Polypharmacy and the Progression of Chronic Kidney Disease: Korean Cohort Study for Outcome in Patients with Chronic Kidney Disease

• Main Authors: Hyang Ki Min, Su Ah Sung, Wookyung Chung, Yeong Hoon Kim, Dong-Wan Chae, Curie Ahn, Kook-Hwan Oh, Sue K. Park, Sung Woo Lee
• Format: Article
• Language: English
• Published: Karger Publishers 2021-06-01
• Series: Kidney & Blood Pressure Research
• Subjects: chronic kidney disease; polypharmacy; progression; renal hazard; serial medication count
• Online Access: https://www.karger.com/Article/FullText/516029

Introduction: The renal hazard of polypharmacy has never been evaluated in predialysis chronic kidney disease (CKD) patients. Objective: We aimed to analyze the renal hazard of polypharmacy in predialysis CKD patients with stage 1–5. Method: The data of 2,238 patients from a large-scale multicenter prospective Korean study (2011–2016), excluding 325 patients with various missing data, were reviewed. Polypharmacy was defined as taking 6 or more medications at the time of enrollment; renal events were defined as a ≥50% decrease in kidney function from baseline values, doubling of the serum creatinine levels, or initiation of renal replacement treatment. Hazard ratio (HR) and 95% confidence interval (CI) were calculated using Cox proportional-hazard regression analysis. Results: Of the 1,913 patients, the mean estimated glomerular filtration rate was 53.6 mL/min/1.73 m2. The mean medication count was 4.1, and the prevalence of polypharmacy was 27.1%. During the average period of 3.6 years, 520 patients developed renal events (27.2%). Although increased medication counts were associated with increased renal hazard with HR (95% CI) of 1.056 (1.007–1.107, p = 0.025), even after adjusting for various confounders, adding comorbidity score and kidney function nullified the statistical significance. In mediation analysis, 55.6% (p = 0.016) of renal hazard in increased medication counts was mediated by the kidney function, and there was no direct effect of medication counts on renal event development. In subgroup analysis, the renal hazard of the medication counts was evident only in stage 1–3 of CKD patients (p for interaction = 0.014). Conclusions: We cannot identify the direct renal hazard of multiple medications, and most of the potential renal hazard was derived from intimate relationship with disease burden and kidney function.