1 Department of Anaesthesiology and Intensive Care Medicine and 2 Department of Cardiovascular Surgery, University Hospital Giessen, Rudolf-Buchheim-Strasse 7, D-35392 Giessen, Germany *Corresponding author
Accepted for publication: May 13, 2002
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Methods. The data sets of 1471 adult patients having received elective cardiac surgery with CPB were recorded using an online anaesthesia record-keeping system. Patients were judged to have required inotropic drug support if they had received one or a combination of the positive inotropic drugs, epinephrine, dobutamine and enoximone. The effects of age, height, weight, body mass index, gender, chronic heart failure, documented preoperative myocardial infarction, left main coronary artery disease, preoperative history of hypertension, chronic renal failure, diabetes mellitus, chronic obstructive pulmonary disease (COPD), preoperative medical treatment, type of surgical procedure, duration of CPB, duration of aortic clamping and reperfusion time were analysed by logistic regression for predictive power of the need for positive inotropic drugs.
Results. Of the patients, 32.4% received positive inotropic drugs in the operating theatre after weaning from CPB. The overall 30-day mortality was 2.2%. Of non-survivors, 81.8% received inotropes compared with 18.2% of survivors (P<0.01). The numbers of previous myocardial infarctions (odds ratio (OR), 2.01), congestive heart failures New York Heart Association class >2 (OR, 1.85), COPD (OR, 1.85) and age >65 yr (OR, 1.62), aortic cross clamping time of >90 min (OR, 2.32) and coronary artery bypass surgery (OR, 0.43) all represented influential factors within the logistic regression model.
Conclusion. The knowledge of these risk factors should be useful in increasing the anaesthetists vigilance in those patients most at risk for inotropic support and in providing for more timely therapeutic intervention and optimizing anaesthesia management.
Br J Anaesth 2002; 89: 398404
Keywords: heart, cardiopulmonary bypass; heart, inotropism; records, anaesthesia, computerized; risk; surgery, cardiovascular
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Previous investigators have focused only on intraoperative and cardiac factors,35 but in this study we have investigated the hypothesis that concomitant diseases and preoperative medical treatment can influence the occurrence of inotropic support after CPB. Accordingly, the main objective was to analyse the incidence and risk factors for inotropic drug support in adults undergoing cardiac surgery with CPB and to evaluate the effects of positive inotropic drugs on mortality.
![]() |
Methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In all patients, anaesthesia consisted of weight-related doses of sufentanil, midazolam and pancuronium bromide. No patient was treated with volatile anaesthetics. CPB was performed with moderate hypothermia (rectal temperature 33°C) using non-pulsatile perfusion (2.4 litres min1 m2) and membrane oxygenators (Sorin 41, Sorin, Torino, Italy). Body temperature was measured using rectal (peripheral compartments) and nasopharyngeal (central compartments) probes. Priming of the extracorporeal circuit consisted of 2000 ml of Ringers solution and 250 ml of 5% albumin solution, as well as electrolytes. Antegrade histidine tryptophane ketoglutarate (HTK) solution was introduced for cardioplegic arrest. The acidbase management followed the alpha-stat procedure, and the haematocrit was held between 20 and 30% throughout CPB.
The reperfusion period before weaning from CPB was one third of the aortic cross-clamping time. After achieving a rectal temperature 35°C, weaning from CPB was performed by a gradual filling of the heart to a systolic blood pressure of 90100 mm Hg or a maximum diastolic pulmonary artery pressure of 20 mm Hg and successively reducing pump flow without routine use of inotropes. Patients received inotropes either based on the observation of reduced cardiac contractility during and after weaning (by direct visual inspection of the right ventricle and/or transoesophageal echocardiography) or after measuring a reduced cardiac index (<2.0 litres min1 m2 measured by thermodilution), or both.
Structure query language statements were defined in order to retrospectively extract data from the database. Patients were judged to have required inotropic drug support if they had received one or a combination of the inotropes, epinephrine, dobutamine or enoximone, as a single dose or infusion. The cardiac anaesthesia risk evaluation (CARE) score7 was used for risk calculation and the 30-day mortality was evaluated.
