Validation of the kidney failure risk equation in predicting the risk of progression to kidney failure in a multi-ethnic Singapore chronic kidney disease cohort

Risk Prediction and Management of Chronic Kidney Disease in People Living with Type 2 Diabetes Mellitus

People with type 2 diabetes mellitus have increased risk of chronic kidney disease and atherosclerotic cardiovascular disease. Improved care delivery and implementation of guideline-directed medical therapy have contributed to the declining incidence of atherosclerotic cardiovascular disease in high-income countries. By contrast, the global incidence of chronic kidney disease and associated mortality is either plateaued or increased, leading to escalating direct and indirect medical costs. Given limited resources, better risk stratification approaches to identify people at risk of rapid progression to end-stage kidney disease can reduce therapeutic inertia, facilitate timely interventions and identify the need for early nephrologist referral. Among people with chronic kidney disease G3a and beyond, the kidney failure risk equations (KFRE) have been externally validated and outperformed other risk prediction models. The KFRE can also guide the timing of preparation for kidney replacement therapy with improved healthcare resources planning and may prevent multiple complications and premature mortality among people with chronic kidney disease with and without type 2 diabetes mellitus. The present review summarizes the evidence of KFRE to date and call for future research to validate and evaluate its impact on cardiovascular and mortality outcomes, as well as healthcare resource utilization in multiethnic populations and different healthcare settings.

Predictive models in chronic kidney disease: essential tools in clinical practice

PURPOSE OF REVIEW

The integration of risk prediction in managing chronic kidney disease (CKD) is universally considered a key point of routine clinical practice to guide time-sensitive choices, such as dialysis access planning or counseling on kidney transplant options. Several prognostic models have been developed and validated to provide individualized evaluation of kidney failure risk in CKD patients. This review aims to analyze the current evidence on existing predictive models and evaluate the different advantages and disadvantages of these tools.

RECENT FINDINGS

Since Tangri et al. introduced the Kidney Failure Risk Equation in 2011, the nephrological scientific community focused its interest in enhancing available algorithms and finding new prognostic equations. Although current models can predict kidney failure with high discrimination, different questions remain unsolved. Thus, this field is open to new possibilities and discoveries.

SUMMARY

Accurately informing patients of their prognoses can result in tailored therapy with important clinical and psychological implications. Over the last 5 years, the number of disease-modifying therapeutic options has considerably increased, providing possibilities to not only prevent the kidney failure onset in patients with advanced CKD but also delay progression from early stages in at-risk individuals.

Multicentre external validation of the prognostic model kidney failure risk equation in patients with CKD stages 3 and 4 in Peru: a retrospective cohort study

OBJECTIVES

To externally validate the four-variable kidney failure risk equation (KFRE) in the Peruvian population for predicting kidney failure at 2 and 5 years.

DESIGN

A retrospective cohort study.

SETTING

17 primary care centres from the Health's Social Security of Peru.

PARTICIPANTS

Patients older than 18 years, diagnosed with chronic kidney disease stage 3a-3b-4 and 3b-4, between January 2013 and December 2017. Patients were followed until they developed kidney failure, died, were lost, or ended the study (31 December 2019), whichever came first.

PRIMARY AND SECONDARY OUTCOME MEASURES

Performance of the KFRE model was assessed based on discrimination and calibration measures considering the competing risk of death.

RESULTS

We included 7519 patients in stages 3a-4 and 2798 patients in stages 3b-4. The estimated cumulative incidence of kidney failure, accounting for competing event of death, at 2 years and 5 years, was 1.52% and 3.37% in stages 3a-4 and 3.15% and 6.86% in stages 3b-4. KFRE discrimination at 2 and 5 years was high, with time-dependent area under the curve and C-index >0.8 for all populations. Regarding calibration in-the-large, the observed to expected ratio and the calibration intercept indicated that KFRE underestimates the overall risk at 2 years and overestimates it at 5 years in all populations.

