Author + information
- Received July 3, 2002
- Revision received September 30, 2002
- Accepted October 10, 2002
- Published online January 15, 2003.
- Viola Vaccarino, MD, PhD*,†,* (, )
- Zhen Qiu Lin, PhD‡,
- Stanislav V Kasl, PhD§,
- Jennifer A Mattera, MPH‡,
- Sarah A Roumanis, RN‡,
- Jerome L Abramson, PhD* and
- Harlan M Krumholz, MD, FACC‡∥
- ↵*Reprint requests and correspondence:
Dr. Viola Vaccarino, Emory University School of Medicine, Department of Medicine, Division of Cardiology, 1256 Briarcliff Road, Suite 1 North, Atlanta, Georgia 30306, USA.
Objectives This study was designed to examine whether female gender is associated with poorer recovery after coronary artery bypass graft (CABG) surgery.
Background The risks and benefits associated with CABG surgery in women are not as well established as they are in men, and there are concerns that women may have worse outcomes. The recovery period after CABG (the first four to eight weeks after the surgery) is a vulnerable time, with higher risks of complications and hospital readmission. There is little information on patients’ experiences during this phase, particularly among women.
Methods We prospectively followed 1,113 patients (804 men and 309 women) who underwent first CABG consecutively between February 1999 and February 2001. Patients were interviewed at baseline and between six and eight weeks after surgery. Clinical data were abstracted from medical records.
Results Compared with men, women were older and more often had unstable angina and congestive heart failure, lower physical function (PF), and more depressive symptoms in the month before surgery. At six to eight weeks after CABG surgery, after adjustment for baseline characteristics, the rate of hospital readmission was 20.5% in women and 11.0% in men (p = 0.005), and the mean number of physical symptoms and side effects was 2.5 in women and 2 in men (p = 0.0009). Whereas, on average, PF remained unchanged in men (an increase in score of 0.3 points, 95% confidence interval [CI], −1.1 to 1.8) and depressive symptoms improved (a decrease of 0.2 depressive symptoms, 95% CI, −0.4 to −0.04), women showed, on average, a 13-point decline in physical function (95% CI, −15.8 to −10.4) and an increase of 0.5 in depressive symptoms (95% CI, 0.1 to 0.9).
Conclusions After CABG surgery, women have a more difficult recovery compared with men, which is not explained by illness severity, presurgery health status, or other patient characteristics.
Coronary artery bypass graft (CABG) surgery is a procedure increasingly performed in patients with coronary heart disease (1). In 1999, more than 180,000 CABG procedures, or about 30% of the total in the U.S., were performed on women (1). Major randomized trials of the efficacy of CABG surgery compared with medical therapy have predominantly included men (2–4). Therefore, the risks and benefits associated with CABG in women are not as well established as they are in men. Observational studies have provided some comparative outcome data on women and men after CABG surgery, and there are concerns that women may have worse outcomes (5).
The focus of previous investigations has been primarily on in-hospital complications and mortality (5–10)or long-term survival and symptoms (6,9,11,12). Patients facing the decision of undergoing CABG surgery, however, are typically interested in a variety of additional outcomes, including side effects and complications beyond the hospitalization period, and the ability to recover from surgery in terms of physical function (PF) and mental health. The recovery period (the first four to eight weeks after the surgery) is a vulnerable time, with higher risks of complications and hospital readmission. Knowledge of the risks associated with CABG surgery in the recovery phase would be valuable to patients and would help the development of interventions to enhance recovery. However, there is little information on patients’ experiences during the recovery phase, and on whether the experiences of men and women differ.
The purpose of this study, therefore, was to examine whether female gender is associated with poorer recovery, with respect to symptoms, side effects, repeated hospitalizations, and functional status; and whether differences in other characteristics, such as age, comorbidity, and health status, might explain recovery differences between men and women.
