Asian Cardiovasc Thorac Ann 2006;14:114-118
© 2006 Asia Publishing EXchange Ltd
Determinants of Morbidity and Intensive Care Unit Stay after Coronary Surgery
Lusine Abrahamyan, MD,
Anahit Demirchyan, MD,
Michael E Thompson, DrPH,
Hrair Hovaguimian, MD
Center for Health Services Research and Development, American University of Armenia, Nork Marash Medical Center, Yerevan, Armenia
For reprint information contact: Lusine Abrahamyan, MD Tel: 374 165 5980 Fax: 374 151 2566 Email: lusine.abrahamyan{at}utoronto.ca, Center for Health Services Research and Development, American University of Armenia, 40 M. Baghramyan Str., Yerevan 375019, Armenia.
 |
ABSTRACT
|
|---|
The study evaluated rates and determinants of hospital morbidity, serious morbid events, and prolonged intensive care unit stay associated with isolated coronary artery bypass. The medical records of 391 patients undergoing isolated coronary artery bypass at our center during 2003 were reviewed. The observed crude hospital mortality rate was 2.05%, similar to the EuroSCORE predicted mortality rate of 2.34%. Arrhythmia was the most frequent postoperative complication (17.6%). The serious hospital morbidity rate was 5.9%. The final logistic regression model of serious morbid events identified the following predictors: drug allergy, diabetes, and EuroSCORE. Prolonged intensive care unit stay (
3 days) was observed in 9.5% of patients. Multivariable logistic regression analysis revealed age, preoperative rhythm disturbances, previous cardiac operation, and hypertension as independent predictors of prolonged intensive care unit stay. The rates of hospital mortality, morbidity, and prolonged intensive care unit stay were comparable to those of other major international cardiac surgery centers. These data can be used as a benchmark for further self- and peer-assessment quality improvement activities.
|
INTRODUCTION
|
|---|
Nork Marash Medical Center is a unique cardiac surgery hospital with inpatient and outpatient services for both adult and pediatric populations, where more than 700 cardiac operations are performed annually. Since 2001, in collaboration with the American University of Armenia Center for Health Services Research and Development, a quality assurance project has been implemented. Given the efforts in continuous quality improvement, evaluating the outcomes of coronary artery bypass grafting (CABG) in terms of hospital mortality, morbidity, and intensive care unit (ICU) stay became imperative. This study evaluated isolated CABG procedure-associated hospital morbidity, serious morbid events, and prolonged ICU stay rates, and assessed the determinants of the last two measurements.
|
PATIENTS AND METHODS
|
|---|
The study utilized a retrospective review of medical records and the hospital surgical database, including all isolated CABG operations from January 1, 2003 through December 31, 2003. The simultaneous use of different sources for one patient allowed double-checking of information and minimized the incidence of incomplete data. Patients who underwent CABG in association with other procedures were excluded. The hospital staff discussed and stated the definitions of both dependent and independent variables included in the study before its initiation. The study involved secondary data analysis and possessed no risk to patients. Furthermore, the study was considered part of the internal evaluation process within the scope of the quality assurance project at the hospital. However, approval to access data was obtained from the Medical Board of the hospital prior to initiation of the study. All records were reviewed in the hospital to ensure patient confidentiality.
For the outcomes of serious morbid events and prolonged ICU stay, univariate logistic regression analyses were performed to find significantly associated variables. Continuous variables were plotted using a locally weighted smoothing scatter plot technique to assess linearity on the logistic scale. In the next step, multivariable logistic regression analyses were carried out to identify the independent predictors. The final models were developed using the likelihood ratio test and assessed by different diagnostic methods (area under the Receiver Operating Characteristic curve and Hosmer-Lemeshow goodness-of-fit test).
|
RESULTS
|
|---|
There were 394 patients who underwent isolated CABG, however the study included 391 as 3 had to be excluded due to incomplete data. The preoperative patient profile was described by frequencies for categorical data and by means and standard deviations for continuous data (Table 1
). The majority of the study population was male and had a previous myocardial infarction, hypertension, three-vessel disease, and a good ejection fraction. The mean EuroSCORE or the predicted mortality rate of the population based on the preoperative characteristics was 2.34, whereas the observed crude hospital mortality was 2.05% (8 patients). No significant difference was found between observed and expected hospital mortality rates ( p = 0.78). Cardiopulmonary bypass (CPB) was employed in 238 (62.8%) CABG procedures. The mean CPB time was 121 ± 39.2 min and the mean aortic crossclamp time was 42.8 ± 15.9 min. Mean hospital stay was 12.1 ± 7.1 days.
