| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Surgical and Morphological Sciences, Cardiac Surgery Unit, Varese University Hospital, University of Insubria, Varese, Italy
b Department of Cardiovascular Sciences, Cardiac Surgery, and Intensive Care Unit, Centro Cardiologico Monzino, Istituto di Ricovero e Cura a Carattere Scientifico, University of Milan, Milan, Italy
c Department of Hypertension, Medical University of Lodz, Lodz, Poland
d Division of Cardiac Surgery, University of Verona Medical School, Verona, Italy
e Department of Surgical and Morphological Sciences, Vascular Surgery Unit, Varese University Hospital, University of Insubria, Varese, Italy
f Cardiac Intensive Care Unit, Cardiac Surgery Unit, Varese University Hospital, University of Insubria, Varese, Italy
Accepted for publication February 6, 2012.
* Address correspondence to Dr Mariscalco, Department of Surgical Sciences, Cardiac Surgery Unit, Varese University Hospital, University of Insubria, Via Guicciardini 9, 21100 Varese, Italy (Email: giovannimariscalco{at}yahoo.it).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods: The study enrolled 4,659 consecutive patients (21% women; age, 67.8 ± 9.2 years) undergoing coronary artery bypass grafting. A propensity score-based optimal-matching algorithm was used to match 1,577 patients receiving preoperative statins with a control group (1:1). Patients were screened for delirium in the intensive care unit according to the Confusion Assessment Method for the intensive care unit.
Results: Delirium affected 89 patients (3%), and preoperative statin administration was not multivariably associated with a decreased incidence of delirium (odds ratio, 1.52; 95% confidence interval, 0.97 to 2.37; p = 0.18) and was also unrelated to a delirium decrease in patient subgroups undergoing isolated coronary artery bypass grafting (odds ratio, 1.31; 95% confidence interval, 0.68 to 2.52; p = 0.51) or combined valvular procedures (odds ratio, 1.72; 95% confidence interval, 0.96 to 3.07, p = 0.08). Similar results were observed for age groups and cardiopulmonary bypass durations. Patients affected by postoperative delirium experienced a longer hospital stay (25th to 75th percentile) of 11 (7 to 18 days) vs 7 days (7 to 8 days, p < 0.001) and 12% hospital mortality vs 1% (p < 0.001).
Conclusions: Preoperative statins were not associated with a decreased incidence of delirium in patients undergoing coronary revascularization.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Although several studies have identified various risk factors of postoperative delirium, the pathophysiologic mechanisms remain unclear [1–9]. Plausible explanations include underlying cerebrovascular disease, advanced age, deleterious effects after cardiopulmonary bypass (CPB), embolic events, and neurotransmitter imbalance occurring during a prolonged ICU stay [1–9]. Experimental and clinical studies have recently documented improved perioperative central neural protection exerted by statins because of their antiinflammatory, immunomodulatory, and antithrombotic properties [11–13]. However, the results of favorable statin effects on delirium after cardiac operations are still uncertain [1, 7, 14, 15]. Therefore, the purpose of the present multicenter study was to investigate whether preoperative statin therapy is associated with a decrease in the incidence of postoperative delirium after coronary revascularization.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Study Population
Between October 2004 and December 2010, 4,659 consecutive patients undergoing coronary operations at 2 university institutions (Varese University Hospital and Centro Cardiologico Monzino, Istituto di Ricovero e Cura a Carattere Scientifico, Italy) were screened for this observational cohort study. Patients with a history of alcoholism or psychiatric illness requiring treatment were not enrolled. Patients undergoing coronary artery bypass grafting (CABG) without CPB and those who died intraoperatively were excluded from the analysis. Accordingly, the final sample comprised 4,079 coronary patients.
Data used in this analysis were retrieved from the institutional databases at both centers, which remained consistent during the study period [16, 17]. All data were prospectively collected, and demographics, comorbidities, medical and surgical history, operative details and postoperative events during the hospital stay were all registered.
Patient Management
Preoperative management, anesthetic, and surgical techniques were standardized for all patients and have been previously reported [16, 17]. Preoperative medications, including antihyperlipidemic therapy, were routinely omitted on the day of the operation and restarted after the operation (first postoperative day), unless clinically contraindicated. Statins were administered orally in extubated patients or by nasogastric tube in those ventilated.
