Summary
Background
The increasing number of obese patients eligible for cardiac surgery requires risks and benefits to be balanced in this population.
Aims
To study the results of cardiac surgery in severely obese patients (body mass index [BMI] ≥ 35 kg/m 2 ).
Methods
In this retrospective study of 3564 patients undergoing elective cardiac surgery between 2004 and 2012, the population was divided into two groups: BMI 20–34.9 kg/m 2 ( n = 3282) and BMI ≥ 35 kg/m 2 ( n = 282). Patients with BMI < 20 kg/m 2 were excluded due to the well-known increased mortality risk. The primary endpoint was 90-day mortality. A multivariable analysis was performed to identify prognostic factors.
Results
Among our patients, 58.2% and 27.7% underwent isolated coronary or valvular surgery, respectively; 9.7% had combined valvular and coronary surgery and 4.4% had other procedures. Severely obese patients were younger: 62.5 ± 9.3 years vs 67.8 ± 10.7 years ( P = 0.0001). Overall 90-day mortality was 4.0%. Severe obesity did not influence postoperative mortality. In the multivariable analysis, the interaction between preoperative renal failure and severe obesity was an important mortality prognostic factor (hazard ratio: 11.17; P = 0.03). Mediastinitis rates were similar between groups in non-diabetic patients; in diabetic patients, severe obesity was associated with higher mediastinitis rates ( P = 0.002). Superficial wound infections were higher in severely obese patients ( P = 0.003).
Conclusion
Elective cardiac surgery in severely obese patients was not associated with increased perioperative morbimortality, but had a higher superficial wound infection risk. Nevertheless, severe obesity itself should not be a contraindication to elective surgery.
Résumé
Contexte
L’augmentation des patients obèses éligibles à la chirurgie cardiaque nécessite de mieux étudier la balance bénéfices–risques.
Objectif
Étudier les résultats de la chirurgie programmée chez les patients obèses sévères (indice de masse corporelle [IMC] ≥ 35 kg/m 2 ).
Méthodes
Il s’agit d’une étude rétrospective de 3564 patients opérés entre 2004 et 2012. La population a été divisée en deux groupes : IMC : 20–34,9 kg/m 2 ( n = 3282) et IMC ≥ 35 kg/m 2 ( n = 282). Les patients avec un IMC < 20 kg/m 2 ont été exclus en raison d’un surrisque bien établi dans la littérature. Le critère de jugement était la mortalité à 90 jours.
Résultats
Respectivement, 58,2 % et 27,7 % des patients ont eu une chirurgie coronarienne ou valvulaire ; 9,7 % une chirurgie combinée et 4,4 % d’autres procédures. Les patients obèses sévères étaient plus jeunes : 62,5 ± 9,3 ans vs 67,8 ± 10,7 ans ( p = 0,0001). La mortalité à 90 jours était de 4,0 % et non influencée par l’obésité sévère. L’association insuffisance rénale préopératoire et obésité sévère était de mauvais pronostic (HR : 11,169 ; p = 0,03). Les taux de médiastinite étaient comparables entre les groupes, chez les non-diabétiques, alors que chez les diabétiques l’obésité sévère était associée à plus de médiastinites ( p = 0,002). Les infections des cicatrices étaient plus élevées chez les patients obèses sévères ( p = 0,003).
Conclusions
La chirurgie cardiaque programmée chez des patients obèses sévères n’accroît pas la morbi-mortalité périopératoire. Elle présente un risque plus élevé d’infection de la cicatrice. Néanmoins, l’obésité sévère en soi ne devrait pas être une contre-indication à cette chirurgie.
Background
Recent data from the World Health Organization showed that worldwide obesity has nearly doubled since 1980 . In 2008, 35% of adults (> 1.4 billion) aged ≥ 20 years were overweight and 11% (> 200 million men and nearly 300 million women) were obese. Being overweight/obese is nowadays the fifth leading risk of global death and at least 2.8 million adults die each year as a result of being overweight or obese.
The proportion of obese patients eligible for cardiac surgery is following the same upwards trend, which raises the issue of the perioperative risk of this population. Low body mass index (BMI < 20 kg/m 2 ) is an independent factor that affects morbidity and mortality negatively after cardiac surgery . On the other hand, the relationship between obesity and operative outcomes is unclear. Some authors have reported that obesity negatively affects operative mortality in patients undergoing valvular surgery , while others have suggested the existence of an ‘obesity paradox’, with such patients having a better survival rate than normal-weight patients .
