|
|
||||||||
J Thorac Cardiovasc Surg 2003;126:2032-2043
© 2003 The American Association for Thoracic Surgery
Surgery for acquired cardiovascular disease |
a Department of Cardiothoracic Anesthesia (G-3), The Cleveland Clinic Foundation, Cleveland, Ohio, USA
b Department of Biostatistics, The Cleveland Clinic Foundation, Cleveland, Ohio, USA
c Department of Cardiovascular Anesthesia, Texas Heart Institute at Saint Luke's Episcopal Hospital, Houston, Tex, USA
d Department of Thoracic and Cardiac Surgery, The Cleveland Clinic Foundation, Cleveland, Ohio, USA
Received for publication May 14, 2003; accepted for publication May 28, 2003.
* Address for reprints: Colleen Gorman Koch, MD, MS, Department of Cardiothoracic Anesthesia (G-3), The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195, USA
kochc{at}ccf.org
| Abstract |
|---|
|
|
|---|
METHODS: Propensity-modeling techniques were used to investigate whether gender adversely impacts outcomes after coronary revascularization. A parsimonious explanatory model was developed by bootstrap bagging with variable selection from 64 baseline and 37 operative variables. Propensity scores were calculated from a logistic model that included the parsimonious model and additional baseline variables. Greedy matching techniques were applied to match female and male patients to the nearest propensity scores. Comparisons were made among the propensity-matched women and men.
RESULTS: Of the 15,597 patients undergoing isolated coronary artery bypass graft surgery, only 26% of the 3596 women were matched on propensity scores with men. Distribution of covariates among the matched pairs was, on average, equal. Postoperative mortality (P = .76), neurologic morbidity (global deficit P = .07, focal deficit P = .51), infection (sepsis P = .88), mediastinitis (P = .18), renal failure (P = .84), intra-aortic balloon pump usage (P = .61), and reoperation for bleeding (P = .10) were similar among women and men. Occurrence of Q-wave myocardial infarction (P = < .01), postoperative inotropic usage (P = < .01), and prolonged ventilatory support (P = .02) were more common in women compared with propensity-matched men.
CONCLUSIONS: The preoperative profiles of women and men are markedly different. Propensity matching women and men was difficult, because only 26% of women were able to be matched with men. However, in well-matched patients, female gender was not associated with increased mortality and had minimal impact on morbidity after coronary artery bypass grafting.
| Methods |
|---|
|
|
|---|
Statistical methods
Data for 64 baseline and 37 operative and postoperative variables were collected. The aim of the analysis was to match female and male patients according to baseline characteristics and compare surgical outcomes between the matched groups.
Before matching patients, a parsimonious explanatory model was developed by bootstrap bagging for variable selection.16 Baseline variables found to be significantly associated with gender were identified.17 A propensity score, in logit units, was then obtained by calculating the predicted value for each observation from a logistic model including the variables identified in the parsimonious model plus additional baseline variables, for a total of 68 covariates. These scores represent a summary of baseline characteristic for each observation. Greedy matching techniques were then used to select male counterparts to the female patients by choosing the patient with the nearest propensity score.18 This resulted in 945 matches (26% of the women and 8% of the men). Univariable comparisons were made among the matched and unmatched groups using the Student t and Wilcoxon rank tests for continuous variables and the chi-square and Fisher's exact tests for categoric variables.
Propensity score justification
Associations found through traditional multivariable regression may be misleading because of underrepresentation of 1 covariate of interest within levels of another. Thus, subclassification with propensity scores is used to adjust for confounding background characteristics.19 To illustrate the advantage of using propensity scores, a logistic model was developed for hospital mortality using the entire original data set irrespective of matching. A parsimonious model was constructed by bagging (bootstrap aggregation) with 200 samples to identify significant covariates in the model. A cluster analysis was then used to identify correlations among them. Variables significant in at least 50% of the samples were placed into the model. The significance of gender was then assessed by first forcing in an indicator variable for female, followed by adding a covariate for the propensity score.
