Machine learning algorithms for predicting mortality after coronary artery bypass grafting

Khalaji, Amirmohammad; Behnoush, Amir Hossein; Jameie, Mana; Sharifi, Ali; Sheikhy, Ali; Fallahzadeh, Aida; Sadeghian, Saeed; Pashang, Mina; Bagheri, Jamshid; Tafti, Seyed Hossein Ahmadi; Hosseini, Kaveh

doi:10.3389/fcvm.2022.977747

Cited by 14 publications

(4 citation statements)

References 36 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, echocardiographic left ventricular ejection fraction (LVEF), estimated glomerular filtration rate (eGFR), HDL-C (mg/dL) and low-density lipoprotein cholesterol (LDL-C) (mg/dL) levels were measured preoperatively. Variable definitions were in line with previous studies on this population [19].…”

Section: Baseline Assessmentmentioning

confidence: 74%

BMI modifies HDL-C effects on coronary artery bypass grafting outcomes

et al. 2022

Self Cite

View full text Add to dashboard Cite

Background Despite the recognized implications of high-density lipoprotein cholesterol (HDL-C) in cardiovascular diseases, the role of body mass index (BMI) in HDL-C association with cardiovascular outcomes remains unclear. This study investigated the possible modifying implications of BMI on the correlation between HDL-C and coronary artery bypass grafting (CABG) outcomes. Methods The present cohort included isolated CABG patients (median follow-up: 76.58 [75.79–77.38] months). The participants were classified into three groups: 18.5 ≤ BMI < 25 (normal), 25 ≤ BMI < 30 (overweight), and 30 ≤ BMI < 35 (obese) kg/m2. Cox proportional hazard models (CPHs) and restricted cubic splines (RCSs) were applied to evaluate the relationship between HDL-C and all-cause mortality as well as major adverse cardio-cerebrovascular events (MACCEs) in different BMI categories. Results This study enrolled a total of 15,639 patients. Considering the final Cox analysis among the normal and overweight groups, HDL-C ≥ 60 was a significant protective factor compared to 40 < HDL-C < 60 for all-cause mortality (adjusted hazard ratio (aHR): 0.47, P: 0.027; and aHR: 0.64, P: 0.007, respectively). However, the protective effect of HDL-C ≥ 60 was no longer observed among patients with 30 ≤ BMI < 35 (aHR: 1.16, P = 0.668). RCS trend analyses recapitulated these findings; among 30 ≤ BMI < 35, no uniform inverse linear association was observed; after approximately HDL-C≈55, its increase was no longer associated with reduced mortality risk. RCS analyses on MACCE revealed a plateau effect followed by a modest rise in overweight and obese patients from HDL-C = 40 onward (nonlinear association). Conclusions Very high HDL-C (≥ 60 mg/dL) was not related to better outcomes among obese CABG patients. Furthermore, HDL-C was related to the post-CABG outcomes in a nonlinear manner, and the magnitude of its effects also differed across BMI subgroups.

show abstract

Section: Baseline Assessmentmentioning

confidence: 74%

BMI modifies HDL-C effects on coronary artery bypass grafting outcomes

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…There is a need for regional models designed for specific populations based on their local demographic features. There have been studies conducted on this topic in CABG patients in Iran 72,73 ; however, this is the first study focusing on hypertensive patients. Moreover, we designed these models with a combination of demographic, clinical, and laboratory characteristics which are widely available in clinical settings, while the use of a combination of variables has been shown to be beneficial in the prediction 74,75 …”

Section: Discussionmentioning

confidence: 99%

Machine learning‐based prediction of 1‐year mortality in hypertensive patients undergoing coronary revascularization surgery