We studied consecutive patients admitted for their first CABG surgery at Yale New Haven Hospital between February 1999 and February 2001. Yale New Haven Hospital, a 900-bed hospital affiliated with Yale University, is the largest community hospital serving New Haven, Connecticut. Potential candidates for the study were identified through a daily screening of the admission records. Patients were excluded if they were younger than 30 years, had a previous CABG, had other cardiac or noncardiac operations performed at the time of the CABG procedure, or were unable to be interviewed because of aphasia, impaired mental status, language barrier, or if they died before consenting to participate in the study. The institutional review board approved the study, and all patients provided informed consent.
After surgery and before discharge, patients were administered a baseline interview in order to collect information on demographic factors, self-reported reasons for admission, and health status in the month before surgery, including functional status, symptoms, side effects, and psychosocial factors. Physical function was determined by means of the PF subscale of the Medical Outcomes Trust Short Form 36-item Health Survey (SF-36) (13), which yields a score ranging from 0 (severe impairment) to 100 (no impairment). Depressive symptoms were assessed by means of the Geriatric Depression Scale (GDS) Short Form (14), which provides a count of depressive symptoms from 0 to 15. Social support was measured by the Enhancing Recovery In Coronary Heart Disease (ENRICHD) Social Support Inventory (ESSI) (15), a five-item scale developed for the measurement of social and emotional support in cardiac patients. The score ranged from 0 (no social support) to 22 (maximum level of social support). Medical charts were reviewed after hospital discharge using the Society of Thoracic Surgery abstraction instrument and clinical definitions (16).
Follow-up was done by phone to assess recovery measures, including functional status and depressive symptoms, measured in the same manner as at baseline; postsurgery symptoms; complications and side effects; and hospital readmission. Patients were contacted six to eight weeks after CABG surgery, with a window of contact up to 10 weeks. This interval was chosen because it corresponds to the time by which most patients are usually expected to have recovered from CABG surgery (17). Patients were asked whether they had experienced chest pain (not related to incision pain), chest tightness, or angina in the past week. Presence of dyspnea at rest was assessed by means of the Dyspnea Index magnitude of task subscale (18). Surgery-related complications and side effects were determined by asking patients whether they had experienced any of the following symptoms and side effects in the previous week: incision pain, incision infection, numbness at incision, sleep problems, constipation, nausea or lack of appetite, and diminished taste. Only problems that were not pre-existing to surgery were considered in this analysis. A score indicating the total number of symptoms and side effects was computed with a range between 0 and 9. If patients had been readmitted to the hospital, they were asked to report the primary reason for hospital admission. Reasons for readmission were categorized as follows: 1) incision infection or other surgery-related infections; 2) congestive heart failure, shortness of breath, or pneumonia; 3) myocardial ischemia (myocardial infarction, angina, or chest pain), with or without repeat revascularization; 4) arrhythmias; 5) other complications, possibly surgery-related; and 6) other reasons or procedures, probably not surgery-related.
We compared sociodemographic and clinical characteristics at baseline and outcome measures at follow-up between women and men using the ttest for continuous variables and the chi-squared test for categorical variables. We used logistic regression models to calculate adjusted readmission probabilities of hospital readmission in women compared with men after adjusting for key preoperative variables, including age; marital status; years of education; baseline SF-36 PF score; baseline GDS score; current smoking; body mass index; number of hospitalizations in year before admission; history of hypertension, diabetes, cerebrovascular accident, and myocardial infarction; congestive heart failure on admission; unstable angina on admission; Canadian Cardiovascular Society angina class; number of diseased vessels; maximal left main stenosis; left ventricular ejection fraction (LVEF); and body surface area. Furthermore, we used analysis of covariance models to calculate mean changes in the SF-36 PF score and the GDS score from preoperative values to follow-up values in women and men after adjusting for the same factors listed earlier. In a similar manner, we derived the adjusted mean number of symptoms and side effects at follow-up. Because the symptom score was slightly skewed, we also fitted a model after log-transformation. However, because the results did not differ substantially, only the untransformed data are presented.
Because it has been reported that gender differences in in-hospital outcomes after CABG surgery are more marked among younger patients (5), in a set of secondary analyses we stratified according to whether the patients were <65 years old or ≥65 years old. All tests for statistical significance were two-tailed with an alpha level of 0.05. Linear and logistic regression model assumptions and adequacy of fit were checked using standard methods (19,20)and found to be satisfactory. All analyses were conducted using SAS software, release 8.02 (SAS Institute, Cary, North Carolina).