Anesthetic management followed routine methods and remained consistent during the study period. Generally, patients received premedication with lorazepam (1 mg), morphine (0.1 mg/kg), and scopolamine (0.25 mg) 1 to 2 hours before the planned operation. Anesthesia was induced with midazolam (0.1 to 0.2 mg/kg), sufentanil (0.75 μg/kg), and rocuronium bromide (1 mg/kg). Anesthesia was thereafter maintained with sufentanil (0.01 μg/kg/min), propofol (5 mg/kg/h), rocuronium bromide boluses (0.15 mg/kg), and sevoflurane (2% to 3% if necessary). During the ICU stay, sedation was achieved with propofol infusion (1 to 3 mg/kg/h).
All surgical procedures were performed through a median sternotomy approach and CPB was of standardized fashion, with ascending aortic cannulation and right atrium or bicaval venous cannulation. Myocardial protection was afforded by potassium-derived intermittent cardioplegia (cold or warm, crystalloid or blood, antegrade or retrograde routes). Pericardial suction blood was recycled in all patients.
After the operation, patients were transferred to a dedicated cardiovascular ICU. Heart rate, electrocardiogram, central venous and arterial pressures, and acid-base and blood gases were continuously monitored during the ICU stay. Inotropic support was provided if ventricular contractility was considered frankly impaired to achieve stable hemodynamic conditions. Patients were generally extubated as soon as they met the following criteria: hemodynamic stability, no excessive bleeding (< 80 mL/h), normothermia, and consciousness with pain control. Perioperative need for blood products was determined on a patient-by-patient basis; in general, blood transfusions were administered when hemoglobin was less than 8 g/dL.
Definitions and Delirium Assessment
Delirium was defined as an acute deterioration of brain function characterized by fluctuating consciousness and an inability to maintain attention, with or without accompanying agitation characterized by excessive motor or verbal behavior that interferes with patient care, patient or staff safety, and medical therapy [18]. Delirium was generally assessed by the cardiovascular ICU nurses and anesthesiologists taking part in daily patient care, on the basis of the Confusion Assessment Method for the ICU [19, 20]. As previously reported [20], a four-step algorithm was adopted, assessing: (1) an acute onset of mental status changes or a fluctuating course; (2) inattention; (3) disorganized thinking; and (4) an altered level of consciousness (ie, other than alert). A patient who features 1 and 2 and either feature 3 or 4 was diagnosed as delirious. The evaluation was tailored to the individual patient to accommodate any reported or perceived auditory or visual deficits. Delirium evaluation was generally performed every 8 hours postoperatively, and patients were classified as having delirium if they experienced it on at least two positive assessments during the ICU stay. The nurses and anesthesiologists were unaware of the study goals.
Study Group
Patients who were receiving any preoperative statin therapy at admission and the day before the operation constituted the treatment group [16]. Commercially available statins screened for in this study were atorvastatin, pravastatin, simvastatin, lovastatin, rosuvastatin, and fluvastatin. Other considered variables were defined as previously described [16, 17]. Postoperative acute kidney injury (AKI) was defined by the consensus RIFLE criteria (Risk, Injury, Failure, Loss of function, and End-stage renal disease) using the maximal change in serum creatinine and estimated glomerular filtration rate (eGFR) during the first 7 postoperative days compared with preoperative baseline values [21]. Urinary output criteria in defining AKI were not considered. The eGFR was calculated with the Chronic Kidney Disease Epidemiology Collaboration equation [22].
Statistical Analysis
Clinical data were prospectively recorded and tabulated with Excel software (Microsoft Corp, Redmond, WA). Continuous variables were tested for normal distribution by the Kolmogorov-Smirnov test and compared between groups with unpaired Student t test for normally distributed values; otherwise, the Mann-Whitney U test was used. In the case of dichotomous variables, group differences were examined by Pearson 
2 or Fisher exact tests, as appropriate.