Behnoush

Khalaji

Rezaee

et al. 2023

Clinical Cardiology

Self Cite

View full text Add to dashboard Cite

Background: Machine learning (ML) has shown promising results in all fields of medicine, including preventive cardiology. Hypertensive patients are at higher risk of mortality after coronary artery bypass graft (CABG) surgery; thus, we aimed to design and evaluate five ML models to predict 1-year mortality among hypertensive patients who underwent CABG.Hyothesis: ML algorithms can significantly improve mortality prediction after CABG.Methods: Tehran Heart Center's CABG data registry was used to extract several baseline and peri-procedural characteristics and mortality data. The best features were chosen using random forest (RF) feature selection algorithm. Five ML models were developed to predict 1-year mortality: logistic regression (LR), RF, artificial neural network (ANN), extreme gradient boosting (XGB), and naïve Bayes (NB). The area under the curve (AUC), sensitivity, and specificity were used to evaluate the models.Results: Among the 8,493 hypertensive patients who underwent CABG (mean age of 68.27 ± 9.27 years), 303 died in the first year. Eleven features were selected as the best predictors, among which total ventilation hours and ejection fraction were the leading ones. LR showed the best prediction ability with an AUC of 0.82, while the least AUC was for the NB model (0.79). Among the subgroups, the highest AUC for LR model was for two age range groups (50-59 and 80-89 years), overweight, diabetic, and smoker subgroups of hypertensive patients.Conclusions: All ML models had excellent performance in predicting 1-year mortality among CABG hypertension patients, while LR was the best regarding AUC. These

show abstract

“…Although anemia was thoroughly discussed in the literature as an important risk factor for in-hospital mortality after CABG, [13][14][15][21][22][23][24][25][26][27][28][29][30][31] the difference in normal Hb levels was not evaluated before.…”

Section: Discussionmentioning

confidence: 99%

“…In addition, patients with lower limits of normal Hb levels were at greater risk compared with mid‐normal values. Although anemia was thoroughly discussed in the literature as an important risk factor for in‐hospital mortality after CABG, 13 , 14 , 15 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 the difference in normal Hb levels was not evaluated before.…”

Section: Discussionmentioning

confidence: 99%

Mortality across the spectrum of hemoglobin level in patients undergoing surgical coronary revascularization

Khalaji

Ajam

Sheikhy

et al. 2023

Clinical Cardiology

Self Cite

View full text Add to dashboard Cite

Background Preoperative hemoglobin (Hb) level is a predictor of in‐hospital and midterm mortality in patients undergoing coronary artery bypass grafting surgery (CABG). However, the debate about the different hazards across Hb levels and sex differences in outcome occurrence is still on the table. Methods This is a registry‐based serial cross‐sectional study at Tehran Heart Center. Nonanemic patients who underwent CABG with complete follow‐up data were included. The Restricted Cubic Splines (RCS) in the Cox model was used to calculate the sex‐specific correlation between in‐hospital, 6‐month, and 1‐year mortalities and normal Hb levels using odds ratio for the in‐hospital and hazard ratios for 6‐month and 1‐year mortality, adjusted for all possible confounders. Results From 2005 to 2016, a total of 13,557 patients were included, of which 134 had in‐hospital mortality as our primary outcome. Preoperative Hb levels were significantly lower in the deceased. Moreover, dead patients had significantly higher rates of diabetes and hypertension, while lower ejection fraction. Cut‐offs for reference Hb level were higher for males compared with females. The correlation between Hb level and in‐hospital mortality was nearly U‐shaped. Quantitatively, Hb of ≥15.62 and ≤13.25 g/dL for men and that of ≥14.92 and ≤13.4 g/dL for women tended to be associated with higher in‐hospital mortality. Conclusions The association between preoperative Hb level and in‐hospital mortality differs in men and women and does not follow a linear correlation among nonanemic patients. Both low and high numbers in the Hb level spectrum are at greater risk.

show abstract

Machine learning algorithms for predicting mortality after coronary artery bypass grafting

Cited by 14 publications

References 36 publications

BMI modifies HDL-C effects on coronary artery bypass grafting outcomes

BMI modifies HDL-C effects on coronary artery bypass grafting outcomes

Machine learning‐based prediction of 1‐year mortality in hypertensive patients undergoing coronary revascularization surgery

Mortality across the spectrum of hemoglobin level in patients undergoing surgical coronary revascularization

Contact Info

Product

Resources

About