JTCS Speed Up Your Browser
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Vivek Rao
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wijeysundera, D. N.
Right arrow Articles by Karkouti, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wijeysundera, D. N.
Right arrow Articles by Karkouti, K.
Related Collections
Right arrow Coronary disease
Right arrowRelated Articles

J Thorac Cardiovasc Surg 2004;127:755-762
© 2004 The American Association for Thoracic Surgery

Cardiopulmonary support and physiology

Calcium antagonists are associated with reduced mortality after cardiac surgery: a propensity analysis

Duminda N. Wijeysundera, MDa, W. Scott Beattie, MD, PhDb,*, Vivek Rao, MD, PhDc, Joan Ivanov, RN, MSc, PhDc, Keyvan Karkouti, MD, MScd

a Department of Anesthesia, University of Toronto, Toronto, Ontario, Canada
b Department of Anesthesia, University Health Network, University of Toronto, Toronto, Ontario, Canada
c Division of Cardiac Surgery, University Health Network, University of Toronto, Toronto, Ontario, Canada
d Department of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada

Received for publication October 1, 2002; accepted for publication April 11, 2003.

* Address for reprints: W. Scott Beattie, MD, PhD, Department of Anaesthesia, University of Toronto, EN 3-453, Toronto General Hospital, 200 Elizabeth St, Toronto, Ontario M5G 2C4, Canada
scott.beattie{at}uhn.on.ca


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
OBJECTIVES: Observational studies have questioned the effectiveness of perioperative calcium channel blockers but failed to correct for selection biases. We therefore performed a prospective observational cohort study of the effects of calcium channel blockers on cardiac surgical mortality. A propensity score technique was used for risk adjustment.

METHODS: We identified 6619 patients who underwent nontransplant cardiac surgery at the Toronto General Hospital (Toronto, Ontario, Canada) between May 1999 and December 2001. Propensity scores for calcium channel blocker use were constructed for the entire sample and for the subgroup (n = 5222) that underwent coronary artery bypass grafting. The calcium channel blocker adjusted odds ratio for in-hospital mortality after cardiac surgery was determined by using multiple logistic regression and propensity matched-pairs analyses. A subgroup analysis was performed for patients who underwent coronary artery bypass grafting: the calcium channel blocker adjusted odds ratio for mortality was determined by using propensity score matched-pairs analyses.

RESULTS: Calcium channel blockers were associated with significantly reduced cardiac surgical mortality after adjustment with both multiple logistic regression (odds ratio, 0.56; 95% confidence interval, 0.33-0.94; P = .028) and propensity score matched-pairs analyses (odds ratio, 0.56; 95% confidence interval, 0.32-0.98; P = .042). Calcium channel blockers were also associated with reduced mortality (odds ratio, 0.48; 95% confidence interval, 0.23-0.98; P = .044) among patients who underwent coronary artery bypass grafting.

CONCLUSIONS: After adjustment for baseline differences, calcium channel blockers were associated with significantly reduced mortality after cardiac surgery. This benefit also extends to the subgroup that underwent coronary artery bypass grafting. A large randomized controlled trial of perioperative calcium channel blockers is therefore warranted.


See related editorials on pages 622 and 625.

 

The role of calcium channel blockers (CCBs) in cardiac surgery is controversial. Prior observational studies concluded that CCBs did not reduce mortality1 or myocardial ischemia.2,3 In contrast, a recent meta-analysis of randomized controlled trials (RCTs) found that CCBs significantly reduced perioperative myocardial infarction (MI) and ischemia during cardiac surgery;4 however, the study was underpowered to determine effects on mortality.