A total of 20 patient-related variables and four operative ones were investigated for predictive power. The patient-related variables were age, height, weight, body mass index (BMI), gender, chronic heart failure (New York Heart Association (NYHA) function class at presentation), documented preoperative myocardial infarction, left main coronary artery disease, preoperative history of hypertension, chronic renal failure (permanent increase in serum creatinine >1.1 mg dl1), preoperative ejection fraction (assessed by ventriculography), diabetes mellitus (non-insulin-dependent and insulin-dependent), chronic obstructive pulmonary disease (COPD) (documented asthma, chronic bronchitis or pulmonary emphysema), and preoperative treatment with angiotensin converting enzyme inhibitors (ACEI), diuretics, digoxin, ß-adrenergic blockers, calcium-antagonists, nitroglycerin or antidysrhythmic drugs. Data for the preoperative ejection fraction were only available in 428 patients and were therefore not included in the logistic regression. The surgical variables included the type of surgical procedure (coronary artery bypass (CABG) surgery, cardiac valve surgery or both simultaneously), duration of CPB, duration of aortic cross-clamping and reperfusion time.
Data were exported from the database into the SPSS® statistics program (SPSS Software GmbH, München, Germany) in order to perform statistical analysis. The dichotomous variable, positive inotropic drugs (yes/no), was used as the target criterion. Univariate analysis was performed using the t-test for metrically scaled variables. Categorical variables were assessed for a significant association with inotropic support using either the 2 test or the exact Fisher test. Data sets were randomly divided into an evaluation (n=761) and validation (n=710) set. The evaluation data set was examined by stepwise logistic regression analysis. Logistic regression was used to filter predictor variables having a significant influence on positive inotropic drugs within a multivariate model. Furthermore, logistic regression was used to estimate the coefficients (ß) of these variables. Based on the results, the probability (score) of the event positive inotropic drugs: yes may be estimated using a logistics function. A forward stepwise algorithm (inclusion criteria: log likelihood test ratio based on maximum likelihood function) was used for developing a logistic regression model. At each step, independent variables not yet included in the equation were tested for possible inclusion. The variable with the strongest significant contribution (P<0.05) improving the model was included. Variables already included in the logistic regression equation were tested for exclusion based on the probability of a log likelihood test ratio. The analysis ended when no further variables for inclusion or exclusion were available. Variables in the logistic regression equation were defined as 1 if present and 0 if absent. Constant values were calculated as absolute values.
We used the HosmerLemeshow goodness-of-fit H statistic to evaluate the overall calibration of the model equation in the validation set. This was accomplished by testing the null hypothesis that the mean predicted and observed incidence of inotropic support was equal.8 9 For this reason, the hypothesis has to be retained for confirming the fitness of the model. The level of significance should be at least P>0.05, but preferably P>0.2 to allow an indirect control of the ß-failure. The observed and expected incidences of inotropic support are shown graphically in calibration curves with the number of patients per group.
The discriminative power of the model was assessed with a receiver operating characteristic (ROC) curve.10 The ROC curve plots the percentage of true positive values (sensitivity) based on the individual score against the percentage of false positive values (1specificity). The area under the curve indicates the accuracy of the calculated model between 0 and 1. This may be interpreted as the probability of correct patient classification in one of the two categories (positive inotropic drugs: yes or no). An area of 0.5 indicates that the predictive accuracy equates a random selection.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The mean CARE score was comparable between patients with positive inotropic drugs (2.6 (SD 0.7)) and patients without positive inotropic drugs (2.6 (0.6)). The overall 30-day mortality was 2.2%. However, 27 of 33 non-survivors (81.8%) received positive inotropic drugs (P<0.01 compared with survivors) and 449 of 1438 survivors received positive intropic drugs.
The results of the univariate analysis concerning the influence of studied variables on the occurrence of positive inotropic drugs are presented in Tables 1 and 2. Mean values of metrically scaled variables are shown in Table 1. Table 2 shows the distribution of ordinally and nominally scaled variables and their relations to the incidence of receiving positive inotropic drugs
|
|
The results of stepwise logistic regression analysis are summarized in Table 3. Among patient-related factors, the number of previous myocardial infarctions showed the strongest association (OR, 2.01), indicating a twofold increased risk per myocardial infarction for the use of positive inotropic drugs. Having a congestive heart failure NYHA class of >2 (OR, 1.85), COPD (OR, 1.85) and age >65 yr (OR, 1.62) also represented patient-related influential factors. Regarding the surgical variables within the logistic regression model, the risk increased most with an aortic cross-clamping time of >90 min (OR, 2.32), whereas CABG surgery possessed a positive influence (OR, 0.43).
|
PPID=1/(1+ez)(1)
where z=0.8524 x (CABG)+0.6141 x (congestive heart failure NYHA >2) + 0.6976 x (number of myocardial infarctions) + 0.6156 x (COPD) + 0.4823 x (age >65 yr) + 0.8421 x (aortic cross-clamping time >90 min) 1.1886.