Propensity matching coronary reoperations
Coronary revascularization procedures pose an increased risk for both postoperative morbidity and mortality. We investigated whether there were gender differences in this subset of patients with the application of propensity-matching techniques. Of the patients who underwent coronary reoperation, 176 women (36%) and 149 men (5%) were propensity matched. By using the Student t and Wilcoxon rank tests for continuous variables and the chi-square and Fisher's exact tests for categoric variables, comparisons were made among the propensity-matched patients who underwent reoperative CABG.
| Results |
|---|
|
|
|---|
|
|
|
|
|
|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
Two large investigations report that female gender is an independent predictor of adverse outcome after CABG within certain subgroups of patients who underwent CABG.1,2 Edwards and colleagues1 analyzed the data from 247,760 men and 97,153 women from the Society of Thoracic Surgeons National Cardiac Surgery database. Women were older, presented more commonly for non-elective procedures, and were reported to have a higher prevalence of preoperative comorbid conditions, such as diabetes, hypertension, and peripheral vascular disease. Female gender was an independent predictor of postoperative death within all of their defined risk groups except within the group with an operative risk of death greater than 30%. By use of the Multicenter National Cardiovascular Network database, Vaccarino and colleagues2 examined the in-hospital mortality for 36,009 men and 15,178 women who underwent CABG surgery from 1993 to 1999. They reported that women aged less than 50 years were at 3 times greater risk of postoperative mortality compared with men of similar age groupings. After adjusting for a number of preoperative comorbid conditions, women aged less than 50 years had twice the likelihood of postoperative death. Gender differences for in-hospital mortality were less marked in the older age subgroups. Our multivariable model for in-hospital mortality on the data set of 15,587 patients did not find a significant interaction between age and gender.
Conflicting results among published studies with regard to the influence of gender on postoperative morbidity and mortality after CABG are influenced by a number of factors: dissimilar study designs and data collection methods, different study sample sizes, and variable application of statistical methods to analyze the data sets. Most of the investigations are nonrandomized observational studies from large databases, which inherently entail systematic differences in baseline characteristics between women and men. Differing approaches to risk adjustment and the model-selection processes among these investigations can also lead to variability in the final model chosen.20
Our investigation of 15,597 patients who underwent CABG found that a majority of women (74%) are so distinctly different from men in their preoperative profiles that they were unable to be matched on propensity scores with men. Although standard diagnostic procedures for regression would not reveal an imbalanced representation of the levels of one covariate within the levels of another, subclassification by propensity makes this imbalance evident. This demonstrates an advantage of propensity modeling over regression adjustment in that the investigator may find that there is effectively no overlap in the distributions of covariates among the groups under investigation. Application of regression adjustment in this circumstance would result in problems in drawing valid conclusions from the data without making assumptions that involve extrapolation.19,21 To demonstrate this concept, we examined the entire data set of patients and applied multivariable logistic regression techniques to determine if there were gender differences in postoperative mortality. Among the variables selected for the final model, female gender was a significant predictor of mortality after CABG. In the propensity-matched patients, postoperative mortality was similar among the women and men. In fact, after the logistic model was adjusted by the weights established by the propensity score, gender did not remain in the model for mortality.
