Development of a Nomogram for Predicting Hyperkalemia in Advanced CKD

Le résumé fourni par la source

Background: With advanced chronic kidney disease (CKD) develop hyperkalemia (HyperK) is frequent. Predicting HyperK is important if we are to prevent it, identify it early, and treat it. We developed a nomogram to predict HyperK in advanced CKD. Methods: In a retrospective cohort study we included ambulatory persons with advanced CKD without HyperK in order to differentiate between clinical presentations and the association with HyperK, death, and kidney replacement therapy (KRT). We analyzed risk using logistic regression with lasso and a multivariate Cox proportional hazards model and created a simple nomogram to predict HyperK according to sex. Results: Between 2020 and 2024, 395 patients with advanced CKD. Women (49.4%), mean age was 65 years. The main comorbidities were diabetes (57%) and hypertension (74.9%). The mean GFR 25 ml. 80% were taking RAASi and 60% diuretics. During the 2.2-year follow-up 76% developed HyperK. The risk of HyperK fell by 49% in men (OR, 0.51; CI, 0.28-0.92) and 58% in patients taking thiazides (OR, 0.42; CI, 0.21-0.86). Given that the risk of HyperK differed according to sex, we ran separate models for women and men. In women, the risk of HyperK fell by 60% in the case of obesity (OR, 0.40; CI, 0.19-0.87) and 61% in line with sodium values (OR, 0.39; CI, 0.17-0.91). The area under the receiver operating characteristic curve was 0.703 for women and 0.731 for men. We constructed a nomogram that predicted HyperK in both sexes with a high degree of accuracy. 29 patients (7%) died, KRT was initiated in 73 patients (18.4%). Conclusion: We developed a nomogram based on simple clinical characteristics to evaluate the risk of HyperK in advanced CKD. Funding: Other NIH Support - NONE

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.