (survivors, nonsurvivors, all) for each organ system. *p < 0.05 between survivors and nonsurvivors with Bonferroni correction.

 
TMS and {Delta}SOFA were significantly (p = 0.05) associated with a longer hospital stay, with an increase of 5% (SE, 0.01) with TMS and of 8% (SE, 0.01) with {Delta}SOFA. Neither age, sex, and total SOFA on day 1, nor any organ SOFA score on day 1, nor maxSOFA or {Delta}maxSOFA were significantly associated with length of hospital stay.

Higher TMS, {Delta}SOFA, and maxSOFA and {Delta}SOFA scores of each component were associated with hospital mortality as shown in Table 5 . Nonsurvivors had significantly (p < 0.001) higher scores than survivors (Table 5) . However, the Cox proportional hazard model on risk of mortality (Table 6 ) showed that the risk of death significantly increased with total SOFA on day 1, TMS, {Delta}SOFA, maxSOFA, and {Delta}maxSOFA, all with an increase of risk ranging from 25 to 31% per unit. Among different organ system scores on day 1, cardiovascular score carried the highest significant correlation with risk of mortality, thus increasing the risk of death > 100% (risk ratio [RR], 2.12; 95% confidence interval [CI], 1.31 to 3.45). Moreover, age seemed to have a slight effect, albeit nonsignificant (p = 0.066), with older people at higher risk (Table 6) . The effects of the maximum score of each organ system considered together were tested with the Cox proportional hazards regression analysis, and showed that maximum cardiovascular score was linked to the highest significant risk of death (RR, 2.81; 95% CI, 1.62 to 4.85; Table 7 ).


View this table:
[in this window]
[in a new window]

  		Table 5.. TMS, {Delta}SOFA, and maxSOFA Scores, for the Six Organ Systems, and Corresponding {Delta}SOFA Score, for All Patients, Survivors, and Nonsurvivors*

 

View this table:
[in this window]
[in a new window]

  		Table 6.. Cox Proportional Hazards Nonstepwise Model Showing the Effect of Sex, Age, Total SOFA on Day 1, maxSOFA, {Delta}maxSOFA, TMS, {Delta}SOFA, and SOFA for Each Organ System on Day 1, on the Risk of Death*

 

View this table:
[in this window]
[in a new window]

  		Table 7.. Cox Proportional Hazards Regression Showing the Effect of the maxSOFA Score for Different Organ Systems on the Risk of Death

 
The AUROC curves showed a consistent mortality discrimination by TMS score (0.89 ± 0.05), {Delta}SOFA (0.86 ± 0.06), maxSOFA (0.88 ± 0.05), and {Delta}maxSOFA (0.88 ± 0.06) that were better than the total SOFA score on day 1 (0.71 ± 0.08). All presented a similar AUROC curve, and the difference with total SOFA score on day 1 was not statistically significant. Discriminant values estimated to maximize both the sensitivity and specificity of the data were chosen as the value on the curve closest to the top left-hand corner of the graph. This threshold was 10 for SOFA score on day 1, 11 for maxSOFA, 3 for {Delta}maxSOFA, 13 for TMS, and 2 for {Delta}SOFA.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Several scoring systems have been devised to stratify cardiac surgical patients for the risk of postoperative morbidity and mortality in order to standardize outcome data for comparative profiling of operative results among institutions or surgeons.7 8 9 10 11 12 Some models have been validated to predict the risk of mortality, while others also offer an estimate of postoperative morbidity.8 9 12

In these systems, preoperative risk factors and type of operation are used to identify high-risk patients. In particular, Higgins et al15 developed a complementary model that also takes into account intraoperative events and ICU admission physiology data. However, the authors considered only coronary artery bypass grafting alone or combined with valve procedures.15 In addition, none of these systems take into account complications arising postoperatively and their impact on ICU length of stay and mortality.

