Risk Factors and Machine Learning Model for Beta-Lactam Drugs Related Acute Kidney Injury

Analysis of Risk Factors of Beta-Lactam Drugs Related Acute Kidney Injury in Hospitalized Patients and Developments of Machine Learning Model

Acute kidney injury (AKI), also known as acute kidney failure (ARF), is a common and complex kidney disease in clinic and an important factor related to poor prognosis of patients in clinic. In the present study, a single-center retrospective study was conducted in our center. The clinical data of hospitalized patients received β-Lactam drugs from January 2018 to December 2020 was retrospectively analyzed. The multiple logistic regression analysis suggested that complicated with hypertension, anemia, pneumonia, shock, sepsis, heart failure, combined use of proton pump inhibitors (PPI), angiotensin-converting enzyme inhibitor (ACEI), angiotensin Ⅱ receptor antagonist (ARB) were independent risk factors for AKI related to β-Lactam drugs. In clinical practice, patients with acute kidney injury risk factors should be closely monitored for changes in their blood creatinine and urine output to avoid acute kidney injury. For patients who have suffered from acute kidney injury, the cause should be removed in time and corresponding symptomatic treatment should be given.

Inclusion Criteria:

• All inpatients who used β-lactam drugs during hospitalization
• Hospital stay ≥ 48h
• Age ≥18 years
• There are two or more blood creatinine tests during hospitalization

Exclusion Criteria:

• Hospital stay < 48h
• Age <18 years
• Glomerular filtration rate (GFR)< 30ml/min/1.73m2 within 48 hours after admission
• AKI was diagnosed on admission
• Less than two Scr test results during hospitalization
• The Scr values were always lower than 40 μmol/L during hospitalization
• Cases with incomplete medical history information