Arrhythmia was the most frequent postoperative complication (17.6%, excluding patients with preoperative arrhythmias). Patients with postoperative arrhythmias had a significantly longer hospital stay (11.5 vs. 14.6 days) and postoperative stay (10 vs. 13 days) than those with normal rhythm ( p = 0.001 for both). Approximately 65% of these patients had paroxysms of atrial fibrillation. Respiratory complications (2.1%) comprised 2 cases of respiratory failure requiring continuous positive airway pressure, 2 cases of respiratory failure requiring prolonged ventilation and tracheostomy, one case of pneumonia, and 3 of acute respiratory distress syndrome. Wound infection occurred in 1.8% of patients and mediastinitis in 0.5%. The postoperative prevalence of transient ischemic attack/stroke and renal failure/dialysis was 0.5% and 0.77%, respectively. No patient had bleeding requiring re-operation or any other CABG-associated re-operation.
A summary variable of serious hospital morbidity was created. It included patients who experienced at least one of the following complications: wound infection, mediastinitis, a cerebrovascular event (transient ischemic attack/stroke), renal failure/dialysis, respiratory complications, or hospital death. Twenty-three patients (5.9%) experienced a serious morbid event. Univariate logistic regression analyses revealed the risk of a serious morbid event was significantly associated with preoperative rhythm disturbances ( p = 0.029), drug allergy ( p = 0.021), diabetes mellitus ( p = 0.001), left main stem disease ( p = 0.021), urgent operation ( p = 0.035), poor (< 30%) ejection fraction ( p = 0.007), and EuroSCORE ( p < 0.001).1 In the group of patients with a serious morbid event, the mean EuroSCORE was nearly double that in patients without such events (4.23 vs. 2.22, p < 0.001).
To examine the independent effect of an individual variable while controlling for the others, multivariable logistic regression analyses were conducted using the forward stepwise selection method. The criterion to enter into the model was a p-value of 0.25 maximum, and the criterion to remain in the model was a p-value of not more than 0.05.2 The final model was checked for potential confounders and interactions (Table 2). Multivariable logistic regression analyses allowed depiction of the single effect of having diabetes while controlling for patient preoperative risk of mortality measured by EuroSCORE. Drug allergy, diabetes, and EuroSCORE were the independent predictors of developing a serious morbid event after an isolated CABG procedure. Diabetic patients had a significantly higher mean EuroSCORE than non-diabetics (p < 0.001).
Patients remained in the ICU until they were extubated, off all continuous drips, did not have any life threatening arrhythmia, and were mobile with one assistant. There is no step-down unit at the center, because all wards have telemetry monitoring and appropriate close care that acts as a step-down unit. Mean ICU length of stay was 61.50 ± 94.95 h, with a range of 24 to 1,316 h. The ICU stay was not taken as a normally distributed continuous variable because a few patients with long ICU stays would serve as influential outliers having a profound effect on the mean value. Intensive Care Unit stay in 275 patients (72.7%) was 
48 h. To eliminate this effect, a new dichotomous variable of prolonged ICU stay was created: ICU stay < 3 days (72 h) and ICU stay 3 days; 37 patients (9.5%) had prolonged ICU stay. To support the construction of such a dichotomous variable, the groups were examined with respect to their crude hospital mortality, mean EuroSCORE, and serious hospital morbid event rates (Table 3). The statistically significant differences supported the assertion that stays over 3 days were associated with increased adverse outcomes. To identify the predictors of prolonged ICU stay, univariate logistic regression analyses were performed. Statistically significant odds ratios were identified for age ( p = 0.003), female sex ( p = 0.023), rhythm disturbances prior to admission ( p = 0.001), and previous cardiac surgery ( p = 0.004).