The SOFA score was designed to describe the sequence of complications in critically ill patients.1 The score is intended to objectively quantify the degree of organ dysfunction over time in order to evaluate the time course of the severity of the disease. Additionally, it allows evaluation of the function of each of six organs separately. A further objective of the system is the simplicity of the collection of the variables needed and ease of calculation of the score.

Initially, the score was not intended to be predictive of risk of mortality. Nevertheless, it was directly observed that a greater SOFA score for each organ was associated with an increasing mortality rate.1 Since its proposal, other studies reported on the good prognostic performance of the model and of its derived scores.2 3 4 5 6 The system was first retrospectively developed and prospectively validated in medical/surgical ICUs.1 2

Furthermore, its descriptive and prognostic performance was confirmed in a subgroup of 181 trauma patients of the original data set, retrospectively.4 Janssens et al3 evaluated the SOFA score in medical cardiovascular patients and concluded that the scoring system described the degree and progression of organ dysfunction in their population. They also found that the model is related to outcome and duration of ICU stay.

We studied the capability of the SOFA score to describe the degree of severity of complicated postoperative course over time in a group of cardiac surgical patients. In particular, we focused our attention on patients requiring prolonged postoperative treatment, since they allow for better assessment of sequential daily scoring and because they impose a greater demand on ICU and hospital resources. Additionally, we tested the reliability of its derived variables in discriminating survivors from nonsurvivors.

As expected, cardiac dysfunction and failure occurred more frequently than in other organs, reflecting the degree of perioperative cardiac damage of this population. The highest cardiovascular scores were reached on postoperative day 1, and persistently high cardiovascular scores were observed in nonsurvivors. Since scoring of severity of cardiovascular dysfunction/failure is based on treatment criteria and not on direct measures of function, it lends itself to subjective or local practice bias. Despite this inherent limitation, the degree of cardiac impairment was significantly correlated to the risk of death from the first postoperative day as well as over time. It should be noted that in our study this limiting factor may have been less significant in contrast to large multicenter or multinational studies. Moreover, cardiac dysfunction/failure categories are significantly spaced to reduce the impact of local or personal bias on the scores.1

After cardiac surgery, inotropic agents are frequently administered regardless of the occurrence of hypotensive episodes. However, in our study, cardiovascular scores were not the result of this practice; rather, they described exclusively patients with a prolonged need for inotropic support or ICU resources. Also, the use of a mechanical ventricular assist device in patients with severe cardiac failure has not been considered by the scoring system.

Hypoxemia (defined by low PaO2/FIO2 ratio) after CPB is not uncommon and is usually short-lived with minimal effect on postoperative course.16 Postoperative derangements of respiratory function may be explained by common causes, including microatelectasis and macroatelectasis with increased shunt, alterations in chest wall mechanics related to sternotomy, and pleural opening.16 17 Other factors are related to changes in the capillary bed and pulmonary parenchyma secondary to left ventricular dysfunction or to primary alveolar-endothelial injury mediated by cellular response mechanisms activated during or immediately after CPB.16 17 18

Respiratory failure was frequent in this population likely due to these reasons. However, high respiratory scores correlated with nonsurvival several days after operation.

Together with preoperative functional reserve, postoperative low cardiac output is the most important cause of acute renal dysfunction,19 and superposition of further hemodynamic insult on ischemic kidneys is usually necessary for acute renal failure to occur.20 Mean renal scores and incidence of renal failure were higher in nonsurvivors late in the postoperative period, suggesting that renal dysfunction was a secondary event in this population. Since continuous extracorporeal renal replacement treatment results in lowered serum creatinine levels, the incidence of renal failure (SOFA score >= 3) may be underreported by the renal component of the SOFA score. It has been suggested that scoring should be based on requirement for renal support.21 However, the definition of acute renal failure and its severity are still a matter of debate,22 and criteria for institution, termination, and the modality of renal replacement treatment are not uniform.