With the application of propensity score methods, baseline characteristics can be summarized as a single score that approximates the background characteristics for the individual patient.22 Tables 1 and 2 demonstrate that covariates are, in general, balanced between the women and men, similar to the balance achieved in a randomized controlled trial. This allows for a "fair" comparison between women and men with regard to morbidity and mortality outcomes after CABG.23
Principal findings
We report similar mortality among the propensity-matched women and men undergoing CABG. Despite similar preoperative BSA, BMI, red blood cell mass, and rate of reoperation for bleeding, women were transfused with red blood cells at a higher rate compared with propensity-matched men. Women required longer postoperative ventilatory support, which, in part, may be related to the amount of intraoperative narcotics and benzodiazepines administered. There were also more cardiac arrests, postoperative myocardial infarctions, and intraoperative inotropic agents administered in the propensity-matched women compared with the men. The increase in cardiac events in the propensity-matched women is similar to what has been shown from other investigations. Brandrup-Wognsen and colleagues,4 in a study of 2129 patients undergoing CABG in western Sweden, reported that women had more indirect signs of myocardial injury after coronary revascularization. Their female patients had more frequent use of circulatory support with an IABP and greater use of postcardiopulmonary bypass inotropics. Christakis and colleagues6 found that female gender was an independent predictor of postoperative mortality and low cardiac output syndrome in their investigation of 7025 patients undergoing isolated CABG from 1990 to 1994. Women had a higher occurrence of postoperative myocardial infarction and need for IABP assistance.6 Woods and colleagues,3 in their prospective cohort study of 5324 patients who underwent CABG, reported that female gender was not an independent predictor of mortality; however, they did find that it was associated with an increase in low cardiac output conditions after revascularization. Finally, Abramov and colleagues11 reported that female gender was a weak independent predictor of perioperative myocardial infarction and IABP usage, but not for mortality after CABG.
Outcomes in unmatched women
Recognizing that a majority of women were unable to be matched on propensity scores with men, we further examined this group of unmatched women and compared them with the propensity-matched women. The unmatched women had a distinctly different preoperative profile compared with the matched women. They had a large number of variables that typify the preoperative profile of the female patient undergoing CABG: lower hematocrits and red blood cell mass, shorter stature, lower bilirubin, increased triglycerides, higher high-density lipoprotein cholesterol, and less prevalence of smoking cigarettes and drinking alcohol.17 Despite these differences in profiles, the unadjusted postoperative mortality and morbidity outcomes were similar.
Outcomes in propensity-matched coronary reoperations
Patients undergoing coronary reoperations represent a subgroup of patients who are reported to be at increased risk of postoperative morbidity and mortality.24-26 The preoperative risk profile for patients undergoing coronary reoperations has been typified by patients who are more frequently male, with more left ventricular dysfunction, with a greater Canadian Cardiovascular Society symptom class, and who present more commonly for urgent surgery compared with patients undergoing primary CABG.26 In particular, female gender has been reported to be an independent predictor of increased postoperative mortality after reoperative CABG. Aidala and colleagues,24 in a small retrospective study, reported an 11.4% mortality for patients undergoing reoperative CABG versus a 3.2% mortality for patients undergoing primary operations. They reported that female gender was a risk factor for postoperative mortality. Weintraub and colleagues,25 from their database of 2030 coronary reoperations, reported that female gender was correlated with in-hospital death both in the univariable and multivariable analyses. Although Yau and colleagues26 did not report female gender to be an independent predictor of postoperative mortality in 1230 patients undergoing reoperative CABG, they reported that it was an independent predictor of postoperative low cardiac output syndrome. We report that among the propensity-matched pairs undergoing coronary reoperations, women have a lower postoperative mortality compared with similarly matched men. Postoperative morbidity was similar among well-matched women and men undergoing coronary reoperations.
Limitations
One of the limitations of this observational investigation was that there may have been unmeasured confounding variables that impacted the propensity matching of women and men who underwent CABG. In addition, the amount of anesthetic medications administered intraoperatively was not a measured variable. The omission of this information may have impacted the findings of gender differences in the duration of postoperative ventilatory support observed in this investigation. Furthermore, we did not record data on all "processes of care" in the operating room. For example, complications from repeat sternotomy, such as graft excision or right ventricular damage, were not recorded variables. This lack of information may have influenced the observed gender differences in postoperative mortality for the propensity-matched patients who underwent reoperative CABG.
