Introduction: Renal infarction (RI) is rare but clinically important because the appropriate treatment depends on the time of diagnosis. RI is often misdiagnosed as acute pyelonephritis (APN) because both diseases have nonspecific symptoms such as flank pain and abdominal pain. We identified predictors for distinguishing RI from APN.
Methods: The data of patients visited the emergency department and diagnosed with RI or APN from March 2016 to May 2020 were prospectively collected and retrospectively analyzed. Patients aged under 18 years, with a history of trauma, or incomplete medical records were excluded. Using a matching ratio of 1:5, RI patients were randomly matched to APN patients. Multivariable logistic regression analysis was performed to identify factors that could distinguish RI from APN. In addition, we constructed a decision tree to identify patterns of risk factors and develop prediction algorithms.
Results: The RI and APN groups included 55 and 275 patients, respectively. Multivariable logistic regression analysis showed that male sex (OR, 6.161; p = 0.009), atrial fibrillation (AF) (OR, 14.303; p = 0.021), costovertebral angle tenderness (CVAT) (OR, 0.106; p < 0.001), aspartate transaminase (AST) level > 21.50 U/L (OR, 19.820; p < 0.001), C-reactive protein (CRP) level < 19.75 mg/L (OR, 10.167; p < 0.001), and pyuria (OR, 0.037; p < 0.001) were significantly associated with RI distinguishing from APN.
Conclusion: Male sex, AF, no CVAT, AST level > 21.50 U/L, CRP level < 19.75 mg/L, and no pyuria were significant factors that could distinguish RI from APN.
Keywords: Alanine transaminase; C-reactive protein; Pyelonephritis; Pyuria; Renal artery obstruction.
Copyright © 2023 Elsevier Inc. All rights reserved.