In the study period, 1,290 consecutive patients met the eligibility criteria for participation. Of these patients, 1,164 (90.2%) were enrolled. Of the 124 eligible patients who were not enrolled, 49 refused participation (13 women and 36 men), 41 were discharged before giving consent for participation (7 women and 34 men), and the remainder had miscellaneous reasons for not being included in the study.
From the sample of 1,164 enrolled patients, we excluded 51 patients, 38 men (4.5%) and 13 women (4.0%) for whom follow-up data were not available because of death, refusal, or inability to contact. Therefore, our sample for this analysis included 1,113 patients, 804 men and 309 women. Only 6% of patients could not be reached by 8 weeks and were interviewed in a time window between 8 and 10 weeks. The time interval between surgery and follow-up interview did not differ between men and women: 6.6 (±1.1) and 6.6 (±1.2) weeks, respectively.
Women were older than men, less educated, more likely to be unmarried and to live alone, and had lower social support scores (Table 1). Women were also less likely than men to have undergone an elective CABG procedure or CABG as a consequence of abnormal test results only.
There were marked gender differences in health status at baseline (Table 1). Women were more likely than men to have unstable angina and congestive heart failure, had higher angina class, lower SF-36 PF score, and higher number of depressive symptoms. However, women had a lower number of diseased vessels, and the LVEF did not differ significantly by gender. Women were less likely to receive internal mammary artery grafts and overall received a lower number of grafts, but other procedural characteristics did not differ by gender. Overall postoperative complications were rare and did not differ markedly between men and women (Table 1).
At the postsurgery follow-up, women were almost twice as likely as men to have been readmitted to the hospital (22.0% vs. 12.7%). Patients were able to report the primary reason for hospitalization in 90.8% of these admissions. In both men and women, the most common reason for readmission was infection at the incision site or other surgery-related infections (23.8% and 29.7% of all readmissions in men and women, respectively). More women than men were readmitted for congestive heart failure or pneumonia (25.7% vs. 12.2%); however, men were more likely than women to be readmitted for arrhythmias (20.0% vs. 8.1%). Only seven admissions in men (6.7%) and four in women (5.4%) occurred for acute ischemia with or without revascularization. The proportion of readmissions due to other reasons was similar in men and women.
Compared with men, women also more often had angina and dyspnea, surgery-related side effects, and continued to report lower PF, more depressive symptoms, and lower social support compared with men (Table 2). Because women and men differed markedly in baseline score levels, we examined score levels at follow-up after stratification according to the corresponding baseline values. Women had a significantly lower PF score than men at follow-up for each approximate quintile of baseline PF score (Fig. 1). These differences are missed in the unstratified sample. Figure 1also shows that only the women in the lowest category of presurgery PF (score <40) showed an improvement at follow-up. In all the other groups, women experienced a decline in functioning. Similarly, higher levels of depressive symptoms at follow-up were noted in women after stratification according to baseline GDS levels (data not shown).
After adjustment for baseline characteristics, significant differences persisted between women and men in outcome measures at follow-up (Table 3). Of the women, 20.5% were readmitted to the hospital versus 11.0% of the men (p = 0.005). Female gender was the strongest predictor of readmission in the multivariable model. Whereas PF remained substantially unchanged in men (a mean increase of 0.3 points in the PF score), women showed, on average, a 13-point decline in PF score from baseline to follow-up, p < 0.0001 for the comparison of changes between women and men. When these results were expressed as the proportion of patients who had a marked decline (defined as a decline of 20 points or more in the PF score, corresponding to about the lowest quartile in the total distribution), the results were consistent. The adjusted probability of decline was 25.7% in women and 11.4% in men (relative risk, 2.25; 95% confidence interval [CI], 1.63 to 4.47, p = 0.0001). Finally, although the number of depressive symptoms significantly improved in men, it worsened in women (Table 3). Again, the results were consistent when, instead of mean changes, we examined the proportion of patients who were at the most unfavorable quartile of change (a worsened GDS of two symptoms or more). The adjusted proportion of patients with worsened GDS was 33.1% in women and 21.1% in men (relative risk, 1.57; 95% CI, 1.18 to 2.91, p = 0.007).
