scholarly journals Machine learning prediction model of acute kidney injury after percutaneous coronary intervention

2022 ◽  
Vol 12 (1) ◽  
Author(s):  
Toshiki Kuno ◽  
Takahisa Mikami ◽  
Yuki Sahashi ◽  
Yohei Numasawa ◽  
Masahiro Suzuki ◽  
...  

AbstractAcute kidney injury (AKI) after percutaneous coronary intervention (PCI) is associated with a significant risk of morbidity and mortality. The traditional risk model provided by the National Cardiovascular Data Registry (NCDR) is useful for predicting the preprocedural risk of AKI, although the scoring system requires a number of clinical contents. We sought to examine whether machine learning (ML) techniques could predict AKI with fewer NCDR-AKI risk model variables within a comparable PCI database in Japan. We evaluated 19,222 consecutive patients undergoing PCI between 2008 and 2019 in a Japanese multicenter registry. AKI was defined as an absolute or a relative increase in serum creatinine of 0.3 mg/dL or 50%. The data were split into training (N = 16,644; 2008–2017) and testing datasets (N = 2578; 2017–2019). The area under the curve (AUC) was calculated using the light gradient boosting model (GBM) with selected variables by Lasso and SHapley Additive exPlanations (SHAP) methods among 12 traditional variables, excluding the use of an intra-aortic balloon pump, since its use was considered operator-dependent. The incidence of AKI was 9.4% in the cohort. Lasso and SHAP methods demonstrated that seven variables (age, eGFR, preprocedural hemoglobin, ST-elevation myocardial infarction, non-ST-elevation myocardial infarction/unstable angina, heart failure symptoms, and cardiogenic shock) were pertinent. AUC calculated by the light GBM with seven variables had a performance similar to that of the conventional logistic regression prediction model that included 12 variables (light GBM, AUC [training/testing datasets]: 0.779/0.772; logistic regression, AUC [training/testing datasets]: 0.797/0.755). The AKI risk model after PCI using ML enabled adequate risk quantification with fewer variables. ML techniques may aid in enhancing the international use of validated risk models.

2019 ◽  
Vol 10 (2) ◽  
pp. 108-115
Author(s):  
David Zahler ◽  
Keren-Lee Rozenfeld ◽  
Ilan Merdler ◽  
Yogev Peri ◽  
Yacov Shacham

Introduction: The ratio of contrast media volume to glomerular filtration rate (contrast/GFR) has been shown to correlate with the occurrence of contrast-induced acute kidney injury (CI-AKI) in unselected patient populations who underwent percutaneous coronary intervention (PCI). Objective: We evaluated the possible utilization of this marker and optimal cutoff among ST-elevation myocardial infarction (STEMI) patients undergoing primary PCI. Methods: We retrospectively included 419 patients with STEMI treated with primary PCI. The occurrence of CI-AKI was defined by the KDIGO criteria as an increase in serum creatinine of ≥0.3 mg/dL within 48 h following PCI. A receiver-operator characteristic (ROC) curve was used to identify the optimal cutoff value of contrast/GFR ratio to predict CI-AKI. This value was then assessed using multivariable logistic regression. Results: The overall incidence of CI-AKI was 9%. The contrast/GFR ratio was significantly higher among patients with CI-AKI (2.7 ± 1.2 vs. 1.9 ± 0.9; p < 0.001). According to the ROC curve analysis, the optimal cutoff value of contrast/GFR ratio to predict AKI was measured as ≥2.13, with 70% sensitivity and 60% specificity (AUC 0.65, 95% CI 0.56–0.74; p = 0.002). In a multivariate logistic regression model, contrast/GFR ratio ≥2.13 was independently associated with CI-AKI (OR 2.46, 95% CI 1.09–5.57; p = 0.03). Conclusions: Among STEMI patients undergoing primary PCI, contrast/GFR ratio ≥2.13 was independently associated with CI-AKI.


Heart ◽  
2017 ◽  
Vol 104 (9) ◽  
pp. 767-772 ◽  
Author(s):  
Johanne Silvain ◽  
Lee S Nguyen ◽  
Vincent Spagnoli ◽  
Mathieu Kerneis ◽  
Paul Guedeney ◽  
...  

ObjectivesContrast-induced acute kidney injury (CI-AKI) is a common and potentially severe complication in patients with ST elevation myocardial infarction (STEMI) treated with primary percutaneous coronary intervention (pPCI). There is no consensus on the best definition of CI-AKI to identify patients at risk of haemodialysis or death. The objective of this study was to assess the association of CI-AKI, using four definitions, on inhospital mortality, mortality or haemodialysis requirement over 1-year follow-up, in patients with STEMI treated with pPCI.MethodsIn this prospective, observational study, all patients with STEMI referred for pPCI were included. We identified independent variables associated with CI-AKI and mortality.ResultsWe included 1114 consecutive patients with STEMI treated by pPCI. CI-AKI occurred in 18.3%, 12.2%, 15.6% and 10.5% of patients according to the CIN, Acute Kidney Injury Network (AKIN), Risk, Injury, Failure, Loss of kidney function, and End-stage kidney disease (RIFLE) Modification of Diet in Renal Disease (MDRD) and RIFLE Chronic Kidney Disease - Epidemiology Collaboration (CKD-EPI) definitions, respectively. The RIFLE (CKD-EPI) definition was the most discriminant definition to identify patients at higher risk of inhospital mortality (27.1% vs 4.0%; adjusted OR 2.7 (95% CI 1.4 to 5.1), p=0.003), 1-year mortality (27.4% vs 6.6%; adjusted OR 2.8 (95% CI 1.5 to 5.3), p=0.002) and haemodialysis requirement at 1-year follow-up (15.6% vs 2.7%; adjusted OR 6.7 (95% CI 3.3 to 13.6), p=0.001). Haemodynamic instability, cardiac arrest, preexisting renal failure, elderly age and a high contrast media volume were independently associated with 1-year mortality. Of interest, contrast-media volume was not correlated to increase of creatininaemia (r=0.06) or decrease in estimated glomerular filtration rate (r=0.05) after percutaneous coronary intervention in our population.ConclusionsCI-AKI is a frequent and serious complication of STEMI treated by pPCI. The RIFLE definition is the most accurate definition to identify patients with CI-AKI at high risk of mortality or haemodialysis.


Sign in / Sign up

Export Citation Format

Share Document