Janssens et al3 suggested that serum bilirubin levels may not be the ideal indicator of liver dysfunction in patients with severe left and right cardiac failure and consequent liver congestion. Early after cardiac surgery, hyperbilirubinemia can be due to surgical factors unrelated to hepatic dysfunction such as blood transfusions, intraoperative and postoperative hemolysis, long operative times, use of intra-aortic balloon counterpulsation, and low-output syndrome.23 24 However, late postoperative rise of serum bilirubin is associated with increased mortality,23 and in our population a growing incidence of liver failure over time was observed in nonsurvivors. Thus, the system may reliably report the severity of hepatic impairment intervening during ICU stay.

As pointed out by Shime et al,25 several factors related to cardiac surgical procedures (eg, hemodilution, consumption, and systemic anticoagulation with heparin for CPB) can worsen coagulation scores. High coagulation scores observed by us early after operation may indeed be affected by factors other than hematopoietic system failure such as a prolonged CPB time or continuing systemic heparinization during intra-aortic balloon counterpulsation.

Therefore, platelet counts may not be the best indicator of coagulatory dysfunction early after cardiac surgical procedures. However, in our study a higher rate of coagulation failure was observed in nonsurvivors from the second postoperative day. Even though the SOFA score was initially intended for sepsis related organ dysfunction/failure, this component of the scoring system may denote severity of illness in cardiac surgical patients as well.

Although severe neurologic complications such as deep levels of coma were infrequent in our population, they were invariably related to a poor outcome. Neurologic assessment in critically ill patients is frequently made difficult or impossible by the use of sedatives and paralyzing agents, and its role in the SOFA scoring system has been argued.25 In our experience, sedation was temporarily discontinued to evaluate mental status with the Glasgow coma scale in accordance with the method described by Kress et al.26 Alternatively, the neurologic score was computed retrospectively when sedatives were stopped. However, it may still be impossible to evaluate the neurologic score of dying patients when the use of sedatives does not cease. Since the neurologic component of the SOFA score is based on a coma scale, it overlooks severe neurologic conditions such as stroke or neuromuscular dysfunction/failure that may be present in critically ill patients.

The preeminence of the cardiovascular system on the entire score was evident in this group of patients, and the degree of cardiac dysfunction correlated to the outcome beginning on the first postoperative day and thereafter, more than that of any other single organ. Our observations in cardiac surgical subjects can be compared to those reported by Janssens et al3 in cardiovascular medical patients. In their study, they also found that cardiovascular score on day 1 and maximum cardiovascular score were associated with risk of death.3 Therefore, it seems unnecessary to modify the scoring method in order to take into account the exceedingly important role played by the cardiovascular system in this group of homogeneous patients.

Cardiovascular scores, both initial and maximum, reflected the specific pattern of organ damage in complicated cardiac surgical patients. In these patients, both the extent of cardiac damage early after operation and the additional functional loss of other organs over time were correlated to a fatal outcome.

As reported by others, a significant relationship between the number of failing organs and mortality was observed.3 Increasing number of organ failures (ie, a SOFA score >= 3) was associated with a higher mortality rate from the first postoperative day and thereafter. In patients with equal scores, the weight of organ failure, as opposed to the simultaneous presence of organ dysfunction in several systems, is of utmost importance in the determination of outcome.

The derived SOFA variables were predictive of mortality in our study, as it has been found in other studies.3 5 6 It has been suggested that the total score on the first day is representative of the conditions of the patient at admission to ICU; the TMS sums up the cumulative organ dysfunction observed in the patient during ICU stay; and the {Delta}SOFA can be used as a measure of the development of dysfunction, the degree of improvement, or its lack, during ICU care.5 6 In our study, the total SOFA score on day 1 is related to the first comprehensive postoperative assessment of organ function and recapitulates patient status before operation and the consequences of perioperative events.

However, the duration of hospital stay was related to the TMS and {Delta}SOFA and not to the other derived variables or to any single organ system score. Therefore, these latter variables may be better indicators of the severity of complications arising during prolonged ICU stay than initial SOFA and maxSOFA scores.