To test the role of social support in our results, we constructed a set of models that included the ESSI social support scores in addition to the other variables listed in the footnote to Table 3. Inclusion of social support in the models, however, did not materially change the estimate for gender for any of the outcome measures examined (data not shown).
Results according to age
Of the 309 women in the study, 91 (29.5%) were younger than 65 years, whereas of the 804 men, 394 (49.0%) were younger than 65 years. Multivariable-adjusted results according to these age groups revealed that gender differences in outcomes tended to be more marked in the younger than in the older patients. The adjusted risk for readmission was more than twice as high in women among patients younger than 65 years (19.7% in women vs. 8.8% in men, p = 0.03), whereas among the older patients the risk for women was 70% higher (20.4% in women vs. 12.1% in men, p = 0.08). In younger men PF improved at six weeks (PF score = +3.9) and depressive symptoms declined (GDS score = −0.5); in women they worsened (PF score = −11.0 and GDS score = +0.67). These differences were overall less marked in older patients. The interaction between gender and age, however, was not significant in these analyses (p > 0.20).
Our results indicate that women experience a more difficult recovery after discharge from CABG surgery compared with men. Six to eight weeks postsurgery, women reported more physical problems and side effects than men, lower PF, more depressive symptoms, and were almost twice as likely as men to have been readmitted to the hospital. These differences remained substantial and statistically significant in multivariable analysis. Women experienced, on average, a decline in their PF and an increase in depressive symptoms relative to their recall of preoperative health status, whereas men reported slight improvements in both these domains. These differences tended to be more marked among younger patients, which is consistent with another recent report of worse outcomes after CABG surgery in younger women compared with men (5).
Previous literature on gender differences in physical and emotional recovery after CABG surgery has yielded conflicting results. In a study of 132 men and 47 women, measures of physical recovery improved significantly between one and six weeks after CABG in both men and women, but women showed a slower recovery (21). At each point in time, women reported greater ambulation dysfunction and more physical symptoms, and with time, the decrease in symptoms was less in women than in men (21). Similarly, in a Swedish study, PF was improved and symptoms of chest pain and dyspnea were reduced in both men and women three months after CABG, but improvements were greater in men than in women (22). In another small study, King et al. (23)also found greater disruption in ambulation and home management in women than in men at one month after surgery, although there were no gender differences in post-CABG angina. Two additional studies, however, showed no significant gender differences in physical outcome measures within three months after CABG (24,25).
These conflicting results may be explained in part by differences in methodology. Most of these studies were based on convenience samples of fewer than 200 patients (21,23–25), which may have been too selected to provide meaningful information. In addition, with one exception (22), these studies did not take into account preoperative differences in health status between the genders. It has been pointed out that, because of underlying differences in preoperative functioning (for example, a higher incidence of degenerative joint disease in older women), the level of improvement from preoperative values, rather than absolute functional capacity, should be considered as an indication of recovery when comparing women with men (23). In addition, when comparing changes over time between groups that differ markedly in the initial values, failure to adjust for baseline levels may produce biased results because of the confounding effects of regression to the mean (26). Our study confirms previous observations of a slower physical recovery in women than in men following CABG surgery after taking into account such baseline variables as presurgery health status and PF.
In addition to PF, our data indicate that CABG surgery has a greater adverse impact on mood in women than in men in the recovery phase after CABG. Only a few small studies have examined this issue before. Some authors have reported higher depression scores in women than in men at each time point in the first few weeks after CABG, but the rate of change of depressive symptoms over time was comparable by gender (21,25). Other studies have found no differences, overall, between women and men in emotional outcomes after CABG (23,24). These studies, however, were limited by small, selected samples and often failed to take into account presurgery depression levels and other preoperative characteristics.
