| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:3-9
© 2007 The Society of Thoracic Surgeons
a Department of Surgery, Brody School of Medicine, East Carolina University, Greenville, North Carolina
b Division of Cardiothoracic Surgery, Brody School of Medicine, East Carolina University, Greenville, North Carolina
c Divisions of Community Health and Preventive Medicine, Brody School of Medicine, East Carolina University, Greenville, North Carolina
Accepted for publication March 9, 2007.
* Address correspondence to Dr Kypson, Division of Cardiothoracic Surgery, Brody School of Medicine, East Carolina University, 600 Moye Blvd, LSB-Rm 177, Greenville, NC 27834 (Email: kypsona{at}ecu.edu).
Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8–11, 2006.
| Adult cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
45). We undertook this study to evaluate outcomes in this cohort. Methods: A retrospective analysis was performed on 14,571 patients in our database who underwent cardiac operations from 1992 to 2005. Patient demographics, comorbidities, and outcomes were recorded. A univariate analysis between two groups: BMI 21 to 34.9 and BMI 45 or more was performed. Logistic regression models were used to identify independent risk factors for 30-day mortality. Long-term follow-up of the extreme obese group was achieved.
Results: We identified 128 extreme obese patients, and 480 patients with a BMI of 21.0 to 34.9 were randomly selected for comparison. Univariate analysis showed significant differences in age, gender, and multiple comorbidities, as well as in cardiopulmonary bypass and cross-clamp times, operative procedure, and transfusion requirements. Extreme obese patients had a higher incidence of infection, acute renal failure, and 30-day mortality. Logistic regression analysis showed BMI, preoperative renal insufficiency, and transfusion status to be independent risk factors for 30-day mortality. Follow-up data did not reveal significant functional improvements. Long-term survival was 33.6% at 12 years.
Conclusions: Extreme obese patients undergoing cardiac surgical procedures have higher perioperative morbidity and mortality compared with a lower BMI group. BMI and preoperative renal insufficiency increase mortality in both groups, whereas transfusion does so only in the extreme obese. These patients can realize acceptable outcomes from cardiac procedures, but continue to suffer from the comorbidities of obesity.
| This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet.org/sections/newsandviews/discussions/index.html
|
Hippocrates recognized that obesity is a medical condition that harbors many associated disease processes [1]. The incidence of obesity has risen to epidemic proportions during the past few decades [2], and in the United States, obesity was set to overtake smoking as the main preventable cause of illness and premature death in 2005 [3]. Recent results from the National Health and Nutrition Examination Survey indicate that the prevalence of morbid obesity, defined as a body mass index (BMI) of 40 or more (calculated as kg/m2), is approaching 10% of the population [4].
The main adverse consequences of obesity are cardiovascular disease, type II diabetes mellitus, and several types of cancer. Adams and colleagues [5] documented an approximate 2.5-fold increase in mortality in nonsmoking men with a BMI of 40. The metabolic syndrome that is associated with central obesity and accumulated abdominal fat is characterized by insulin resistance, hypertriglyceridemia, low high-density lipoprotein cholesterol, hypertension, and a proinflammatory and a prothrombotic state [6]. Metabolic syndrome is also associated with an increased risk of type II diabetes and has been reported in up to 40% of patients with cardiovascular disease [7].
These conditions of obesity, type II diabetes, and metabolic syndrome have been linked with operative or late mortality after coronary artery bypass grafting (CABG) [8]. Obesity and diabetes have been extensively examined and have generally been associated with increased morbidity, but the effect on operative mortality has generated equivocal results [9]. Of importance is that most of these studies have concentrated on patient populations with BMIs between 30 and 40. Few studies have examined outcomes in the extreme obese, those patients with a BMI of 45 or more. In addition, long-term follow-up and quality of life have not been fully described in this obesity subset. We therefore investigated perioperative morbidity and mortality, as well as long-term follow-up, of extreme obese patients who underwent cardiac procedures at our institution during the past 13 years.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
We identified 169 patients with a BMI of 45 or more who had undergone CABG, valve repair or replacement, combined CABG/valve, or other cardiac procedures requiring median sternotomy and cardiopulmonary bypass. Patients undergoing off-pump CABG, redo operations, minimally invasive incisions, and emergency surgical procedures were excluded, resulting in 128 patients in the cohort for this study. These exclusion criteria were created to allow for comparison with patient populations with lower BMI.