The factors for CABG, congestive heart failure NYHA >2, COPD, age >65 yr and aortic cross-clamping time >90 min were defined as 1 if present and 0 if absent. By notating number of myocardial infarctions, the number of previous myocardial infarctions is used as the factor.
ORs were determined by a logistical function for the validation set. The resulting probabilities were used for calculating the ROC curve (Fig. 1). The area under the ROC curve, acting as the measure of accuracy, was 0.68. The 95% confidence interval is represented by area (0.640.72). The results of the HosmerLemeshow H statistic showed acceptable calibration (H=12.7; P=0.09; degrees of freedom, 7). The calibration curve (Fig. 2) plots the observed and expected positive inotropic drug frequency together with the number of categorized patients (bar diagram), depending on the probability of inotropic support
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Butterworth and colleagues3 reported that positive inotropic drugs were given in 52% of patients undergoing cardiac valve surgery, a frequency comparable to patients undergoing CABG surgery in a study from Royster and colleagues.4 Rao and colleagues5 reported an overall prevalence of low cardiac output after CABG surgery, defined as the need for a postoperative intra-aortic balloon pump or inotropic support in the intensive care unit of 9.1%. However, in contrast to Rao and colleagues, we studied a case mix (CABG surgery and/or cardiac valve surgery) in which CABG surgery correlates with a lower incidence of positive inotropic drugs. Furthermore, Rao and colleagues defined in their study that patients receiving a renal dose of dopamine do not suffer from low cardiac output. A recent study has shown that dosing dopamine based on body weight does not yield consistent, predictable plasma concentrations.12 Thus, stimulating effects of a renal dose on cardiac ß-receptors cannot be excluded. Inotropes, including dopamine at doses <5 µg kg1 min1, were not routinely administered during and after weaning from CPB at our institution. The use of positive inotropic drugs thus represented failure of our usual therapeutic regimen and indicated a poor ventricular function, at least in the first hours after CPB.
In the present study, CABG surgery was associated with a lower incidence of positive inotropic drugs than were cardiac valve surgery and combined valve and CABG surgery. The strongest independent association with positive inotropic drugs was found among factors of impaired preoperative ventricular function (number of pre-existing myocardial infarctions and anamnestic congestive heart failure NYHA class >2). A close association between congestive heart failure NYHA class >2 and mortality and morbidity after cardiac surgery has been described.7 A low preoperative ejection fraction has also been identified by other investigators as a predictor of low cardiac output after CPB in patients undergoing CABG surgery as well as in those undergoing valve surgery.3 4 In a subgroup of 428 patients, we analysed the influence of the ejection fraction. A significant influence of a depressed ejection fraction on positive inotropic drug frequency could be observed in this subgroup as well. However, the ejection fraction was not included in the multivariate analysis because it was less well documented compared with the NYHA classification of congestive heart failure.
Our results demonstrate that patients with COPD have an increased risk of needing positive inotropic drugs. Patients with COPD commonly have a mild-to-moderate increase in pulmonary artery pressure.13 An increased afterload may influence right ventricular performance after CPB. Weightman and colleagues14 have shown that chronic lung disease increases the risk of in-hospital mortality after CABG surgery.
An age of >65 yr indicates an increased risk for inotropic support in the present study. Advanced age is always accompanied by a general decline in organ function, especially because of changes in the structure and function of the heart and vasculature that will ultimately affect cardiovascular performance, even in the absence of overt coexisting diseases.15 This may explain why post-ischaemic systolic functional recovery was markedly worse in the older group of an ovine model.16
With regards to intraoperative factors, we found that an aortic cross-clamping time of >90 min is an independent predictor of post-CPB need for positive inotropic drugs. A greater need for inotropic agents during long periods of aortic cross-clamping time was also observed by Bar-El and colleagues.17 To achieve cardioplegic arrest, cold (4°C) crystalloid cardioplegia, HTK solution, was applied by hydrostatic pressure as a single shot into the aortic root or directly into the coronary ostia when necessary. On rare occasions, a further dose was given via a venous graft during CABG surgery. Using this strategy for cardioplegic arrest, a prolonged cross-clamping time can result in a decrease in adenosine triphosphate compared with blood cardioplegia.18 In a study comparing blood and crystalloid cardioplegia in patients with unstable angina, blood cardioplegia significantly reduced the incidence of peri operative myocardial infarction, low-output syndrome and operative mortality.19
One major limitation is that we have not been able to assess the impact of the quality of surgery (i.e. perioperative myocardial infarction, re-operation). The CARE score was used for risk adjustment. The overall mean CARE score was 2.6. A CARE score of 3.0 predicts a mortality rate of 2.2%,7 which is comparable to our overall 30-day mortality. However, we could not predict the need of inotropic support by using the CARE score.