Why do women show more physical problems and fewer improvements in the recovery phase after bypass surgery? Consistent with other studies (5,10,11,21,23,25,27), in our sample women had, preoperatively, more severe angina and more coexisting illnesses compared with men, although there were no gender differences in previous history of myocardial infarction and in LVEF. Women actually had fewer diseased coronary arteries than men. Therefore, although women were at higher risk in terms of symptoms and coexisting diseases and risk factors, they were at lower risk in terms of extent of coronary artery disease. Pre-existing conditions and risk factors, however, did not explain differences in recovery by gender in our study.
The explanation for the gender differences in recovery after CABG surgery could relate, in part, to the different social roles of women and men (23). Women, traditionally, have more responsibilities in home management and family caregiving than men. Therefore, women may feel greater disruption than men when they cannot resume their roles upon returning home after surgery. It is conceivable that family responsibilities are more pressing in younger women, and this fact may explain why the gender differences in recovery were more marked among younger patients in our study. Consistent with this hypothesis, other authors have found that home management is a functional area in which disparities between men and women are greatest during recovery after bypass surgery (21,23), and women have been found to struggle to maintain their usual family role after a coronary event (28)and to resume household activities early during recovery. This explanation may also account for why, in our study, and consistent with others (21,23), women tended to have fewer sleep problems than men. If women resume household activities early during recovery, they may feel more tired and therefore have less trouble sleeping compared with men. This observation has also been noted in other studies (23). In addition, women in our study were more likely to be unmarried, to live alone, and to report a lower level of social support than men. This has been shown in other studies as well (21,23,27,29). As a consequence, women may have fewer sources of help compared with men after CABG surgery. However, in our study, the level of social support at baseline did not substantially explain gender differences in recovery. This could be due to the fact that the social support scale we used is geared towards the measurement of emotional support and may not capture the level of help in daily life available to patients.
Our study has several possible limitations. Because willingness to participate was necessary in order to be included in the study, our results could be affected by response bias. However, the response rate was excellent and very few patients were missed at follow-up. Because patients were asked to recall their preoperative health status after surgery, women and men may have differed in their propensity to recall health problems. It has been shown, however, that conclusions about health status changes, measured with scales similar to those used in our study, are materially unaffected when data on preadmission health status are collected retrospectively (30); therefore, this method should not have introduced bias. Finally, referral bias might play a role in our results. Because women with symptoms of coronary heart disease may be referred less often or later than men for revascularization procedures (31), it may be that only women with more severe coronary disease receive CABG surgery. The higher risk status of the women, therefore, could explain their poorer short-term outcomes, even after adjusting for known risk factors. However, as noted earlier, the women in our study, despite being older, had less extensive coronary artery disease than men. This observation, reported in several other studies as well (6,9,32–34), argues against a later referral of women compared with men.
In conclusion, women appear to have a more difficult recovery after CABG surgery compared with men, which is not explained by illness severity, presurgery health status, or other patient characteristics. Special efforts should be directed to understanding the reasons for these differences. This information should lead to the development of interventions aimed at improving the health status of women after CABG surgery.
☆ This study was supported by a grant from the Ethel F. Donaghue Women’s Health Investigator Program, New Haven, Connecticut; and by a grant from the Quality Care Research Fund, Aetna Foundation, Hartford, Connecticut.
- coronary artery bypass graft
- confidence interval
- Geriatric Depression Scale
- left ventricular ejection fraction
- physical function
- Short Form 36-item Health Survey
- Received July 3, 2002.
- Revision received September 30, 2002.
- Accepted October 10, 2002.
- American College of Cardiology Foundation
- American Heart Association
- CASS Principal Investigators and their Associates
- Vaccarino V.,
- Abramson J.L.,
- Veledar E.,
- Weintraub W.S.
- O’Connor G.T.,
- Morton J.R.,
- Diehl M.J.,
- et al.
- Loop F.D.,
- Golding L.R.,
- MacMillan J.P.,
- Cosgrove D.M.,
- Lytle B.W.,
- Sheldon W.C.
- Ware J.E.
- ↵(1999) ACC/AHA guidelines for coronary artery bypass graft surgery. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 34:1262–1347.
- Weisberg S.
- Hanley J.,
- McNeil B.
- Davis C.E.
- Hogue C.W. Jr.,
- Barzilai B.,
- Pieper K.S.,
- et al.