In the next step, forward stepwise selection was used to conduct multivariable logistic regression analysis. A p value of 0.25 maximum was the criterion to enter into the model and at least 0.05 to remain in the model.2 The final multivariable logistic regression model was checked for potential confounders and interactions. Selected final models are presented in Table 4. An important factor in model selection is the inclusion of clinically relevant variables. This approach favored model 2 where the history of hypertension, although of borderline significance, was included in the model as an important predictor of prolonged ICU stay. The likelihood ratio test comparing the two models resulted in a significant p-value (0.036) supporting the selection of the more saturated model 2. Patient age and previous cardiac surgery are among the predictor variables of EuroSCORE. This is why the latter was not included in the final model to avoid collinearity. Based on model 2, age, preoperative rhythm disturbances, previous cardiac operation, and history of hypertension were the independent predictors of prolonged ICU stay.
|
DISCUSSION
|
|---|
Coronary artery bypass grafting is the most accepted method of myocardial revascularization.3 Increasing attention is being paid to outcomes after CABG procedures.4,5 Such results may be used as a tool for comparison between healthcare institutions. Mortality and morbidity following CABG are determined mainly by the preoperative status of the patient; the core variables include age, sex, left ventricular ejection fraction, diabetes, body size, and comorbidity.5,6 Coronary artery bypass grafting outcomes are also influenced by surgical and anesthetic techniques, CPB, efficacy of myocardial protection, patient hemodynamic management, and care in the ICU.7 The probability of a major morbid event is an important concern.6 Neurologic events are frequent; reported rates of stroke range from 0.4% to 13.8%.3 Some possible predictors of this complication include prior stroke, hypertension, advanced age, diabetes mellitus, and atherosclerosis of the aorta and carotid artery.3,8 Deep sternal wound infection is associated with substantial morbidity and mortality (prevalence, 0.7%–3.9%).3 Predictors include obesity, re-operation, use of internal mammary arteries, and diabetes.3,8,9 The predictors of postoperative renal dysfunction, which can occur in 3.5% of cases, include advanced age, a history of congestive heart failure, prior renal disease, and previous bypass operation.3,8,9 Postoperative angina, myocardial infarction, arrhythmia, bleeding, and re-operation are among the other CABG-associated complications.8,9 A patient’s prognosis on arrival in the ICU could be different from the preoperative prognosis.5 Intensive care unit stay and risk-adjusted ICU mortality are used to assess the quality of care in this department. Up to 37.1% of CABG patients might experience a prolonged ICU stay.10 Among the identified predictors of prolonged ICU stay are lung disease, rhythm disturbances, serious valve pathology, re-operation, non-elective surgery, and CPB.11
Morbidity rates for CABG procedures provide valuable insights into target areas for improved quality of care. Based on the findings of this study, postoperative arrhythmias were the most prevalent morbidity, associated with significantly longer postoperative and hospital stay, and thus higher treatment costs. Serious respiratory complications, renal failure, wound infection, mediastinitis, and cerebrovascular accident had low rates within the ranges reported by other major cardiac surgery centers.3,8,9 The predictors of serious morbid events were drug allergy, diabetes, and EuroSCORE. A history of allergy to various medications (analgesics, aspirin, antibiotics) was noted in 13.7% of the study population. Drug allergy was an unexpected finding as a predictor of serious morbidity. Nevertheless, the patients with reported drug allergy had significantly longer ICU stay and a higher rate of postoperative complications. One hypothesis is that a history of drug allergy could be a marker of changed immune response, which predisposes to development of post-surgical complications. Another hypothesis is that drug allergy could be a proxy for other unidentified comorbid conditions or risk factors. Diabetes was found to be a predictor of serious hospital complications following CABG in other observational studies.3,8,9 Additional factors such as blood gases, heart rate, and cardiac index at the time of ICU admission, use of an intra-aortic balloon pump during the operation, and duration of ventilation in the ICU could also play important roles.7 Identification of patients at risk of longer ICU stay is valuable for appropriate resource planning and patient follow-up within the hospital.