To control for selection bias related to the choice of preoperative statin therapy, a propensity score for the likelihood of receiving preoperative statin therapy was developed and a hierarchic logistic regression model to calculate propensity score was applied. The propensity score was based on 24 preoperative and intraoperative variables: age (years), sex, body surface area, emergency status, cardiogenic shock, reoperation, prior acute myocardial infarction, left main stem stenosis, hypertension, diabetes, dyslipidemia, chronic obstructive pulmonary disease, peripheral vascular disease, history of cerebrovascular accident, the additive European System for Cardiac Operative Risk Evaluation, eGFR, preoperative left ventricular ejection fraction (LVEF), type of operation (CABG isolated or combined with a valvular procedure), CPB time (min), aortic cross-clamp time (minutes), number of grafts, preoperative β-blocker agents, calcium channel blockers, and angiotensin-converting enzyme inhibitor.
This process generated a propensity score between 0 and 1, and patients receiving preoperative statin therapy were matched 1:1 with patients not receiving it, using the Rosenbaum optimal matching algorithm [23]. This approach minimizes the overall distance between observations and was conducted using the Mahalanobis distance within propensity score calipers (no matches outside the calipers) [23].
After the propensity score match was performed, we assessed differences between the two groups as above. Absolute standardized differences were estimated to evaluate the prematch imbalance and postmatch balance [23]. An absolute standardized difference of 0% indicates no residual bias and differences of less than 10% are considered inconsequential [24]. Then, univariable analyses for postoperative delirium occurrence were performed, and variables were subsequently tested in a multivariable model. A stepwise approach was used and confirmed by backward and forward methods. The models were built using variables that demonstrated a p value of less than 0.25 in univariable analysis.
The significance within the models was evaluated with the Wald test, whereas the strength of the association of variables with postoperative delirium was estimated by calculating the odds ratio (OR) and 95% confidence intervals (CIs). The model was calibrated by the Hosmer-Lemeshow goodness-of-fit test, as well as residual diagnostics (deviance and df of β); model discrimination was evaluated by using the area under the receiver operating characteristic curve. All tests were two-sided with the α level set at 0.05 for statistical significance. Extracted database variables were tabulated using Excel software and statistical analysis was computed using SPSS 19.0 (SPSS Inc, Chicago, IL) and NCCS 2007 7.1 software (NCCS, Kaysville, UT).
Sample Size Estimates
Data from previously published studies were used to calculate the minimum samples needed to detect significant differences for the outcome of interest (delirium occurrence). In a prior study of delirium after elective operations of more than 250,000 patients, the incidence of postoperative delirium among cardiac surgical patients who took preoperative statins was 6.7% (1,297 of 9,272) compared with 3% (7,975 of 9,272) among those who did not [7]. Applying these estimates and aiming for a study power of 80% and an α of 0.05, 416 study patients in each group would be needed. Given these estimates, the inclusion of more than 1,500 individuals in each group would be sufficient to assess a significant difference in delirium incidence even if the treatment effect was smaller than previously reported.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
After propensity score-matching analysis, 1,577 patients (98%) who received preoperative statin therapy were matched with a control group. Postmatch standardized differences for all measured covariates were less than 10% (most were < 5%), suggesting substantial covariate balance across groups (Fig 1). Both groups also appeared completely comparable in preoperative and perioperative characteristics and risk, with no statistically significant average risk scores as calculated by the European System for Cardiac Operative Risk Evaluation (4.9 ± 3.3 vs 4.9 ± 3.0, p = 0.48; Table 1). In the propensity-matched population (n = 3,154), the overall incidence of delirium was 3% (n = 89), and among patients who experienced delirium, the occurrence was most frequently on the same day of operation (53%), followed by postoperative day 1 (25%). Of the 1,577 patients with preoperative statin therapy, postoperative delirium occurred in 53 (3.4%) compared with 36 (2.3%) of those without it. A higher delirium rate was observed patients aged 65 years or older (4% vs 2%, p < 0.001). Patients affected by delirium had a higher rate of emergency and combined procedures (Table 2) and presented a more severe profile of comorbidities with a prolonged CPB and aortic cross-clamp time.
|
|
|
|
2(8 df) = 6.01, p = 0.65) and area under the receiver operating characteristic curve analysis (0.83) revealed good calibration and discrimination for the multivariable model analysis.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
However, Katznelson and colleagues [1], retrospectively enrolling 1,059 cardiac surgical patients undergoing CPB, observed that administration of statins conferred a 46% reduction in risk of delirium development (OR, 0.54; 95% CI, 0.35 to 0.84). The discrepancy in statin results on delirium between that study and ours could be explained by the different data collection protocol [1]. Those investigators did not control for preoperative variables, and a selection bias may have been present because there was no homogeneous indication for preoperative statin therapy. Their analysis did not account for this by assessing the role of statins while stratifying on the propensity score, and the statin users and nonusers were not comparable in preoperative characteristics.