CCBs have multiple benefits that explain the findings of the meta-analysis. They improve balance between myocardial oxygen supply and demand through negative chronotropic, negative inotropic, afterload-reducing, and coronary vasodilatatory properties. Imbalance between myocardial oxygen supply and demand might cause ischemia and MI.5 CCB-mediated vasodilatation might reduce coronary artery bypass graft (CABG) spasm, another cause of postoperative ischemia. Nondihydropyridine CCBs have also been shown to prevent supraventricular tachyarrhythmias in both animal studies6 and the recent meta-analysis.4 They might therefore reduce postoperative atrial fibrillation, which is associated with neurocognitive dysfunction7 and prolonged hospitalization.8

The cause of the discrepancies between RCTs and observational studies might be inadequate risk adjustment for important selection biases. Patients receiving CCBs are likely to be older, with more chronic obstructive pulmonary disease, peripheral vascular disease, and diabetes mellitus. These systemic diseases increase cardiac surgical mortality.9 Furthermore, they reduce the likelihood of receiving ß-blockers,10 which reduce mortality after CABG surgery.11 Indeed, only 30% of patients who receive CCBs after MI will receive ß-blockers concurrently.10 Furthermore, CCB users might have more severe coronary artery disease that failed symptomatic control with ß-blockers alone. Failure to adjust for these selection biases will prevent identification of the treatment effects of CCBs.

We therefore undertook a prospective observational cohort study of cardiac surgical patients, aiming to determine the effects of CCBs on in-hospital mortality while adjusting for selection biases with propensity score techniques.12


   Methods
Top
 Abstract
 Methods
Results
 Discussion
 References
 
Patient population and data sources
Preoperative, intraoperative, and postoperative information for individuals undergoing cardiac surgery at the Toronto General Hospital, University Health Network (Toronto, Ontario, Canada), was prospectively collected in cardiac surgery and anesthesia registries, with the approval of the institutional research ethics board. These databases have been previously described.13,14 The databases document CCB and ß-blocker use in the preoperative period but not in the intraoperative or postoperative periods. All cardiac surgical patients at the Toronto General Hospital routinely receive their last preoperative dose of cardiac medication on the morning of the operation.

The study sample consisted of individuals who underwent nontransplant cardiac surgical procedures between May 1, 1999, and December 31, 2001. A total of 6635 individuals underwent eligible procedures. We excluded 16 (0.2%) patients with missing preoperative medication data. Ninety-two (1.4%) patients had missing data in elements aside from preoperative medications. An unknown left ventricular ejection fraction was considered equal to a normal ejection fraction (>60%), a previously used approach15 recommended by Pierpont and associates.16 Missing values for dichotomous variables were assigned the most frequent value, whereas continuous variables were assigned the median value, as previously described.11 Reanalysis after exclusion of patients with missing data did not materially alter the results.

General analysis issues
All analyses were performed with SAS Version 8.20 software (SAS Institute, Cary, NC).

Propensity score development
Given that CCB therapy assignment was nonrandom, confounding and selection biases were accounted for through a propensity score for CCB use. The rationale and methods underlying the use of propensity scores for proposed causal exposure variables have been previously described.17 Multiple logistic regression18 was used to construct a propensity score for preoperative CCB use, without regard for outcome. We developed a full nonparsimonious model, including 26 covariates and 18 first-order interactions (Table 1). This model had an area under the receiver operating characteristic (ROC) curve of 0.73.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Propensity score model for entire study sample

 
Multiple logistic regression risk adjustment
The effects of CCBs and ß-blockers on in-hospital mortality were determined by using multiple logistic regression. Treatment effects were adjusted by using the following covariates: year, age, sex, left ventricular ejection fraction (>60%, 40%-60%, 20%-40%, and <20%), procedure (CABG, single-valve surgery, and complex procedures), prior cardiac surgery, timing of operation, triple-vessel coronary disease (stenoses >70% in 3 major coronary arteries), left main coronary artery disease (stenosis >50%), MI within 30 days, Canadian Cardiovascular Society class IV angina, preoperative intra-aortic balloon pump support, diabetes mellitus, peripheral vascular disease, cerebrovascular disease, and chronic obstructive pulmonary disease. Operation timing was graded as follows: elective, semiurgent (cannot leave hospital without surgical intervention), urgent (surgical intervention required within 72 hours of presentation), and emergency (surgical intervention required within 12 hours of presentation).

Three models were generated. The first model consisted of the risk-adjustment covariates alone. The second model consisted of the risk-adjustment covariates, CCB use, and ß-blocker use. The third model consisted of the risk-adjustment covariates, CCB use, ß-blocker use, and propensity score. We calculated adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for CCB and ß-blocker use in the 2 latter regression models.

