JMIR Publications

ABSTRACT

Background:

Acute kidney injury critically impacts the outcomes of cardiogenic shock secondary to acute myocardial infarction (CS-AMI). Predicting and early initiation of renal replacement therapy (RRT) could improve the outcomes of patients with CS-AMI.

Objective:

This study aimed to predict RRT in this population via least absolute shrinkage and selection operator (LASSO) regression.

Methods:

This retrospective cohort study included 1431 CS-AMI patients divided into patients who did not need RRT (n= 1241) and patients who required RRT (n= 190).

Results:

Patients who required RRT were significantly older than those who did not (64.17±12.14 vs. 59.75±11.77 years, p<0.001). Patients with RRT had a greater prevalence of diabetes, peripheral arterial disease (PAD), congestive heart failure and cerebrovascular accident (CVA). Laboratory investigations revealed lower hemoglobin, creatinine clearance, and sodium bicarbonate levels and higher potassium levels in patients who required RRT. The need for RRT was associated with a significantly higher in-hospital mortality (76% vs. 39%; p<0.001). The probability of RRT in the CS. AMI patients are equal to: 1/(1 + exp(−1.92+0.82×mechanical ventilation+0.56×PAD -0.39×COPD-0.37×STEMI+0.25×mechanical circulatory support+0.22×age≥62+0.21×CABG performed-0.09×medically treated+0.19×creatinine clearance<30ml/min - 0.17× NaHCO3>19mEq+0.13x male -0.11×prior PCI+0.09× coronary arteries affected >3+ 0.08×right ventricular dysfunction+0.08×CVA+0.05×WBCs ≥13000/µL -0.03× history of CABG- 0.1×SCAI stage B- 0.05×SCAI stage C)). The cross-validated mean AUC of the model was 0.81±0.058.

Conclusions:

RRT is common after CS-AMI and is associated with high mortality. The predictive model developed via LASSO regression showed good discrimination ability. Early identification and management of high-risk patients with CS-AMI is crucial and multidisciplinary management of CS-AMI, including nephrologists, is warranted.