There were some limitations of this study. Although the definitions of variables were discussed with the medical staff, it is possible that they varied across the residents and surgeons, especially considering the retrospective design of the study. Another limitation was the low prevalence of some preoperative characteristics identified a priori as important. There were too few cases to develop reliable univariate and multivariable risk estimates for some factors. A larger sample size is needed for more accurate conclusions.
The results were comparable with those from other cardiac surgery centers. Older patients and those identified with significant predictors should be managed more cautiously both in terms of possible negative outcomes and treatment costs. The data on quality indicators such as hospital morbidity and prolonged ICU stay will be used as a baseline for further self-assessment activities and could become a benchmark for other cardiac surgery centers in Armenia and the region, should any be established. Further research with a larger sample size would allow the development of more robust models to calculate the risk of prolonged ICU stay and/or serious morbid events, and would enhance the planning of further medical care of the patients. The present study underlined the importance of conducting new interventional studies related to arrhythmias after CABG.
|
REFERENCES
|
|---|
- Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9–13.[Abstract/Free Full Text]
- Hosmer DW, Lemeshow S. Model-building strategies and methods for logistic regression. In: Applied logistic regression. 2nd ed. New York: John Wiley & Sons, 2000, 91–142.
- Gupta V, Grover V. Risk assessment and outcome after CABG. Ann Card Anaesth 2002;5:156–8.[Medline]
- Peterson ED, Coombs LP, Ferguson TB, Shroyer AL, DeLong ER, Grover FL, et al. Hospital variability in length of stay after coronary artery bypass surgery: results from the Society of Thoracic Surgeon’s National Cardiac Database. Ann Thorac Surg 2002;74:464–73.[Abstract/Free Full Text]
- Ferguson TB Jr, Hammill BG, Peterson ED, DeLong ER, Grover FL; STS National Database Committee. A decade of change—risk profiles and outcomes for isolated coronary artery bypass grafting procedures, 1990 1999: a report from the STS National Database Committee and the Duke Clinical Research Institute. Society of Thoracic Surgeons. Ann Thorac Surg 2002;73:480–90.[Abstract/Free Full Text]
- Tu JV, Sykora K, Naylor CD. Assessing the outcomes of coronary artery bypass graft surgery: how many risk factors are enough? Steering Committee of the Cardiac Care Network of Ontario. J Am Coll Cardiol 1997;30:1317–23.[Abstract]
- Higgins TL, Estafanous FG, Loop FD, Beck GJ, Lee JC, Starr NJ, et al. ICU admission score for predicting morbidity and mortality risk after coronary artery bypass grafting. Ann Thorac Surg 1997;64:1050–8.[Abstract/Free Full Text]
- Eagle KA, Guyton RA, Davidoff R, Ewy GA, Fonger J, Gardner TJ, et al. ACC/AHA guidelines for coronary artery bypass graft surgery: executive summary and recommendations: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to revise the 1991 guidelines for coronary artery bypass graft surgery). Circulation 1999;100:1464–80.[Free Full Text]
- Shroyer AL, Coombs LP, Peterson ED, Eiken MC, DeLong ER, Chen A, et al. The Society of Thoracic Surgeons: 30-day operative mortality and morbidity risk models. Ann Thorac Surg 2003;75:1856–65.[Abstract/Free Full Text]
- Bucerius J, Gummert JF, Walther T, Doll N, Falk V, Schmitt DV, et al. Predictors of prolonged ICU stay after on-pump versus off-pump coronary artery bypass grafting. Intensive Care Med 2004;30:88–95.[Medline]
- Janssen DP, Noyez L, Wouters C, Brouwer RM. Preoperative prediction of prolonged stay in the intensive care unit for coronary bypass surgery. Eur J Cardiothorac Surg 2004;25:203–7.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