Combined CCB and ß-blocker use might represent a harmful interaction given their combined negative inotropic effects. We therefore tested the statistical significance of adding an interaction term between CCBs and ß-blockers in the third regression model.

The models’ discrimination and calibration were determined by using ROC curve areas and Hosmer-Lemeshow goodness-of-fit statistics, respectively.18

Propensity score Matched-Pairs analysis
To determine the robustness of the results from model-based risk-adjustment techniques, we also calculated the treatment effects of CCBs by using a matched-pairs analysis. Using an SAS macro (available at www2.sas.com/proceedings/sugi26/p214-26.pdf), we matched CCB users to unique control subjects by using their propensity scores.* We attempted to match each CCB user to a non-CCB user with a propensity score identical to 5 digits. If this was not possible, we attempted 4-, 3-, 2-, and 1-digit matches. Once this threshold was passed, that specific CCB user was excluded.

Conditional logistic regression was used to determine the effects of CCBs on mortality and several postoperative complications: dialysis, low cardiac output state, permanent pacemaker insertion, reoperation, and major bleeding. Low cardiac output state was defined as the need for intra-aortic balloon pump support or inotropic medication (dopamine in excess of 4 µg · kg–1 · min–1, dobutamine, milrinone, or epinephrine) for at least 30 minutes to maintain systolic blood pressure of greater than 90 mm Hg and cardiac output of greater than 2.2 L · min–1 · m–2.13 Major bleeding was defined as a requirement for more than 10 units of blood during the operative period, chest tube drainage of greater than 2400 mL over 24 hours, or reoperation for significant bleeding.19

Subgroup analyses
CCBs might have different effects among individuals with significant coronary artery disease. We therefore performed a subgroup analysis for patients who underwent CABG with and without any concurrent procedures. This subgroup consisted of 5222 (78.9%) patients.

A propensity score for CCB use within this subgroup was developed by using multiple logistic regression. We developed a full nonparsimonious model, including 24 covariates and 16 first-order interactions; these variables are available from the authors. The ROC area for this model was 0.68. Matched pairs were generated on the basis of propensity scores, as described above. Conditional logistic regression modeling was used to determine the effects of CCBs on mortality, dialysis, low cardiac output state, permanent pacemaker insertion, reoperation, and major bleeding.

Sensitivity analyses
Given that ß-blockers are associated with reduced mortality after CABG,11 we carried out sensitivity analyses to determine whether preoperative ß-blocker use affected the treatment effects of CCBs. Patients who underwent CABG were further divided into 2 subgroups defined by whether they were or were not receiving ß-blockers preoperatively. We used multiple logistic regression to determine the adjusted effects of CCBs on mortality within each subgroup. Although age, sex, and CCB use were forced into the models, forward stepwise selection was used to identify other covariates. We used stepwise selection because of the limited statistical power of the subgroups. The covariates considered were year, left ventricular ejection fraction, procedure other than CABG, prior cardiac surgery, timing of operation, triple-vessel coronary disease, left main coronary artery disease, MI within 30 days, Canadian Cardiovascular Society class IV angina, preoperative intra-aortic balloon pump support, diabetes mellitus, peripheral vascular disease, cerebrovascular disease, and chronic obstructive pulmonary disease. A P value of less than .05 was required for inclusion.


   Results
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Patient characteristics
The study sample consisted of 6619 patients; 31.7% (n = 2097) were receiving CCBs preoperatively. The prevalence of CCB use decreased from 35% in 1999 to 29% in 2001. The in-hospital mortality rate was 1.8% (n = 117). The patients’ baseline characteristics are summarized in Table 2. Renal dysfunction was defined as a preoperative creatinine concentration of greater than 130 µmol/L, which is associated with increased perioperative mortality and morbidity.20


View this table:
[in this window]
[in a new window]
  		TABLE 2. Baseline characteristics of study sample