The control group was 512 patients with a BMI between 21 and 35 who underwent cardiac operations from 2004 and 2006 and were selected from the same database. This time period and BMI range gave us the best representation of the contemporary cardiac surgical population at our facility. To select four controls for each study patient, a computerized random numbers list was created to match patient file numbers. Control patients were then selected by implementing a systematic random sampling approach. For both study and control groups, the same inclusion and exclusion criteria were applied. Those undergoing off-pump CABG, redo operations, minimally invasive incisions, and emergency procedures were excluded, resulting in 480 patients.
Preoperative data collected included age, gender, ejection fraction, smoking status, diabetic status, and comorbidities such as hypertension, chronic lung disease, pulmonary hypertension, sleep apnea, renal insufficiency, and dialysis requirement.
Operative information included cardiopulmonary bypass and cross clamp times, type of procedure, use of the internal mammary artery (IMA) or saphenous vein, total intensive care unit (ICU) time, total ventilator time, perioperative transfusion requirements, reintubation, readmission to the ICU, and hospital length of stay.
Postoperative data included deep and superficial sternal infections, leg infections, myocardial infarction, stroke, septicemia, urinary tract infection, acute renal failure, dialysis requirement, 30-day hospital readmission rate, respiratory failure (>48 hours on the ventilator), and 30-day mortality.
Statistical Analysis
A univariate analysis of all variables between the extreme obese and control groups was performed. Continuous variables were analyzed by the Student t test. Dichotomous variables were analyzed by 
2 analysis with the Yates continuity correction. Unconditional logistic regression models were fit to test for associations between preoperative and operative risk factors and 30-day mortality. Odds ratios (OR) and 95% confidence intervals (CI) were calculated for each risk factor. Because we observed a significant difference in age and gender between BMI groups, all logistic regression models were adjusted for age and gender. Survival data were analyzed with the Kaplan-Meier technique for the extreme obese group. All statistical analyses were performed using SPSS 13.0 (SPSS Inc, Chicago, IL).
Follow-Up
Patients in the extreme obese group were contacted and asked to complete a phone questionnaire that inquired about postoperative hospitalizations and operations, current symptoms of cardiac disease, weight loss or gain, current medications, activity level, and general assessment of quality of life compared with their preoperative period. One hundred and eleven patients (87%) were contacted. Current mortality status was inferred for the group by the questionnaire or the National Social Security Death Index Database. Mean follow-up was approximately 5 years. Patients in the control group were not contacted.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Patients were managed postoperatively in dedicated cardiac surgery ICU and step-down units according to well-defined care path protocols. All postoperative care was directed by the surgical team. Patients were followed up routinely at 1 month and subsequently as indicated by clinical circumstances.
Patient demographics and preoperative characteristics are summarized in Table 1. The extreme obese group was significantly younger and had a higher proportion of women with both noninsulin and insulin-dependent diabetes. The incidence of preoperative renal insufficiency, chronic lung disease, and sleep apnea was also significantly higher in the study group (BMI 45). The incidence of pulmonary hypertension was significantly higher in the control group (BMI 21.0 to 34.9). Most patients in both groups had normal or mildly impaired left ventricular function (mean, 50.6%; 95% CI, 49.7% to 51.6%).
|
|
|
|
Of the 102 patients with successful follow-up, 35 had died, and 67 answered the questionnaire. Table 5 summarizes the questionnaire data. This group was significantly free from subsequent coronary intervention, but 42% had symptoms of chest pain and 64% had shortness of breath. A quarter of patients underwent a subsequent surgical procedure, but none of these were bariatric procedures. In fact, subsequent weight loss after the cardiac surgical procedure was disappointing: only 24% lost more than 20 pounds in follow-up, and 12% actually gained weight. Despite increased activity compared with their preoperative state, only 27% of these patients returned to work. Overall, 63% of this sampling of patients reported a subjective improvement in their quality of life. Although they may have "felt better" after successful treatment of their cardiac disease, that improvement did not translate into significant functional improvement for most patients.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
A BMI of 35 to 40 has been the upper limit for analysis in most reported studies, with only one small study of 28 patients with a BMI exceeding 40 [10]. Multiple studies have found that morbidity was increased for the obese [11–18], whereas others show more deleterious effects in cachectic populations [12, 18]. Major complications in these obese patient studies have included sternal infection, renal failure, and respiratory failure. Preoperative risk factors in these studies included diabetes, hypertension, ejection fraction, smoking, chronic lung disease, and renal insufficiency. Our extreme obese group demonstrated similar comorbidities, including diabetes, sleep apnea, preoperative renal insufficiency, and chronic lung disease. The univariate analysis found an increased incidence of sternal and saphenous vein harvest site infections, postoperative renal failure, and the need for dialysis (Table 3).