Our data, however, are consistent with the Mathew and colleagues study [14], which demonstrated no effects of preoperative statins on cognitive function after CABG. Redelmeier and colleagues [7], conducting a retrospective cohort analysis of 284,158 consecutive patients admitted for elective operations (9,272 underwent cardiac operations), confirmed that the use of statins was not associated with a reduction in the risk of postoperative delirium among elderly patients (OR, 1.30; 95% CI, 0.96 to 1.70). Similarly, Koenig and colleagues [15] did not observe an association between statins and a decreased incidence of stroke and encephalopathy in 5,121 patients undergoing isolated CABG during a 10-year period.
The lack of statin protection could be certainly related to the complex multifactorial pathogenesis of postoperative delirium [9]. The etiology of delirium involves an interaction of multiple factors, and it remains distinctly possible that the pleiotropic effects of statin therapy are ruined by the complex events that occur during cardiac operations [3, 7–9, 14, 15, 25]. Changes in catecholamine levels after CPB use, cerebral embolization detected during the surgical procedure, decreased cardiac output observed after the operation, and prolonged ICU stay seem to overcome the potential protective effects of preoperative statins on delirium [3, 7–9, 14, 15, 25]. Mathew and colleagues [14] documented that cytokine and C-reactive protein levels did not differ in patients with or without postoperative cognitive dysfunction independently of the preoperative statin therapy. These authors concluded that the relevant proinflammatory effects of CPB overwhelmed the antiinflammatory effects of statins [14].
The acute withdrawal of statins, especially occurring in the first hours after the operation, may also reduce the protective effect of these drugs [14]. Experimental data in a mouse model of cerebral ischemia demonstrated that acute termination of statin treatment resulted in a rapid loss of cerebral protection [27]. In addition, other detrimental factors, as we documented in our series, could interfere with cerebral function despite the protective effects of statins.
On one hand, age, hypertension, and history of cerebrovascular accident have been repeatedly observed to be related to increased cerebral atherosclerosis, which may inhibit cerebral blood flow, exacerbate the nonpulsatility of CPB, and increase the risk of cerebral embolization [1, 3, 7, 9, 25, 26]. Advanced age is also associated with lack of cholinergic reserve, one of the main neurochemical changes that predispose patients to delirium [28]. On the other hand, preexisting LV failure or decreased cardiac output resulting from postoperative complications, such as atrial fibrillation or AKI, may lead to suboptimal cerebral perfusion [1, 3, 8, 9, 10, 15, 29]. In addition, combined surgical procedures require prolonged CPB with its deleterious effects and heart manipulations with attendant embolic phenomena [25].
An interesting observation of our study was the relatively low incidence of observed postoperative delirium compared with other series [1, 3, 5–9, 15, 26]. However, its incidence is different across studies, occurring in 1% to 52% of the patients, because of the adopted study designs, the method of assessing delirium, and the enrolled populations [1–4, 6–9]. Another confusing aspect is the frequent mistaken diagnosis of delirium vs postoperative cognitive dysfunction, as reported in several studies [9,30]. The diagnosis of delirium is based on the detection of impairment of the consciousness and attention, of abstract thinking and comprehension, as well as disturbance of cognition with hallucinations with or without accompanying agitation [18–20]. In addition, a typical feature of postoperative delirium is that the patient immediately recalls with an impaired recent memory but an intact remote memory [30]. Postoperative cognitive dysfunction, however, often takes several days or weeks to be recognized, with a minority of patients impaired for life [30]. Postoperative cognitive dysfunction is not associated with a change in consciousness, and its detection requires sensitive testing methods [30].
In the present study, delirium detection was generally based on the Diagnostic and Statistical Manual of Mental Disorders-IV criteria, and Confusion Assessment Method for the ICU was the assessment tool [18–20]. Contrary to other experiences, we diagnosed patients as having delirium if they experienced it on at least two positive Confusion Assessment Method for the ICU assessments. However, our data are consistent with those noted in other reports [2, 4, 7]. Redelmeier and colleagues [7] observed a rate of 1% delirium occurrence in a large study cohort undergoing elective operations, and the definition of delirium was based on the International Classification of Diseases codes (293.0 to 293.9). Norkiene and colleagues [2], in a case series drawn from 1,367 consecutive patients undergoing CABG, documented a delirium rate of a 3%. Delirium was defined according to Diagnostic and Statistical Manual of Mental Disorders-IV criteria, and the diagnosis was made by an intensive care clinician, taking part in daily patient care [2].