 
Multiple logistic regression risk adjustment
The model consisting of the risk-adjustment covariates alone had an ROC area of 0.85, with good calibration ({chi}2 = 6.22, P = .62). CCBs were associated with significant reductions in mortality both before (OR, 0.55; 95% CI, 0.33-0.92; P = .022) and after (OR, 0.56; 95% CI, 0.33-0.94; P = .028) adjustment for propensity score (Table 3). ß-Blockers were also associated with significant reductions in mortality in both models (Table 3). The interaction term between CCB use and ß-blocker use was statistically nonsignificant (P = .59).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Adjusted effects of CCBs and ß-blockers

 
The risk-adjusted model without a propensity score had an ROC area of 0.86, with good calibration ({chi}2 = 5.75, P = .68). The model with a propensity score had similar discrimination (ROC area of 0.86) and calibration ({chi}2 = 7.16, P = .52).

Propensity score matched-pairs analysis
Of the 2097 CCB users in the entire study sample, 90% (n = 1890) were matched to unique control subjects. These matched pairs were well balanced for all known covariates (Table 2). The in-hospital mortality rate was 1.4% (n = 53).

Within the matched-pairs sample, CCBs were associated with significant reductions in mortality (OR, 0.56; 95% CI, 0.32-0.98; P = .042; Table 4). CCBs were not associated with statistically significant increases in postoperative complications (Table 4).


View this table:
[in this window]
[in a new window]
  		TABLE 4. Outcomes for propensity score matched pairs

 
Subgroup analyses
The baseline characteristics of the 5222 patients undergoing CABG are summarized in Table 5; 37.1% (n = 1935) were receiving CCBs preoperatively. The in-hospital mortality rate was 1.5% (n = 79).


View this table:
[in this window]
[in a new window]
  		TABLE 5. Subgroup who underwent CABG

 
Of the 1935 CCB users within the CABG subgroup, 89% (n = 1726) were matched to unique control subjects. These matched pairs were well balanced for all known covariates (Table 5). The in-hospital mortality rate for matched pairs was 1.1% (n = 38).

Among the matched pairs, CCBs were associated with significant reductions in mortality (OR, 0.48; 95% CI, 0.23-0.98; P = .044; Table 6). CCBs were not associated with statistically significant increases in postoperative complications (Table 6).


View this table:
[in this window]
[in a new window]
  		TABLE 6. Outcomes for propensity matched patients undergoing CABG

 
Sensitivity analyses
The CABG subgroup that received ß-blockers preoperatively consisted of 3900 patients, of which 35% (n = 1362) were receiving CCBs. The effect of CCBs on mortality among patients receiving ß-blockers (Table 7) was similar in magnitude to that determined by matched-pairs analyses (Table 6).


View this table:
[in this window]
[in a new window]
  		TABLE 7. Sensitivity analyses examining whether preoperative ß-blockers influence the treatment effects of CCBs on CABG mortality

 
The CABG subgroup that was not receiving ß-blockers preoperatively numbered 1322, of which 43% (n = 573) were receiving CCBs. The effect of CCBs on mortality within this subgroup (Table 6) was also similar to that determined by matched-pairs analyses (Table 7).


   Discussion
Top
 Abstract
 Methods
 Results
 Discussion
References
 
To our knowledge, this is the largest observational study to examine the effects of CCBs on perioperative mortality after cardiac surgery. Approximately 32% of patients were receiving CCBs preoperatively, with use gradually declining from 1999 to 2001. Multiple selection biases determined CCB use, including differences in age, sex, cardiac disease, systemic disease, and preoperative medications. After adjusting for selection biases, CCBs were associated with significantly reduced in-hospital mortality among patients undergoing cardiac surgery, including the subgroup undergoing CABG.

In addition to demonstrating that CCBs were associated with reduced mortality, this study also found that CCBs were free of major side effects. Although CCBs are negative chronotropes, negative inotropes, and platelet inhibitors,21 we found no evidence that they increased the incidence of related postoperative complications. Furthermore, the benefits of CCBs appeared to be consistent, regardless of whether patients were or were not receiving ß-blockers preoperatively.