The disease condition of diabetes is also present in a significant percentage of patients with obesity, and the relative contributions of diabetes versus obesity toward surgical mortality and morbidity have been conflicting [19, 20]. The potential additional adverse contributions of the metabolic syndrome have also been recently studied [6–8]. Diabetes was not a significant risk factor for 30-day mortality in our analysis, but BMI did reach significance as a continuous variable (Table 4). Thus, our study supports the data from reports addressing less obese populations that although diabetes in obese cardiac surgical patients increases the risk for morbidity (infection, renal failure), the impact of obesity alone appears to outweigh any influence of diabetes on 30-day mortality. It is possible that contributing factors from the metabolic syndrome in obese patients impacts on this finding. We did not collect laboratory or abdominal girth data on this population to be able to address this question in this obesity subset.
Our study also indicates that the use of the IMA for bypass grafting was significantly less in extreme obese patients in the CABG cohort. This difference in IMA use might be explained by the difference in time periods each population was drawn from. The population with a BMI of 21.0 to 34.9 consisted of patients operated on in 2004 to 2006, whereas the cohort with a BMI of 45 or more was drawn from patients operated on from 1992 to 2005. Changes in operative strategy over that time period might explain the lower use of the IMA for grafting. Another cause might be the increased technical difficulty in dissecting out the IMA in extreme obese patients, steering the surgeon to use other conduits. Because IMA use is solely at the discretion of the operating surgeon, a third reason may have been surgeon reluctance to use the IMA in this group, perhaps to decrease the potential for postoperative respiratory insufficiency in these patients.
Only 1 patient in the extreme obese group had bilateral IMA grafting, indicating reluctance to subject these patients to a potentially increased incidence of mediastinitis. However, this reduced level of IMA use (72.7%) in this younger patient population (mean age, 59.9 years) would be expected to adversely affect the long-term outcomes data for the extreme obese group [21].
Thirty-Day Mortality
We also demonstrated that 30-day mortality was significantly higher in the extreme obese group. Perioperative outcomes data from other studies on obesity in adult cardiac surgery are variable. Prabhakar and colleagues [11] have demonstrated an increase due to obesity, whereas Jin and colleagues [12] and others [10, 13–17] show no significant increase in 30-day mortality in obese patients, even in patients with BMI exceeding 40. In contrast, we observed a nearly threefold increase in mortality compared with our control group, and the finding of a marginal but significantly increased mortality with BMI in this extreme obese study population suggests that there is an increased risk in operating on these patients and that the risk/benefit and outcomes should be well-defined preoperatively.
Of importance, the most definite association with 30-day mortality was transfusion status, confirming an increasing body of evidence [22–23]. Although patients in the extreme obese group received fewer transfusions overall, transfusion was significantly associated with increased mortality in this group (Table 4). This study thus suggests an adverse association between mortality and transfusion status in this extreme obese population. Thus all these analyses, including ours, suggest that it is the increased body mass due to obesity that conveys the additional mortality and morbidity in adult cardiac surgical procedures.
Longitudinal Follow-Up of Extreme Obese Patients
The long-term survival data, although incomplete, suggest the notion that cardiac operations can be a "stepping stone" to obesity surgical procedures or aggressive weight reduction programs is unsubstantiated. This appears to be the case, not because the procedure was not successful, but because the significant lifestyle changes necessary to accomplish these goals were not undertaken after the surgical intervention.
These data emphasize that the comorbidities associated with severe morbid obesity have a strong adverse effect on quality of life in this population. Although a direct comparison was not possible, this survival curve for this age population is significantly less favorable than what would be expected from a matched group of nonobese patients of a similar age [5]. However, recent data from two large analyses of survival in obese patients document the significant increase in overall and cardiovascular mortality in these patients [5, 24]. Of note, a BMI of up to 40 was the largest included in these analyses, again significantly less than the population studied in this report.
Limitations
This observational study reporting on a subset of adult cardiac surgical patients has a number of limitations. First, the retrospective design makes it subject to selection bias.
Second, the control population consisted of a heterogeneous population of normal (BMI < 25), overweight (BMI 25.0 to 29.9), and obese (BMI > 30). In addition, a more appropriate control population (BMI 21 to 34.9) would consist of patients who had their operations over the same time span as our obese population. However, given the outlier characteristics of the extreme obese subgroup compared with these other classifications of body mass, it is unlikely that these factors would impact upon the findings of this study. Nonetheless, a case–control study design could provide stronger evidence of an association between BMI as a continuous variable and 30-day mortality. Additional case–control studies comparing normal, overweight, obese, and the extreme obese are needed to better delineate the association between 30-day mortality and BMI.