A second intriguing finding of the present study was the significantly greater rate of delirium in subjects with poor preoperative LVEF (<0.30) taking statins. Although the role of statins in patients with heart failure has still not been established and controversial results have been presented [31], we observed a similar association in other high-risk patient populations, especially for patients undergoing prolonged and complex surgical procedures (Table 3). However, our data do not necessarily imply a harmful effect exerted by statins in this patient setting. Complexity clearly exists in distinguishing a possible direct effect of statins from the severe comorbidities of patients in whom statins are prescribed. A plausible explanation refers to the severe profile of comorbidities of statin users, including peripheral and cerebrovascular disease. The protective cerebral effect of statin therapy could also be overcome by the inherent cerebral blood alterations exacerbated during CPB and by other CPB-related perioperative events.
Finally, our data unequivocally confirm that postoperative delirium is related to increased early postoperative morbidity and a prolonged ICU and hospital stay, with significant adverse effects. Delirium was associated here with a 2-day to 3-day increase in ICU and hospital stay and a 10-fold increase in hospital death, underlying the need to identify patients with a high risk of developing delirium, ensuring that patients can be treated efficiently and the postoperative morbidity and death can thereby be reduced.
Our study has several limitations. It is a nonrandomized study, being a multicenter retrospective investigation based on prospectively collected data. A selection bias may have been present, although we attempted to adjust for these differences by assessing the role of preoperative statins while stratifying on the propensity score. Despite this careful model approach, measurable and immeasurable factors may still exist, such as the perioperative events related to CPB use [9]. CPB provides a brain perfusion that is very different from normal physiology, and embolization is known to occur with microemboli as well as macroemboli. CPB is also associated with a pronounced systemic inflammatory response that involves the release of mediators with important effects on the central nervous system. CPB may also lead to a higher prevalence of postoperative delirium because of hypotensive periods during CPB and nonpulsatile blood flow. All these events have been demonstrated as being associated with delirium development [9]. Moreover, physician bias may have influenced patient selection, statin type, and dose, a difficulty inevitably shared with other studies analyzing the same subject [1, 7, 14–16].
Second, we were unable to determine the influence of the duration of preoperative statin therapy on the risk of postoperative delirium. In the statin group, we were able to document the duration of statin administration from the starting day of hospitalization only. We were not able to perform a detailed analysis of the association between the different statin medications and delirium occurrence because of the limited sample size of each subgroup; nor could we measure or control for the effect of postoperative statin administration. Similarly, we were unable to clarify how long the possible protective effects of statin therapy lasted after intake the day before the operation.
In a murine model of transient focal cerebral ischemia, chronic statin pretreatment conferred a 40% reduction in infarct size, but protection was lost when statin therapy was stopped 2 days before the induction of cerebral ischemia [32]. In a prospective study of 669 patients undergoing major vascular operations, statin withdrawal of more than 4 days was an independent predictor of myonecrosis (OR, 2.9; 95% CI, 1.6 to 5.5), whereas early resumption of statins (
4 days) was associated with a cardiac protective effect [33]. Therefore, our policy of the cessation of statin therapy the day of the operation and its early resumption (postoperative day 1) could not have influenced the possible protective cerebral effects.
A further limitation is the lack of data concerning the clinical contraindications for not restarting statin therapy in the postoperative period, which could have possibly influenced the rate of delirium occurrence. In addition, the lack of delirium assessments after ICU discharge could have led to underestimation of its occurrence. Finally, we did not perform a baseline assessment of inflammatory markers, although inflammation is a possible mechanism associated with delirium development [1, 7, 15].
In conclusion, we did not find an association between preoperative statin treatment and postoperative delirium in a large cohort of patients undergoing isolated or combined surgical revascularization.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  S. Mitchell Invited Commentary Ann. Thorac. Surg., May 1, 2012; 93(5): 1447 - 1448. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