Comparison with prior studies
Our results contradict prior observational studies and support a recent meta-analysis.4 Two early studies2,3 looked only at the effects of CCBs on myocardial ischemia, a surrogate outcome, and did not use risk-adjustment techniques. More recently, Weightman and colleagues1 found that CCBs had no effect on mortality, whereas ß-blockers exerted a significant beneficial effect. However, their multiple logistic regression analysis had less statistical power than ours. The power of logistic regression models depends in large part on the number of outcome events.22 There were only 53 deaths in the study by Weightman and colleagues. In contrast, we used more rigorous statistical risk-adjustment techniques in a sample with 119 deaths.

The recent meta-analysis of preoperative, intraoperative, and postoperative (first 48 hours) CCB use was underpowered to demonstrate significant changes in perioperative mortality.4 However, CCBs were associated with trends toward decreased mortality among patients undergoing CABG (OR, 0.66; 95% CI, 0.26-1.70; P = .4). The similarity in treatment effects observed in the meta-analysis and this observational study strengthens arguments that CCBs have important mortality benefits during cardiac surgery.

Our results argue against concerns that CCBs increase perioperative mortality, as suggested by Legault and coworkers.23 Their RCT of nimodipine use during valve surgery was terminated early because of an excess of deaths in the CCB arm, largely caused by excessive bleeding. However, these patients all underwent hypothermic cardiopulmonary bypass. The latter inhibits platelet function,24 whereas its effects on perioperative blood loss are mixed.24,25 Furthermore, only 46% received aminocaproic acid, an antifibrinolytic agent that reduces blood loss during cardiac surgery by 30% to 40%.26 By comparison, patients undergoing cardiac surgery at our hospital do not undergo hypothermic cardiopulmonary bypass and routinely receive antifibrinolytic agents. In contrast to the study by Legault and coworkers,23 our analyses found no association between CCB use and major bleeding, thereby supporting a recent review that concluded that most of the clinical data linking CCBs and bleeding points against an increased risk.27

Study limitations
Although this study used a variety of statistical methods to correct for confounders and selection biases, it is an observational study. Hence interpretation of results from this study should be limited to associations between variables of interest. No causal inferences should be drawn. Furthermore, cardiac surgical outcomes are confounded by a multitude of factors. The constraints of feasibility limit data collection in observational studies of this design; therefore all factors that determined preoperative CCB use might not have been collected. Although propensity scores reduce bias caused by observed covariates, they cannot remove confounding from these unobserved covariates. Only an RCT is able to do the latter.

Our analyses did not have sufficient power to determine whether CCBs benefit the subgroup receiving preoperative ß-blockers. Nonetheless, our results suggest that the 2 medications have additive benefits, without any negative interactions. The logistic regression interaction term between CCBs and ß-blockers was statistically nonsignificant. In addition, sensitivity analyses showed that preoperative ß-blocker use did not affect the magnitude of benefits of CCBs among patients undergoing CABG.

Our data set did not allow us to determine the effects of different classes of CCBs: benzothiazepines (eg, diltiazem), phenylalkyamines (eg, verapamil), and dihydropyridines (eg, nifedipine and amlodipine). This is important given that dihydropyridines have been associated with increased mortality in nonsurgical studies.28 In addition, we were not able to determine whether CCB therapy was continued throughout the immediate postoperative period. More than 50% of patients previously receiving CCBs might not receive them postoperatively.29 The findings of the above meta-analysis,4 however, suggest that ongoing CCB use intraoperatively and postoperatively would likely improve perioperative outcomes as well.

Finally, this study might have been affected by site confounding, given that our results reflect the effects of CCBs in a single tertiary care institution. Furthermore, the 1.8% perioperative mortality rate seen in this study contrasts with the 3.1% rate reported in a recent North American population-based study of CABG surgery.11 The generalizability of our results to other hospitals will therefore require further study. It is nonetheless reassuring that CCBs have benefits even when baseline operative risks are relatively low.

Clinical implications
CCBs remain relatively underused in cardiac surgical practice, largely because of the results of prior observational studies.1-3 This observational study, using a larger sample size and more sophisticated risk-adjustment techniques, suggests instead that preoperative CCBs are associated with significant reductions in perioperative mortality. In light of a meta-analysis showing that intraoperative and postoperative CCB use significantly reduces MI and ischemia,4 there is now strong justification for prospectively investigating perioperative CCB use among cardiac surgical patients. The most appropriate method for determining their efficacy is a large simple RCT, a feasible option given that 750,000 individuals annually undergo cardiac surgical procedures in the United States alone.30 We would further suggest that this trial should target individuals undergoing CABG, the subgroup most likely to benefit from perioperative CCBs.