The consequence of severe obesity (BMI ≥ 35 kg/m 2 ) in perioperative care after cardiac surgery is an interesting issue. The EuroSCORE risk calculation in this patient group might be incomplete in predicting operative mortality, as it does not take into account the weight of the patients. The present study aims to evaluate perioperative morbidity and mortality in severely obese patients undergoing elective cardiac surgery.
Methods
Population
This was a retrospective study carried out using a prospective database of 3564 patients undergoing elective cardiac surgery between January 2004 and December 2012. All patients gave written consent to inclusion of their medical information in our institutional database and to the use of this information for research purposes. The study was approved by the Institutional Review Board of the French Society of Thoracic and Cardio-Vascular Surgery (CERC-SFCTCV-2013-8-2-22-12-38-HyIl).
Patients were divided into two groups according to their BMI: group I ( n = 3282), 20 kg/m 2 ≤ BMI ≤ 34.9 kg/m 2 ; group II ( n = 282), BMI ≥ 35 kg/m 2 . Patients with a BMI < 20 kg/m 2 were deliberately excluded due to their established increased mortality risk, as reported unambiguously in the literature. Overweight (25 kg/m 2 ≤ BMI ≤ 29.9 kg/m 2 ) and obese (30 kg/m 2 ≤ BMI ≤ 34.9 kg/m 2 ) patients were pooled with normal-weight patients (20 kg/m 2 ≤ BMI ≤ 24.9 kg/m 2 ), and severely obese patients (35 kg/m 2 ≤ BMI ≤ 39.9 kg/m 2 ) were pooled with morbidly obese patients (BMI ≥ 40 kg/m 2 ).
Preoperative renal function was estimated by glomerular filtration rate (GFR), calculated using the Cockcroft-Gault formula. A patient was considered to have preoperative chronic renal failure when the GFR was < 60 mL/min, and GFR was used as a dichotomous variable in further statistical analyses.
Operative management
During the 8-year study period, patients were operated on by three senior surgeons. All patients were monitored routinely in the operating room during cardiac surgery. Cannulation techniques and cardiopulmonary bypass (CPB) were conducted as usual. Myocardial protection was achieved by antegrade cold blood cardioplegia repeated every 30 minutes. In coronary patients, internal thoracic arteries were harvested in a skeletonized fashion by all surgeons. Sternal closure was done in a similar fashion in the two groups, mostly with seven simple wires. After the surgical procedure, patients were monitored in an intensive care unit (ICU) for at least 24 hours. Strict control of glycaemia (< 1.6 g/L) was maintained during the first two postoperative days, by means of an intravenous insulin perfusion. Thereafter, preoperative diabetic treatment (subcutaneous insulin and/or oral antidiabetics) was progressively reintroduced. All patients underwent a transthoracic echocardiogram before discharge, to confirm the absence of pericardial effusion.
Postoperative outcomes
Operative mortality included all patients who died within 90 days after surgery. Wound infection was defined as a bacteriologically positive collection in the presternal surgical site without sternal involvement or fracture. Mediastinitis was defined as the presence of a bacteriologically positive collection in the anterior mediastinum, behind the sternum, or bacteriologically positive samples from the sternum. All patients had a medical consultation with their surgeon 6 months after the operation. In our database, this marked the end of follow-up, giving us information only about postoperative and short-term outcomes.
Statistical analysis
Analyses were conducted using SAS software (SAS version 9.2; SAS Institute Inc., Cary, NC, USA). Continuous variables are expressed as means ± standard deviations or medians (interquartile ranges), as appropriate. Categorical variables are presented as absolute numbers and percentages. The comparison between severely obese patients (BMI ≥ 35 kg/m 2 ) and patients with a BMI between 20 kg/m 2 and 35 kg/m 2 was made using Student’s t test for means of continuous variables, and the Chi 2 test or Fisher’s exact test, as appropriate, for proportions of categorical variables.
