Supplementary Material for: Disparities in Chronic Kidney Disease Progression by Medicare Advantage Enrollees

<b><i>Introduction:</i></b> The prevalence of chronic kidney disease (CKD) in Medicare beneficiaries has quadrupled in the past 2 decades, but little is known about risk factors affecting the progression of CKD. This study aims to understand the progression in Medicare Advantage enrollees and whether it differs by provider recognition of CKD, race and ethnicity, or geographic location. In a large cohort of Medicare Advantage (MA) enrollees, we examined whether CKD progression, up to 5 years after study entry, differed by demographic and clinical factors and identified additional risk factors of CKD progression. <b><i>Methods:</i></b> In a cohort of 1,002,388 MA enrollees with CKD stages 1–4 based on 2013–2018 labs, progression was estimated using a mixed-effects model that adjusted for demographics, geographic location, comorbidity, urine albumin-to-creatinine ratio, clinical recognition via diagnosed CKD, and time-fixed effects. Race and ethnicity, geographic location, and clinical recognition of CKD were interacted with time in 3 separate regression models. <b><i>Results:</i></b> Mean (median) follow-up was 3.1 (3.0) years. Black and Hispanic MA enrollees had greater kidney function at study entry than other beneficiaries, but their kidney function declined faster. MA enrollees with clinically recognized CKD had estimated glomerular filtration rate levels that were 18.6 units (95% confidence interval [CI]: 18.5–18.7) lower than levels of unrecognized patients, but kidney function declined more slowly in enrollees with clinical recognition. There were no differences in CKD progression by geography. After removal of the race coefficient from the eGFR equation in a sensitivity analysis, kidney function was much lower in all years among Black MA enrollees, but patterns of progression remained the same. <b><i>Discussion/Conclusions:</i></b> These results suggest that patients with clinically recognized CKD and racial and ethnic minorities merit closer surveillance and management to reduce their risk of faster progression.

**引言：** 过去二十年间，联邦医疗保险（Medicare）参保人群中慢性肾脏病（chronic kidney disease, CKD）的患病率翻了四倍，但目前针对影响CKD进展的危险因素仍知之甚少。本研究旨在探究联邦医疗保险优势计划（Medicare Advantage, MA）参保者的CKD进展情况，以及该进展是否因医疗机构对CKD的识别情况、种族族裔或地理位置而异。本研究纳入大型MA参保者队列，分析研究入组后最长5年内的CKD进展是否因人口统计学与临床特征存在差异，并明确CKD进展的其他危险因素。
**方法：** 本队列纳入1002388名基于2013至2018年实验室检测结果确诊为1~4期CKD的MA参保者。采用混合效应模型评估CKD进展情况，模型校正了人口统计学特征、地理位置、合并症、尿白蛋白肌酐比值、经确诊CKD体现的临床识别情况以及时间固定效应。在3个独立回归模型中，分别将种族族裔、地理位置以及CKD临床识别情况与时间设置为交互项。
**结果：** 本研究的平均（中位）随访时长为3.1（3.0）年。黑人与西班牙裔MA参保者在研究入组时的肾功能水平高于其他参保者，但他们的肾功能下降速度更快。经临床识别的CKD患者的估算肾小球滤过率（estimated glomerular filtration rate, eGFR）水平较未被识别的患者低18.6个单位（95%置信区间[confidence interval, CI]：18.5~18.7），但这类经临床识别的参保者的肾功能下降速度更缓慢。不同地理位置间的CKD进展无显著差异。在敏感性分析中移除eGFR方程中的种族系数后，所有随访年份中的黑人MA参保者的肾功能水平均显著更低，但CKD进展模式未发生改变。
**讨论与结论：** 上述结果提示，经临床识别的CKD患者以及少数种族族裔群体需要接受更密切的监测与管理，以降低其CKD快速进展的风险。