   Acknowledgments
 
We acknowledge the superb data collection and management of Susan Collins at the University Health Network.


   Footnotes
 
Dr Wijeysundera is the recipient of the Allan K. Laws Clinician Scientist Fellowship from the University of Toronto.

* Parsons LS. Reducing bias in a propensity score matched-pair sample by using greedy matching techniques. Presented at the 26th Meeting of SAS Users’ Group International (SUGI 26); April 22-25, 2001; Long Beach, Calif. Back


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. 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;88:286–291[Abstract/Free Full Text]
  2. Chung F, Houston PL, Cheng DC, Lavelle PA, McDonald N, Burns RJ, et al. Calcium channel blockade does not offer adequate protection from perioperative myocardial ischemia. Anesthesiology. 1988;69:343–347[Medline]
  3. Slogoff S, Keats AS. Does chronic treatment with calcium entry blocking drugs reduce perioperative myocardial ischemia? Anesthesiology. 1988;68:676–680[Medline]
  4. Wijeysundera DN, Beattie WS, Rao V, Karski J. Calcium antagonists reduce cardiovascular complications after cardiac surgery: a meta-analysis. J Am Coll Cardiol. 2003;41:1496–1505[Abstract/Free Full Text]
  5. Slogoff S, Keats AS. Does perioperative myocardial ischemia lead to postoperative myocardial infarction? Anesthesiology. 1985;62:107–114[Medline]
  6. Talajic M, Papadatos D, Villemaire C, Nayebpour M, Nattel S. Antiarrhythmic actions of diltiazem during experimental atrioventricular reentrant tachycardias. Importance of use-dependent calcium channel-blocking properties. Circulation. 1990;81:334–342[Abstract/Free Full Text]
  7. Stanley TO, Mackensen GB, Grocott HP, White WD, Blumenthal JA, Laskowitz DT, et al. The impact of postoperative atrial fibrillation on neurocognitive outcome after coronary artery bypass graft surgery. Anesth Analg. 2002;94:290–295[Abstract/Free Full Text]
  8. Mathew JP, Parks R, Savino JS, Friedman AS, Koch C, Mangano DT, et al. Atrial fibrillation following coronary artery bypass graft surgery: predictors, outcomes, and resource utilization. JAMA. 1996;276:300–306[Abstract/Free Full Text]
  9. Jones RH, Hannan EL, Hammermeister KE, Reid JL. Identification of preoperative variables needed for risk adjustment of short-term mortality after coronary artery bypass graft surgery. J Am Coll Cardiol. 1996;28:1478–1487[Abstract]
  10. Gottlieb SS, McCarter RJ, Vogel RA. Effect of beta-blockade on mortality among high-risk and low-risk patients after myocardial infarction. N Engl J Med. 1998;339:489–497[Abstract/Free Full Text]
  11. Ferguson TB Jr, Coombs LP, Peterson ED. Preoperative beta-blocker use and mortality and morbidity following CABG surgery in North America. JAMA. 2002;287:2221–2227[Abstract/Free Full Text]
  12. Rosenbaum PR, Rubin DB. Reducing bias in observational studies using subclassification on the propensity score. J Am Stat Assoc. 1984;79:516–524
  13. 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;112:38–51[Abstract/Free Full Text]
  14. Karkouti K, Cohen MM, McCluskey SA, Sher GD. A multivariable model for predicting the need for blood transfusion in patients undergoing first-time elective coronary bypass graft surgery. Transfusion. 2001;41:1193–1203[Medline]
  15. Tu JV, Jaglal SB, Naylor CD. Multicenter validation of a risk index for mortality, intensive care unit stay, and overall hospital length of stay after cardiac surgery. Circulation. 1995;91:677–684[Abstract/Free Full Text]
  16. Pierpont GL, Kruse M, Ewald S, Weir EK. Practical problems in assessing risk for coronary artery bypass grafting. J Thorac Cardiovasc Surg. 1985;89:673–682[Abstract]