CONCLUSIONS

The four-variable KFRE models have good discrimination but poor calibration in the Peruvian population. The model underestimates the risk of kidney failure in the short term and overestimates it in the long term. Further research should focus on updating or recalibrating the KFRE model to better predict kidney failure in the Peruvian context before recommending its use in clinical practice.

Diagnosis of Chronic Kidney Disease and Assessing Glomerular Filtration Rate

Chronic kidney disease (CKD) is a silent progressive disease. It is diagnosed by assessing filtration and markers of kidney damage such as albuminuria. The diagnosis of CKD should include not only assessing the glomerular filtration rate (GFR) and albuminuria but also the cause. The CKD care plan should include documentation of the trajectory and prognosis. The use of a combination of serum cystatin C and creatinine concentration offers a more accurate estimation of GFR. Social determinants of health are important to address as part of the diagnosis because they contribute to CKD disparities.

ESKD Risk Prediction Model in a Multicenter Chronic Kidney Disease Cohort in China: A Derivation, Validation, and Comparison Study

BACKGROUND AND OBJECTIVES

In light of the growing burden of chronic kidney disease (CKD), it is of particular importance to create disease prediction models that can assist healthcare providers in identifying cases of CKD individual risk and integrate risk-based care for disease progress management. The objective of this study was to develop and validate a new pragmatic end-stage kidney disease (ESKD) risk prediction utilizing the Cox proportional hazards model (Cox) and machine learning (ML).

DESIGN, SETTING, PARTICIPANTS, AND MEASUREMENTS

The Chinese Cohort Study of Chronic Kidney Disease (C-STRIDE), a multicenter CKD cohort in China, was employed as the model's training and testing datasets, with a split ratio of 7:3. A cohort from Peking University First Hospital (PKUFH cohort) served as the external validation dataset. The participants' laboratory tests in those cohorts were conducted at PKUFH. We included individuals with CKD stages 1~4 at baseline. The incidence of kidney replacement therapy (KRT) was defined as the outcome. We constructed the Peking University-CKD (PKU-CKD) risk prediction model employing the Cox and ML methods, which include extreme gradient boosting (XGBoost) and survival support vector machine (SSVM). These models discriminate metrics by applying Harrell's concordance index (Harrell's C-index) and Uno's concordance (Uno's C). The calibration performance was measured by the Brier score and plots.

RESULTS

Of the 3216 C-STRIDE and 342 PKUFH participants, 411 (12.8%) and 25 (7.3%) experienced KRT with mean follow-up periods of 4.45 and 3.37 years, respectively. The features included in the PKU-CKD model were age, gender, estimated glomerular filtration rate (eGFR), urinary albumin-creatinine ratio (UACR), albumin, hemoglobin, medical history of type 2 diabetes mellitus (T2DM), and hypertension. In the test dataset, the values of the Cox model for Harrell's C-index, Uno's C-index, and Brier score were 0.834, 0.833, and 0.065, respectively. The XGBoost algorithm values for these metrics were 0.826, 0.825, and 0.066, respectively. The SSVM model yielded values of 0.748, 0.747, and 0.070, respectively, for the above parameters. The comparative analysis revealed no significant difference between XGBoost and Cox, in terms of Harrell's C, Uno's C, and the Brier score (p = 0.186, 0.213, and 0.41, respectively) in the test dataset. The SSVM model was significantly inferior to the previous two models (p < 0.001), in terms of discrimination and calibration. The validation dataset showed that XGBoost was superior to Cox, regarding Harrell's C, Uno's C, and the Brier score (p = 0.003, 0.027, and 0.032, respectively), while Cox and SSVM were almost identical concerning these three parameters (p = 0.102, 0.092, and 0.048, respectively).

CONCLUSIONS

We developed and validated a new ESKD risk prediction model for patients with CKD, employing commonly measured indicators in clinical practice, and its overall performance was satisfactory. The conventional Cox regression and certain ML models exhibited equal accuracy in predicting the course of CKD.