| Conclusions |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. A. Fox Pro: Newly Appreciated Pathophysiology of Ischemic Heart Disease in Women Mandates Changes in Perioperative Management Anesth. Analg., July 1, 2008; 107(1): 29 - 32. [Full Text] [PDF] |
||||
![]() |
C. G. Koch Con: Newly Appreciated Pathophysiology of Ischemic Heart Disease in Women Mandates Changes in Perioperative Management Anesth. Analg., July 1, 2008; 107(1): 33 - 36. [Full Text] [PDF] |
||||
![]() |
M. H Mandegar, M. Marzban, A. H Lebaschi, P. Ghaboussi, A. R. Alamooti, and A. Ardalan Gender Influence on Hospital Mortality after Coronary Artery Bypass Surgery Asian Cardiovasc Thorac Ann, June 1, 2008; 16(3): 231 - 235. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Parolari, L. Dainese, M. Naliato, G. Polvani, C. Loardi, M. Trezzi, M. Fusari, C. Beverini, E. Tremoli, P. Biglioli, et al. Do Women Currently Receive the Same Standard of Care in Coronary Artery Bypass Graft Procedures as Men? A Propensity Analysis Ann. Thorac. Surg., March 1, 2008; 85(3): 885 - 890. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ranucci, A. Pazzaglia, C. Bianchini, G. Bozzetti, and G. Isgro Body Size, Gender, and Transfusions as Determinants of Outcome After Coronary Operations Ann. Thorac. Surg., February 1, 2008; 85(2): 481 - 486. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. L. Lazar, T. Keilani, C. A. Fitzgerald, O. M. Shapira, C. T. Hunter, R. J. Shemin, H. C. Marsh Jr, U. S. Ryan, and the TP10 Cardiac Surgery Study Group Beneficial Effects of Complement Inhibition With Soluble Complement Receptor 1 (TP10) During Cardiac Surgery: Is There a Gender Difference? Circulation, September 11, 2007; 116(11_suppl): I-83 - I-88. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Basaran, O. Selimoglu, H. Ozcan, H. Ogus, E. Kafali, C. Ozcelebi, and T. N. Ogus Being an elderly woman: is it a risk factor for morbidity after coronary artery bypass surgery? Eur. J. Cardiothorac. Surg., July 1, 2007; 32(1): 58 - 64. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Goto, T. Baba, A. Ito, K. Maekawa, and T. Koshiji Gender Differences in Stroke Risk Among the Elderly After Coronary Artery Surgery Anesth. Analg., May 1, 2007; 104(5): 1016 - 1022. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. H. Humphries, M. Gao, A. Pu, S. Lichtenstein, and C. R. Thompson Significant Improvement in Short-Term Mortality in Women Undergoing Coronary Artery Bypass Surgery (1991 to 2004) J. Am. Coll. Cardiol., April 10, 2007; 49(14): 1552 - 1558. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Ad, S. D. Barnett, and A. M. Speir The performance of the EuroSCORE and the Society of Thoracic Surgeons mortality risk score: the gender factor Interactive CardioVascular and Thoracic Surgery, April 1, 2007; 6(2): 192 - 195. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Pilote, K. Dasgupta, V. Guru, K. H. Humphries, J. McGrath, C. Norris, D. Rabi, J. Tremblay, A. Alamian, T. Barnett, et al. A comprehensive view of sex-specific issues related to cardiovascular disease Can. Med. Assoc. J., March 13, 2007; 176(6): S1 - S44. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. A. Seifert, D. L. Howard, J. H. Silber, and D. R. Jobes Female gender increases the risk of death during hospitalization for pediatric cardiac surgery J. Thorac. Cardiovasc. Surg., March 1, 2007; 133(3): 668 - 675. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. I. Duncan, J. Lin, C. G. Koch, A. M. Gillinov, M. Xu, and N. J. Starr The Impact of Gender on In-Hospital Mortality and Morbidity After Isolated Aortic Valve Replacement Anesth. Analg., October 1, 2006; 103(4): 800 - 808. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. D. McCully, Y. Toyoda, H. Wakiyama, A. J. Rousou, R. A. Parker, and S. Levitsky Age- and gender-related differences in ischemia/reperfusion injury and cardioprotection: effects of diazoxide. Ann. Thorac. Surg., July 1, 2006; 82(1): 117 - 123. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Brunelli, F. Xiume', M. Al Refai, M. Salati, R. Marasco, and A. Sabbatini Gemcitabine-Cisplatin Chemotherapy Before Lung Resection: A Case-Matched Analysis of Early Outcome Ann. Thorac. Surg., June 1, 2006; 81(6): 1963 - 1968. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Doenst, J. Ivanov, M. A. Borger, T. E. David, and S. J. Brister Sex-specific long-term outcomes after combined valve and coronary artery surgery. Ann. Thorac. Surg., May 1, 2006; 81(5): 1632 - 1636. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. A. M. Rogers, K. M. Langa, C. Kim, B. K. Nallamothu, L. F. McMahon Jr, P. N. Malani, B. E. Fries, S. R. Kaufman, and S. Saint Contribution of infection to increased mortality in women after cardiac surgery. Arch Intern Med, February 27, 2006; 166(4): 437 - 443. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. A. Nussmeier Are women different from men in ways that matter? Maybe J. Thorac. Cardiovasc. Surg., February 1, 2006; 131(2): 264 - 265. [Full Text] [PDF] |
||||
![]() |
R. H. Habib, A. Zacharias, T. A. Schwann, C. J. Riordan, S. J. Durham, and A. Shah Effects of Obesity and Small Body Size on Operative and Long-Term Outcomes of Coronary Artery Bypass Surgery: A Propensity-Matched Analysis Ann. Thorac. Surg., June 1, 2005; 79(6): 1976 - 1986. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. H. Edwards, V. A. Ferraris, D. M. Shahian, E. Peterson, A. P. Furnary, C. K. Haan, and C. R. Bridges Gender-Specific Practice Guidelines for Coronary Artery Bypass Surgery: Perioperative Management Ann. Thorac. Surg., June 1, 2005; 79(6): 2189 - 2194. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. J. Birdas, J. T. Landis, D. Haybron, D. Evers, P. K. Papasavas, and P. F. Caushaj Outcomes of Coronary Artery Bypass Grafting in Patients With Connective Tissue Diseases Ann. Thorac. Surg., May 1, 2005; 79(5): 1610 - 1614. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Bucerius, J. F. Gummert, T. Walther, M. A. Borger, N. Doll, V. Falk, and F. W. Mohr Impact of Off-Pump Coronary Bypass Grafting on the Prevalence of Adverse Perioperative Outcome in Women Undergoing Coronary Artery Bypass Grafting Surgery Ann. Thorac. Surg., March 1, 2005; 79(3): 807 - 812. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Brunelli, A. Sabbatini, F. Xiume', A. Borri, M. Salati, R. D. Marasco, and A. Fianchini Inability to perform maximal stair climbing test before lung resection: a propensity score analysis on early outcome Eur. J. Cardiothorac. Surg., March 1, 2005; 27(3): 367 - 372. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. H. Habib, A. Zacharias, T. A. Schwann, C. J. Riordan, S. J. Durham, and A. Shah Worse early outcomes in women after coronary artery bypass grafting: Is it simply a matter of size? J. Thorac. Cardiovasc. Surg., September 1, 2004; 128(3): 487 - 488. [Full Text] [PDF] |
||||
![]() |
K. M. King, P. D. Faris, and W. A. Ghali Sex Differences in Mortality After Coronary Artery Bypass Graft Surgery--Reply JAMA, July 7, 2004; 292(1): 41 - 41. [Full Text] [PDF] |
||||
![]() |
R. H. Habib, A. Zacharias, T. A. Schwann, C. J. Riordan, S. J. Durham, and A. Shah Sex Differences in Mortality After Coronary Artery Bypass Graft Surgery JAMA, July 7, 2004; 292(1): 40 - 41. [Full Text] [PDF] |
||||
![]() |
A. Azakie and I. A. Russell Gender differences in pediatric cardiac surgery: The surgeon's perspective J. Thorac. Cardiovasc. Surg., July 1, 2004; 128(1): 4 - 6. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||