  17. Joffe MM, Rosenbaum PR. Invited commentary: propensity scores. Am J Epidemiol. 1999;150:327–333[Abstract/Free Full Text]
  18. Hosmer DW, Lemeshow S. Applied logistic regression. New York: John Wiley; 1989.
  19. Woodman RC, Harker LA. Bleeding complications associated with cardiopulmonary bypass. Blood. 1990;76:1680–1697[Abstract/Free Full Text]
  20. Anderson RJ, O’Brien M, MaWhinney S, VillaNueva CB, Moritz TE, Sethi GK, et al. Renal failure predisposes patients to adverse outcome after coronary artery bypass surgery. Kidney Int. 1999;55:1057–1062[Medline]
  21. Kiyomoto A, Sasaki Y, Odawara A, Morita T. Inhibition of platelet aggregation by diltiazem. Comparison with verapamil and nifedipine and inhibitory potencies of diltiazem metabolites. Circ Res. 1983;52(suppl):I115–119
  22. Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med. 1993;118:201–210[Abstract/Free Full Text]
  23. Legault C, Furberg CD, Wagenknecht LE, Rogers AT, Stump DA, Coker L, et al. Nimodipine neuroprotection in cardiac valve replacement: report of an early terminated trial. Stroke. 1996;27:593–598[Abstract/Free Full Text]
  24. Boldt J, Knothe C, Welters I, Dapper F, Hempelmann G. Normothermic versus hypothermic cardiopulmonary bypass: do changes in coagulation differ? Ann Thorac Surg. 1996;62:130–135[Abstract/Free Full Text]
  25. Stensrud PE, Nuttall GA, de Castro MA, Abel MD, Ereth MH, Oliver WC, et al. A prospective, randomized study of cardiopulmonary bypass temperature and blood transfusion. Ann Thorac Surg. 1999;67:711–715[Abstract/Free Full Text]
  26. Vander Salm TJ, Kaur S, Lancey RA, Okike ON, Pezzella AT, Stahl RF, et al. Reduction of bleeding after heart operations through the prophylactic use of epsilon-aminocaproic acid. J Thorac Cardiovasc Surg. 1996;112:1098–1107[Abstract/Free Full Text]
  27. Kizer JR, Kimmel SE. Epidemiologic review of the calcium channel blocker drugs. An up-to-date perspective on the proposed hazards. Arch Intern Med. 2001;161:1145–1158[Abstract/Free Full Text]
  28. Furberg CD, Psaty BM, Meyer JV. Nifedipine. Dose-related increase in mortality in patients with coronary heart disease. Circulation. 1995;92:1326–1331[Abstract/Free Full Text]
  29. Smith RC, Leung JM, Mangano DT. Postoperative myocardial ischemia in patients undergoing coronary artery bypass graft surgery. Anesthesiology. 1991;74:464–473[Medline]
  30. American Heart Association. 2002 heart and stroke statistical update. Dallas: American Heart Association; 2001.

Related Articles

Preoperative calcium-channel blockade in cardiac surgery: the good, the bad, the issues
T. Bruce Ferguson, Jr
J. Thorac. Cardiovasc. Surg. 2004 127: 622-624. [Extract] [Full Text] [PDF]

Calcium antagonists and good results: association or causation?
Donald D. Glower
J. Thorac. Cardiovasc. Surg. 2004 127: 625-626. [Extract] [Full Text] [PDF]



This article has been cited by other articles:


Home page
 J. Thorac. Cardiovasc. Surg.Home page
P. C. Austin
Propensity-score matching in the cardiovascular surgery literature from 2004 to 2006: a systematic review and suggestions for improvement.
J. Thorac. Cardiovasc. Surg., November 1, 2007; 134(5): 1128 - 1135.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Author home page(s):
Vivek Rao
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wijeysundera, D. N.
Right arrow Articles by Karkouti, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wijeysundera, D. N.
Right arrow Articles by Karkouti, K.
Related Collections
Right arrow Coronary disease
Right arrowRelated Articles

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