Event-free survival curves were estimated using the Kaplan–Meier method and compared using the log-rank test. Median follow-up time was estimated with the reverse Kaplan–Meier method. Univariate Cox analyses were performed to identify independent predictors of an event (mediastinitis, wound infection and death were considered one at a time, mediastinitis and wound infection being censored at the time of death). As the main aim of this article was to study the predictive value of severe obesity, interactions between severe obesity and the other covariates were systematically tested, by comparing models with interactions with models without interactions using likelihood ratio tests. Owing to a significant interaction between severe obesity and diabetes for the hazard of mediastinitis, wound infection and death, analyses were done separately for non-diabetic and diabetic patients. The log-linearity assumption for continuous variables and the proportional hazard assumption were tested by Kolmogorov-type supremum tests as implemented in PROC PHREG in SAS software (SAS version 9.2; SAS Institute Inc., Cary, NC, USA) and inspected visually with residual plots. In case of violation of the former assumption, the continuous variable was dichotomized, the cut-off value being established visually; in case of violation of the latter assumption, a piecewise model was used to model the hazard ratio (HR) as a step function of time. Multivariable Cox models were built using best subset selection and were selected using Schwarz’s Bayesian criterion, Harrell’s c-statistic as a measure of calibration and Kent and O’Quigley’s ρ 2 as a measure of discrimination. A two-tailed type I error rate < 0.05 was considered for statistical significance.
Results
Preoperative and perioperative characteristics
During the study period, 3849 patients (mean BMI 27.1 ± 5.2 kg/m 2 ) underwent elective cardiac surgery in our department. Of these patients, 3564 had a BMI ≥ 20 kg/m 2 and were included for further analysis. Preoperative patient data are summarized in Table 1 . The whole cohort of patients had a mean age of 67.3 ± 10.6 years and most were men (71.9%). Overall, 58.2% and 27.7% of the patients had isolated coronary or valvular surgery, respectively; 9.7% had combined valvular and coronary surgery and 4.4% had other procedures. Table 2 summarizes the perioperative data in the different groups. There was no difference in the number of harvested internal thoracic arteries between the groups ( P = 0.08; Table 2 ). Similarly, this number did not differ between diabetic and non-diabetic patients in the severely obese group ( P = 0.10).
| Variables | Sample, ( n = 3564) | BMI 20–35 kg/m 2 , ( n = 3282) | BMI ≥ 35 kg/m 2 , ( n = 282) | P |
|---|---|---|---|---|
| Age (years) | 67.3 ± 10.6 | 67.8 ± 10.7 | 62.5 ± 9.3 | 0.0001 |
| Men | 2542 (71.9) | 2289 (70.3) | 253 (89.7) | 0.0001 |
| LVEF (%) | 56.4 ± 13.3 | 56.5 ± 13.4 | 54.2 ± 11.7 | 0.004 |
| Dyspnoea, NYHA class | 0.03 | |||
| I | 948 (29.1) | 892 (29.7) | 56 (21.6) | |
| II | 1421 (43.6) | 1292 (43.1) | 129 (49.8) | |
| III | 777 (23.9) | 710 (23.7) | 67 (25.9) | |
| IV | 112 (3.4) | 105 (3.5) | 7 (2.7) | |
| Diabetes mellitus | 1229 (34.7) | 1080 (33.1) | 149 (53.4) | 0.0001 |
| COPD | 539 (15.1) | 478 (14.6) | 61 (21.6) | 0.002 |
| Preoperative stroke | 268 (7.7) | 244 (7.6) | 24 (8.9) | 0.44 |
| Hypertension | 2518 (71.7) | 2294 (71.0) | 224 (80.0) | 0.001 |
| Preoperative MI | 820 (25.3) | 741 (24.8) | 79 (30.9) | 0.03 |
| Warfarin take | 289 (11.8) | 261 (11.7) | 28 (13.6) | 0.41 |
| Clopidogrel take | 622 (25.4) | 562 (25.1) | 60 (29.0) | 0.22 |
| Preoperative renal failure | 466 (13.4) | 426 (13.3) | 40 (14.8) | 0.48 |
| Preoperative dialysis | 19 (0.6) | 18 (0.6) | 1 (0.4) | 1.00 |
| PAD | 742 (21.6) | 673 (21.3) | 69 (25.5) | 0.11 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