As outlined by Moreno et al,5 both the initial degree of severity and the summation of organ dysfunction/failure developing during ICU stay play an important role on outcome. In our study, survival status was associated with decreasing scores as the days from operation increased, whereas mortality was associated with persistently high scores. Accordingly, Ferreira et al6 reported a decrease in mortality rates with decreasing scores in critically ill patients and emphasized the advantage of describing the time factor together with the severity of the disease in the evaluation of patients’ response to treatment. They also observed an increasing mortality rate for persistently high or worsening initial scores with threshold values similar to ours.

In conclusion, although peculiarities of cardiac surgical operations and particularly the use of CPB introduce important confounding factors on the parameters utilized to calculate the severity of organ dysfunction with the SOFA score, our data show that the scoring system can reliably and objectively describe the ongoing course of organ impairment after cardiac surgery, without specific adaptations. Furthermore, initial scores and their daily follow-up outlined the evolution of the process of disease and ICU care, as well as the relevance of each single organ. A further advantage of this method is consistency in mortality prediction, thus confirming findings by other authors that the SOFA score can be applied to specific groups of critically ill patients.3 4 However, we support the view that the system is a powerful tool for group analysis and not a clinical parameter for decision making on individual patients. The adoption of a standardized method to quantify postoperative morbidity may offer a great advantage for forthcoming studies in this field.


   Footnotes
 
Abbreviations: AUROC = area under receiver operating characteristic; CI = confidence interval; CPB = cardiopulmonary bypass; df = degrees of freedom; FIO2 = fraction of inspired oxygen; maxSOFA = maximum sequential organ failure assessment; RR = risk ratio; SOFA = sequential organ failure assessment; TMS = total maximum sequential organ failure assessment