Third, more complete follow-up data would give a clearer picture of the long-term outcomes for this extreme obese population, but again the study inferences are likely to be the same. Ideally, one would like to realize that at a certain BMI, the risk of surgical intervention would far outweigh the benefits. To more accurately quantitate the additional risks of these patients, and the link between successful surgery and significant change and improvement in lifestyle, a larger observational study with propensity analysis would be appropriate.
Conclusion
Our analysis demonstrated that extreme obese patients (BMI 45) can be operated on safely but with anticipated increased perioperative morbidity and 30-day mortality. In our study, this population had higher rates of diabetes, preoperative renal insufficiency, chronic lung disease, and sleep apnea than nonextreme obese patients.
Caution should be exercised in the perioperative management of these patients, because they would be anticipated to have a significantly higher incidence of complications postoperatively. Indeed, we demonstrated significant associations between BMI, transfusion, and preoperative renal insufficiency with 30-day mortality. Perhaps alterations in perioperative management such as the increased use of off-pump surgery might improve outcomes in those patients undergoing CABG. Even more beneficial might be the loss of weight preoperatively; however, frequently, these patients are severely limited by their cardiac disease. Nevertheless, long-term follow-up data suggest that acceptable results for cardiac surgical procedures can be obtained in the extreme obese population, but most of these patients fail to realize significant functional improvement owing to the comorbidities associated with their obesity.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Number one: Did you compare your outcomes, particularly mortality outcomes, to the STS predicted risk? If you have done so, do you feel the STS model for predicted mortality is adequately weighting BMI? As the BMI of our general population changes, I wonder if that model needs to be adjusted.
Number two: I was curious to know just how confident you are with only 128 patients in your super obese group that your multivariable analysis is adequate to separate out BMI itself as a causal influence rather than as a comorbidity that is closely associated with other important risk factors like diabetes, peripheral vascular disease, renal insufficiency, poor saphenous vein quality, et cetera?
And finally, given the fact that you have demonstrated a lack of functional improvement in these patients, is it time to reevaluate the "we accept anybody for surgery" mentality? Should some of these super obese patients be getting drug-eluting stents instead of surgery?
DR TYSON: Thank you for your questions. The first one, did we compare our mortality data to the STS database, we did not do any direct comparison. The reason being, this population, again, is very unique, and even our comparison we did not feel was the best comparison we could do because of all the differences between the groups. We wanted to have a benchmark and we chose to go with patients that had been operated on at our institution under the same conditions under a similar cohort of surgeons.
As far as whether or not this model is adequate, again, as we look at the data that have been collected thus far, this super obese group of patients really kind of falls off the curve. (Slide) This is a graph off the latest New England Journal of Medicine issue from August in which they look at mortality in the obese population, just baseline, not having any kind of procedure, and as you can see, the extent, this is the one in men, goes up to and stops at 40 BMI, and in women it goes up to 45 and stops. So I think, without question, the model needs to be relooked at as far as comparing it with this population.
As far as using a multivariate analysis for only 128 patients, that was definitely a concern of ours when doing that analysis. One of the things that we noted, particularly when looking at body mass index as an independent risk factor for 30-day mortality, was how low the beta value was, and, again, this was low because it didn’t have a very large effect on 30-day mortality but was significant. And we controlled all of our logistic models on age and gender as well. But it definitely was a concern.
As far as lack of functional improvement, again, we can get these patients through their operation, but unless they have some sort of support system set up to where they can actually lose weight and treat their obesity, we are not going to see functional improvement.
DR JOHN W. HAMMON, JR (Winston-Salem, NC): I enjoyed your presentation and congratulations for presenting a group of patients that most of us would rather forget about. This is a very challenging group, obviously, obese with all the comorbid factors of insulin-resistant diabetes. We were challenged by the fact that many of our patients in this regard had very poor saphenous vein survival, and we have been trying to use mostly arterial grafts in them and wondered if that was your experience as well. We also have tried to use at least one mammary artery, and we have been trying to skeletonize the mammary to prevent sternal dehiscence and have had good luck with that. I noticed that saphenous vein harvest site infections and difficulties was one of your most powerful statistical differences between the two groups. We have been using endoscopic saphenous vein harvesting and wondered if you had used that as well. Thank you very much again.
DR TYSON: Thank you for your question. Yes, we are at this point using endoscopic vein grafts. Again, because the cohort of patients came from 1992 to 2005, definitely a good portion of these patients probably did have open harvests. The question as far as using the internal mammary artery, that was something that we noted as well, having a less use of the internal mammary in this large group and how it might affect adversely the outcomes data that we present as well.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
40. Eur J Cardiothorac Surg 2006;29:434-440.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