To conclude, we have shown that preoperative myocardial infarction, congestive heart failure NYHA class >2, COPD, age >65 yr and aortic cross-clamping time >90 min are independent predictors of ventricular dysfunction after CPB requiring inotropic support. Using the described anaesthetic strategy, inotropes following CPB are associated with higher 30-day mortality. However, the knowledge of these risk factors should be useful in increasing the anaesthetists vigilance in those patients most at risk for inotropic support and in providing for more timely therapeutic intervention and optimizing anaesthesia management.
![]() |
Acknowledgement |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2 Mangano DT. Alteration of ventricular function during coronary artery surgery. Acta Chir Scand Suppl 1989; 550: 5762[Medline]
3 Butterworth JF, Legault C, Royster RL, Hammon JW Jr. Factors that predict the use of positive inotropic drug support after cardiac valve surgery. Anesth Analg 1998; 86: 4617[Abstract]
4 Royster RL, Butterworth JF, Prough DS, et al. Preoperative and intraoperative predictors of inotropic support and long-term outcome in patients having coronary artery bypass grafting. Anesth Analg 1991; 72: 72936[Abstract]
5 Rao V, Ivanov J, Weisel RD, Ikonomidis JS, Christakis GT, David TE. Predictors of low cardiac output syndrome after coronary artery bypass. J Thorac Cardiovasc Surg 1996; 112: 3851
6 Benson M, Junger A, Quinzio L, et al. Clinical and practical requirements of online software for anesthesia documentation an experience report. Int J Med Inf 2000; 57: 15564[ISI][Medline]
7 Dupuis JY, Wang F, Nathan H, Lam M, Grimes S, Bourke M. The cardiac anesthesia risk evaluation score: a clinically useful predictor of mortality and morbidity after cardiac surgery. Anesthesiology 2001; 94: 194204[ISI][Medline]
8 Hosmer DW, Lemeshow S. Applied Logistic Regression. New York: John Wiley and Sons Inc., 1989
9 Hosmer DW, Hosmer T, Le Cessie S, Lemeshow S. A comparison of goodness-of-fit tests for the logistic regression model. Stat Med 1997; 16: 96580[ISI][Medline]
10 Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982; 143: 2936[Abstract]
11 Fremes SE, Tamariz MG, Abramov D, et al. Late results of the Warm Heart Trial: the influence of nonfatal cardiac events on late survival. Circulation 2000; 102: III33945[Medline]
12 MacGregor DA, Smith TE, Prielipp RC, Butterworth JF, James RL, Scuderi PE. Pharmacokinetics of dopamine in healthy male subjects. Anesthesiology 2000; 92: 33846[ISI][Medline]
13 Weitzenblum E. The pulmonary circulation and the heart in chronic lung disease. Monaldi Arch Chest Dis 1994; 49: 2314[Medline]
14 Weightman WM, Gibbs NM, Sheminant MR, Whitford EG, Mahon BD, Newman MA. Drug therapy before coronary artery surgery: nitrates are independent predictors of mortality and beta-adrenergic blockers predict survival. Anesth Analg 1999; 88: 28691
15 Priebe HJ. The aged cardiovascular risk patient. Br J Anaesth 2000; 85: 76378
16 Misare BD, Krukenkamp IB, Levitsky S. Age-dependent sensitivity to unprotected cardiac ischemia: the senescent myocardium. J Thorac Cardiovasc Surg 1992; 103: 604[Abstract]
17 Bar-El Y, Adler Z, Kophit A, et al. Myocardial protection in operations requiring more than 2 h of aortic cross-clamping. Eur J Cardiothorac Surg 1999; 15: 2715
18 Catinella FP, Cunningham JN Jr, Spencer FC. Myocardial protection during prolonged aortic cross-clamping. Comparison of blood and crystalloid cardioplegia. J Thorac Cardiovasc Surg 1984; 88: 41123[Abstract]
19 Christakis GT, Fremes SE, Weisel RD, et al. Reducing the risk of urgent revascularization for unstable angina: a randomized clinical trial. J Vasc Surg 1986; 3: 76472[ISI][Medline]