Received for publication March 20, 2002. Accepted for publication July 8, 2002.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Vincent, JL, Moreno, R, Takala, J, et al (1996) The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure: on behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 22,707-710[ISI][Medline]
  2. Vincent, JL, de Mendonça, A, Cantraine, F, et al Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study; Working Group on "Sepsis-Related Problems" of the European Society of Intensive Care Medicine. Crit Care Med 1998;26,1793-1800[ISI][Medline]
  3. Janssens, U, Graf, C, Graf, J, et al Evaluation of the SOFA score: a single-center experience of a medical intensive care unit in 303 consecutive patients with predominantly cardiovascular disorders. Intensive Care Med 2000;26,1037-1045[CrossRef][ISI][Medline]
  4. Antonelli, M, Moreno, R, Vincent, JL, et al Application of SOFA score to trauma patients. Intensive Care Med 1999;25,389-394[CrossRef][ISI][Medline]
  5. Moreno, R, Vincent, JL, Matos, R, et al The use of maximum SOFA score to quantify organ dysfunction/failure in intensive care: results of a prospective, multicentre study; Working Group on Sepsis-Related Problems of the ESICM. Intensive Care Med 1999;25,686-696[CrossRef][ISI][Medline]
  6. Ferreira, FL, Bota, DP, Bross, A, et al Serial evaluation of the SOFA score to predict outcome in critically ill patients. JAMA 2001;286,1754-1758[Abstract/Free Full Text]
  7. Parsonnet, V, Dean, D, Bernstein, AD A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79,I3-I12
  8. Tuman, KJ, McCarthy, RJ, March, RJ, et al Morbidity and duration of ICU stay after cardiac surgery: a model for preoperative risk assessment. Chest 1992;102,36-44[Abstract/Free Full Text]
  9. Higgins, TL, Estafanous, FG, Loop, FD, et al Stratification of morbidity and mortality outcome by preoperative risk factors in coronary artery bypass patients: a clinical severity score. JAMA 1992;267,2344-2348[Abstract]
  10. Tu, JV, Mazer, CD, Levinton, C, et al A predictive index for length of stay in the intensive care unit following cardiac surgery. Can Med Assoc J 1994;151,177-185[Abstract]
  11. Nashef, SA, Roques, F, Michel, P, et al European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16,9-13[Abstract/Free Full Text]
  12. Dupuis, JY, Wang, F, Nathan, H, et al The cardiac anesthesia risk evaluation score: a clinically useful predictor of mortality and morbidity after cardiac surgery. Anesthesiology 2001;94,194-204[CrossRef][ISI][Medline]
  13. Zweig, MH, Campbell, G Receiver-operating characteristics (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 1993;39,561-577[Abstract/Free Full Text]
  14. van der Schouw, YT, Straatman, H, Verbeek, AL ROC curves and the areas under them for dichotomized tests: empirical findings for logistically and normally distributed diagnostic test results. Med Decis Making 1994;14,374-381[Abstract/Free Full Text]
  15. Higgins, TL, Estafanous, FG, Loop, FD, et al ICU admission score for predicting morbidity and mortality risk after coronary artery bypass grafting. Ann Thorac Surg 1997;64,1050-1058[Abstract/Free Full Text]
  16. Weiss, YG, Merin, G, Koganov, E, et al Postcardiopulmonary bypass hypoxemia: a prospective study on incidence, risk factors, and clinical significance. J Cardiothorac Vasc Anesth 2000;14,506-513[CrossRef][ISI][Medline]
  17. Shenkman, Z, Shir, Y, Weiss, YG, et al The effects of cardiac surgery on early and late pulmonary functions. Acta Anaesthesiol Scand 1997;41,1193-1199[ISI][Medline]
  18. Cohen, AJ, Moore, P, Jones, C, et al Effect of internal mammary harvest on postoperative pain and pulmonary function. Ann Thorac Surg 1993;56,1107-1109[Abstract]
  19. Abrahamov, D, Tamariz, M, Fremes, S, et al Renal dysfunction after cardiac surgery. Can J Cardiol 2001;17,565-570[ISI][Medline]
  20. Hilberman, M, Myers, BD, Carrie, BJ, et al Acute renal failure following cardiac surgery. Thorac Cardiovasc Surg 1979;77,880-888
  21. Ryan, TA, Rady, MY Scoring organ dysfunction [letter]. Chest 1998;114,942[Free Full Text]
  22. Bellomo, R, Kellum, J, Ronco, C Acute renal failure: time for consensus. Intensive Care Med 2001;27,1685-1688[CrossRef][ISI][Medline]
  23. Wang, MJ, Chao, A, Huang, CH, et al Hyperbilirubinemia after cardiac operation: incidence, risk factors, and clinical significance. Thorac Cardiovasc Surg 1994;108,429-436
  24. Michalopoulos, A, Alivizatos, P, Geroulanos, S Hepatic dysfunction following cardiac surgery: determinants and consequences. Hepatogastroenterology 1997;44,779-783[Medline]
  25. Shime, N, Kageyama, K, Ashida, H, et al Application of modified sequential organ failure assessment score in children after cardiac surgery. J Cardiothorac Vasc Anesth 2001;15,463-468[CrossRef][ISI][Medline]
  26. Kress, JP, Pohlman, AS, O’Connor, MF, et al Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342,1471-1477[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 ChestHome page
A. N. Aggarwal, R. Agarwal, D. Gupta, and S. K. Jindal
Nonpulmonary Organ Dysfunction and Its Impact on Outcome in Patients With Acute Respiratory Failure
Chest, September 1, 2007; 132(3): 829 - 835.
[Abstract] [Full Text] [PDF]

Home page
 HeartHome page
R De Maria, M Mazzoni, M Parolini, D Gregori, F Bortone, V Arena, and O Parodi
Predictive value of EuroSCORE on long term outcome in cardiac surgery patients: a single institution study
Heart, June 1, 2005; 91(6): 779 - 784.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (11)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ceriani, R.
Right arrow Articles by Parodi, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ceriani, R.
Right arrow Articles by Parodi, O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS