External validation, recalibration, and clinical utility of the kidney failure risk equation in patients with advanced CKD: a nationwide retrospective cohort analysis in Peru

External validation, recalibration, and clinical utility of the kidney failure risk equation in patients with advanced CKD: a nationwide retrospective cohort analysis in Peru | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article External validation, recalibration, and clinical utility of the kidney failure risk equation in patients with advanced CKD: a nationwide retrospective cohort analysis in Peru Jessica Ivonne Bravo-Zúñiga, Percy Soto-Becerra, Edgar Juan Coila-Paricahua, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5520011/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Dec, 2025 Read the published version in BMC Nephrology → Version 1 posted 6 You are reading this latest preprint version Abstract Background The Kidney Failure Risk Equation (KFRE) is widely used for predicting kidney failure, but its external validity in Latin America is limited. A previous study in Peru found that KFRE was miscalibrated but did not evaluate its recalibration or clinical utility. Methods We conducted a retrospective cohort study using data from EsSalud’s Renal Health Surveillance Program (2013–2022), including 30,031 patients with chronic kidney disease (CKD) stages G3-4. Kidney failure was defined by dialysis initiation or nephrologist-confirmed end-stage renal disease. Calibration was assessed using observed-to-expected (O/E) ratios and differences, calibration slope, and intercept, while discrimination was evaluated using the concordance index (C-index). Recalibrated models were developed, and decision curve analysis (DCA) was performed to evaluate clinical utility. Results The original KFRE demonstrated good discrimination (C-index: 0.88 at 2 years, 0.85 at 5 years) but poor calibration in-the-large: O/E ratios indicated mean underestimation of risk at 2 years (O/E ratio: 1.84) and a slight mean overestimation at 5 years (O/E ratio: 1.06). Original KFRE also had poor weak (slope: 0.58) and poor moderate calibration. Recalibrated models improved calibration in-the-large, but none achieved good weak (all slope < 1) and moderate calibration. However, DCA showed a higher net benefit for KFRE-based nephrology referrals (in original and recalibrated by method D) compared to Peruvian and international guidelines, especially over a 5-year horizon. Conclusions Despite miscalibration, KFRE remains valuable for guiding nephrology referrals in Peru, with recalibrated models offering potential improvements. This is the first study in Latin America to rigorously assess the clinical utility of KFRE. Chronic Kidney Disease Kidney Failure Risk Equation External Validation Clinical Utility Peru Decision Curve Analysis Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 BACKGROUND Chronic kidney disease (CKD) is a growing global health challenge, affecting about 10% of the population and often leading to kidney failure, requiring costly interventions like dialysis or transplantation 1 , 2 . Recent guidelines, including NICE 2021 and KDIGO 2024 3,4 , emphasize that accurate, individualized risk prediction models can help reduce the burden of CKD by identifying high-risk patients who may benefit from early interventions. This enables healthcare providers to make informed decisions about when to refer patients to nephrology, allowing for the implementation of nephroprotective strategies to slow disease progression and facilitating timely planning for renal replacement therapy (RRT), ultimately improving patient outcomes and optimizing healthcare resources. While international guidelines recommend the use of the Kidney Failure Risk Equation (KFRE) for predicting kidney failure 3 , 4 , its implementation in Peru remains limited 5 , 6 . The Peruvian Clinical Practice Guideline instead relies on a combination of estimated glomerular filtration rate (eGFR) and albuminuria for nephrology referrals, similar to the earlier approach of NICE 2014 guidelines 7 . Despite endorsements from the Latin American Nephrology Society to recommend KFRE 8 , concerns have persisted due to the lack of region-specific evidence supporting its utility. This highlights the importance of externally validating prognostic models like KFRE within the specific settings where they are intended to be used 9 – 11 , particularly in the Peruvian and broader Latin American context. Although KFRE has shown strong prognostic performance in international contexts 12 – 28 , recent studies have identified miscalibration in various countries 16 , 17 , 19 , 22 , 23 , 29 , 30 . In Peru, a recent study found KFRE to perform well in a cohort from Lima, the country's capital and largest urban center, but did not undertake recalibration, raising questions about its nationwide applicability 31 . Moreover, region-specific evidence in Latin America is scarce, with no studies evaluating the clinical utility of KFRE 32 , 33 . Addressing this gap is essential to provide evidence that supports the use of KFRE in the Peruvian context and informs tailored CKD management strategies. Therefore, this study aims to perform an external validation and recalibration of the KFRE to predict kidney failure, as well as assess its clinical utility in guiding decisions for referral and planning for RRT in a nationwide cohort of patients with CKD stages G3-4 in Peru. METHODS Study design, population and data source We conducted a retrospective cohort study following the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) guidelines 11 , 34 to externally validate, recalibrate and assess the clinical utility of the KFRE model for predicting the risk of kidney failure at 2-year and 5-year horizons in patients diagnosed with chronic kidney disease (CKD) stages 3–4 (eGFR > 15 and < 60 ml/min per 1.73 m² ) between 2013 and 2022 across all 45 care networks of EsSalud (see TRIPOD checklist in online Supplementary Material ). The data were collected and managed by the National Renal Health Center as part of the national renal health surveillance system (VISARE, by its Spanish acronym). For health centers within the Rebagliati Network, the largest in Lima, data were directly sourced from the UMERC informatics application designed specifically for this network to provide information to VISARE, as previously described 31 . Additionally, data from the Kaelin Hospital were provided directly by the hospital itself due to its special status as a public-private partnership, which maintains its own renal health surveillance and does not routinely report to the CNSR. Validation model and predictors We validated the 4-variable KFRE, which includes age (scaled to 10 years), eGFR (ml/min per 1.73 m² using the CKD-EPI formula), sex, and urinary albumin/creatinine ratio (ACR) in mg/g 25 . Standardized laboratory protocols were followed across health facilities to measure these predictors (see Supplementary Methods for details on KFRE equations, coding of predictors, CKD-EPI formula and laboratory considerations). Outcome variable The primary outcome was the time to kidney failure, defined as the date of the first hemodialysis or peritoneal dialysis based on administrative data from the National Center for Renal Health or through specific ICD-10 diagnoses (e.g., N18.5, N18.6, Z99.2, Z49.1, Z49.2, Z94.0). This computational phenotyping approach has been validated in previous studies and is further detailed in the Supplementary Methods . Death was considered a competing risk, with mortality data obtained from both the National Death System of Peru and the National Registry of Identification and Civil Status, which together cover more than 90% of all deaths in the country. For more information on the outcome definitions and data sources, please refer to the supplementary material . Follow-up time Patients were followed until kidney failure, death or the administrative censorship date (December 31, 2022), whichever occurred first. Censoring occurred at loss to follow-up or study end. Sample size Due to the comprehensive nature of the national dataset, a specific sample size calculation was unnecessary. All patients meeting inclusion criteria were analyzed (see Fig. 1 and Table S4). Since the number of events per EsSalud health network was significantly low—often fewer than 100 events, and in many instances, nearly zero—it was decided to evaluate the performance of KFRE across the entire country without considering regions as separate clusters. Statistical Analysis Initial data analyses were performed to identify extreme values, inconsistencies, and missing data. Winsorization of albumin-creatinine ratio (ACR) values at the 1.5th and 98.5th percentiles was conducted 35 . Missing data were managed using multiple imputation via Additive Regression, Bootstrapping, and Predictive Mean Matching, creating 100 imputed datasets 36 . To ensure that the imputation model was congenial (at least semicompatible) with the competing risk models used in the substantive analysis, interactions between all predictors and the cumulative baseline hazards for kidney failure and death were included in the imputation process 37 . This step aimed to improve the accuracy of the imputations and their compatibility with the intended analyses (see Supplementary Methods for a detailed description). For the external validation of the KFRE, model discrimination was assessed using the concordance index (C-index) at 2 and 5 years. In line with the TRIPOD guidelines, model performance was evaluated through both discrimination and calibration assessments, incorporating the consideration of competing risks based on recent methodological recommendations 11 . Calibration was evaluated through assessment of calibration in-the-large: observed-to-expected (O/E) ratios, difference and calibration intercepts; weak calibration: slopes calibration; and moderate calibration via calibration plots. All these measures were obtained accounting for competing risk 38 , 39 . The calibration curves were obtained using LOESS smoothing applied to the cumulative baseline hazard predicted by the Fine-Gray model for subdistribution hazards 38 , 39 . See Supplementary Methods for further details. Recalibration was performed using four methods 35 : Methods A and B used the traditional Cox model without accounting for competing risks, as originally proposed in the KFRE model by Tangri et al 25 . Method A adjusted baseline risk only, while Method B adjusted both baseline risk and the magnitude of the linear predictor. Methods C and D, in contrast, employed the Cause-Specific Cox model to account for competing risks. Method C adjusted the baseline risk, and Method D combined baseline risk adjustment with linear predictor adjustment while considering competing risks. Clinical utility was assessed using decision curve analysis 40 by comparing the net benefit of the original and recalibrated KFRE models against the Peruvian National Guidelines 5 , 6 and NICE 2014 Guidelines for nephrology referral 7 . For each prediction horizon (2 and 5 years), we evaluated the utility of using KFRE to guide decisions for referral and planning for renal replacement therapy. Predefined reasonable decision thresholds (based on existing literature) were used to identify patients who might benefit from referral. For long-term management and nephroprotection to halt or reverse CKD progression, a 5-year horizon was used, with thresholds typically ranging between 3–5% 5,6 . For planning renal replacement therapy, the 2-year horizon was considered, with decision thresholds set at higher probabilities, usually around 20–40% 5,6 . The final estimates and standard errors were pooled across the imputed datasets using Rubin's rules, and 95% confidence intervals were calculated based on these standard errors 41 . All statistical analyses were performed using R version 4.3. The reproducible code used for this analysis is available in an open GitHub repository ( https://github.com/psotob91/kfre-ckd-nationwide-essalud-peru ). Language Editing Assistance The authors utilized ChatGPT-4o, a Large Language Model (LLM) developed by OpenAI, to review and improve the English grammar and style of this manuscript. The AI tool was employed solely for language editing purposes and was not used to generate or create any content. RESULTS Study Population Out of 152,084 patients screened between January 1, 2013, and December 30, 2022, in all EsSalud facilities nationwide under the VISARE program, 30,031 met the selection criteria for CKD stages G3-4 (Fig. 2 ). Only 38.4% (11,540) had complete data for the four variables required for KFRE estimation. After multiple imputations, all eligible individuals were included in the analysis. Table 1 summarizes the key sociodemographic and clinical characteristics of the study population after imputation. Of the total, 56.4% were women, and ages ranged from 18 to 109 years, with a median age of 73.8 years. The prevalence of diabetes mellitus (41.5%) and hypertension (75.8%) was high, with most participants classified as stage G3a (61.6%) or G3b (27.3%). A detailed comparison of population characteristics with and without imputation is provided and according to outcome are showed in Table S5 and S6 , respectively. Table 1 Baseline Characteristics (at first recorded evaluation in VISARE) of patients with CKD G3-4 Included in the Analysis Characteristic N = 30,031 Sex Female 13,097 (43.6%) Male 16,934 (56.4%) Age (years) Mean (SD) 73.8 (11.1) Median (Q1 - Q3) 75.0 (67.0–82.0) Min - Max 18.0–109.0 EsSalud Network Metropolitan Lima 14,784 (49.2%) Other Regions 15,247 (50.8%) Hypertension No 7,254 (24.2%) Yes 22,777 (75.8%) Diabetes Mellitus No 17,567 (58.5%) Yes 12,464 (41.5%) Persistent Albuminuria Categories A1 14, (48.8%) A2 10,095 (33.6%) A3 5,274 (17.6%) eGFR Categories G3a 18,491 (61.6%) G3b 8,201 (27.3%) G4 3,339 (11.1%) CKD KDIGO Classification Low risk 0 (0.0%) Moderately increased risk 9,403 (31.3%) High risk 10,169 (33.9%) Very high risk 10,459 (34.8%) Serum Creatinine (mg/dL) Mean (SD) 1.4 (0.4) Median (Q1 - Q3) 1.3 (1.2–1.6) Min - Max 0.9–4.5 eGFR using CKD-EPI (ml/min/1.73m²) Mean (SD) 45.9 (10.8) Median (Q1 - Q3) 48.5 (39.4–54.5) Min - Max 15.0–60.0 Albumin-Creatinine Ratio (mg/g) Mean (SD) 802.1 (3,534.1) Median (Q1 - Q3) 32.0 (8.1–160.3) Min - Max 0.6–27,817.5 Urine Albumin (mg/dl) Mean (SD) 35.8 (161.5) Median (Q1 - Q3) 1.6 (0.4–8.4) Min - Max 0.0–1,365.2 Urine Creatinine (mg/dL) Mean (SD) 60.8 (50.5) Median (Q1 - Q3) 49.2 (27.0–85.0) Min - Max 0.1–221.3 Death at 2 years* No 27,640 (92.0%) Yes 2,391 (8.0%) Outcomes at 2 years Alive w/o Kidney Failure 27,227 (90.7%) Kidney Failure 793 (2.6%) Death w/o Kidney Failure 2,011 (6.7%) Death at 5 years* No 24,261 (80.8%) Yes 5,770 (19.2%) Outcomes at 5 years Alive w/o Kidney Failure 23,579 (78.5%) Kidney Failure 1,308 (4.4%) Death w/o Kidney Failure 5,144 (17.1%) *Death after o before kidney failure. SD: standard deviation, IQR: first quartile and third quartile, ACR, urine albumin to creatinine ratio; CKD, chronic kidney disease; eGFR, glomerular filtration rate estimated by CKD Epidemiology Collaboration formula Table 1 Cumulative incidences of kidney failure at 2 and 5 years were 2.73% (95% CI: 2.55%-2.92%) and 4.76% (95% CI: 4.51%-5.02%), respectively ( Table S2 ). The cumulative incidences of death without kidney failure were notably higher, with values of 6.96% (95% CI: 6.67%-7.26%) at 2 years and 19.71% (95% CI: 19.22%-20.2%) at 5 years. The detailed cumulative incidence curve for both kidney failure and death are shown in Fig. 3 . 2-year and 5-year KFRE’s predicted risk is shown in Figure S1 . The distribution of four KFRE equation variables is shown in Figure S2 . Predictive Performance of the Original 4-Variable KFRE Equation KFRE showed good discrimination across all time horizons, with C-indices of 0.88 (95% CI: 0.86–0.89) at 2 years and 0.85 (95% CI: 0.84–0.87) at 5 years (Table 2 ). However, calibration was poor across all time horizons. Table 2 External Validation Metrics for Predictive Performance of the Original 4-Variable KFRE Model Validation aspect and performance measure Time horizon t = 2 years t = 5 years Calibration Average predicted risk 1.48% 4.48% Average observed risk (95% CI) 2.73% (2.54–2.92%) 4.76% (4.51–5.02%) O/E ratio (95% CI) 1.84 (1.7 to 1.99) 1.06 (1 to 1.13) O-E difference (95% CI) 1.25% (1.05–1.44%) 0.29% (0–0.58%) Calibration intercept (95% CI) 0.02 (-0.13 to 0.18) -0.47 (-0.59 to -0.35) Calibration slope (95% CI) 0.58 (0.53 to 0.63) 0.58 (0.54 to 0.62) Discrimination C-index up to t years (95% CI) 0.88 (0.86 to 0.89) 0.85 (0.84 to 0.87) %, percentage; C-index, truncated agreement index; CKD, chronic kidney disease; O/E and O-E, observed vs expected ratio and differences, respectively; t, time At 2 years, the model underestimated the risk of kidney failure, as indicated by an observed-to-expected (O/E) ratio of 1.84 (95% CI: 1.7–1.99). Although the calibration intercept was not statistically different from zero (0.02, 95% CI: -0.13 to 0.18), the pronounced O/E ratio suggests that the model underestimates the actual risk on average. At 5 years, the O/E ratio was closer to 1 at 1.06 (95% CI: 1 to 1.13), which might initially suggest an alignment between predicted and observed risks. However, the calibration intercept was significantly negative (-0.47, 95% CI: -0.59 to -0.35), reinforcing the presence of systematic underestimation of the mean risk. In both time horizons, the calibration slope was below 1 (0.58 for both 2 and 5 years), indicating overly extreme predictions — underestimating risks for low-risk individuals and overestimating for high-risk individuals. Calibration curves in Fig. 4 further illustrate these trends. Overall, these results highlight that KFRE has poor mean, weak and moderate calibration in Peruvian population of EsSalud. Recalibration of the KFRE Model Recalibration was performed using four methods (A-D). Recalibrated equations has shown in Table 3 . Table 3 Original and recalibrated equations Time horizon Equations Original model 2 years \(\:1-{0.9832}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) 5 years \(\:1-{0.9365}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) Method A: Baseline risk adjustment without considering competing risk 2 years \(\:1-{0.9688}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) 5 years \(\:1-{0.9363}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) Method B: Baseline risk adjustment + adjustment of linear predictor magnitude without considering competing risk 2 years \(\:1-{0.9550}^{{e}^{0.7126042\times\:(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) 5 years \(\:1-{0.9130}^{{e}^{0.7126042\times\:(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) Method C: Baseline risk adjustment considering competing risk 2 years \(\:1-{0.9699}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) 5 years \(\:1-{0.9425}^{{e}^{(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) Method D: Baseline risk adjustment + adjustment of linear predictor magnitude considering competing risk 2 years \(\:1-{0.9572}^{{e}^{0.7126042\times\:(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) 5 years \(\:1-{0.9240}^{{e}^{0.7126042\times\:(-0.2201\times\:\left(\frac{age}{10}-7.036\right)+0.2467\times\:\left(sex-0.5642\right)-0.5567\times\:\left(\frac{eGFR}{5}-7.222\right)+0.4510\times\:\left(\text{log}\left(ACR\right)-5.137\right))}}\) eGFR, estimated glomerular filtration rate; urine albumin/creatinine ratio (ACR) As shown in Table 4 , Method D had the best calibration-in-the-large at both 2 and 5 years, with O/E ratios close to 1 (1.02 at 2 years; 1.04 at 5 years) and calibration intercepts not significantly different from zero. Despite this, the calibration slope (0.81 for both time horizons) indicates some underestimation in low-risk individuals and overestimation in high-risk ones, failing in achieve good weak calibration. Table 4 Validation metrics for predictive performance of recalibrated KFRE models Validation aspect and performance measure Method A 1 Method B 2 Method C 3 Method D 4 t = 2 years Calibration Average predicted risk 2.52% 2.82% 2.45% 2.69% Average observed risk (95% CI) 2.73% (2.54–2.92%) 2.73% (2.54–2.92%) 2.73% (2.54–2.92%) 2.73% (2.54–2.92%) O/E ratio (95% CI) 1.08 (1.01 to 1.16) 0.97 (0.9 to 1.04) 1.11 (1.04 to 1.19) 1.02 (0.95 to 1.09) O-E difference (95% CI) 0.21% (0.02–0.4%) -0.09% (-0.28–0.1%) 0.28% (0.09–0.47%) 0.04% (-0.15–0.23%) Calibration intercept (95% CI) -0.6 (-0.74 to -0.46) -0.1 (-0.2 to 0) -0.57 (-0.71 to -0.42) -0.05 (-0.15 to 0.05) Calibration slope (95% CI) 0.58 (0.53 to 0.63) 0.82 (0.75 to 0.88) 0.58 (0.53 to 0.63) 0.82 (0.75 to 0.88) Discrimination C-index up to t years (95% CI) 0.88 (0.86 to 0.89) 0.88 (0.86 to 0.89) 0.88 (0.86 to 0.89) 0.88 (0.86 to 0.89) t = 5 years Calibration Average predicted risk 4.48% 5.19% 4.14% 4.59% Average observed risk (95% CI) 4.76% (4.51–5.02%) 4.76% (4.51–5.02%) 4.76% (4.51–5.02%) 4.76% (4.51–5.02%) O/E ratio (95% CI) 1.06 (1.01 to 1.12) 0.92 (0.87 to 0.97) 1.15 (1.09 to 1.22) 1.04 (0.98 to 1.1) O-E difference (95% CI) 0.28% (0.02–0.54%) -0.42% (-0.68% to -0.17%) 0.62% (0.37–0.88%) 0.18% (-0.08–0.43%) Calibration intercept (95% CI) -0.47 (-0.57 to -0.37) -0.16 (-0.23 to -0.08) -0.36 (-0.47 to -0.26) -0.02 (-0.09 to 0.06) Calibration slope (95% CI) 0.58 (0.54 to 0.62) 0.81 (0.76 to 0.86) 0.58 (0.54 to 0.62) 0.81 (0.76 to 0.86) Discrimination C-index up to t years (95% CI) 0.85 (0.84 to 0.87) 0.85 (0.84 to 0.87) 0.85 (0.84 to 0.87) 0.85 (0.84 to 0.87) %, percentage; C-index, truncated agreement index; CKD, chronic kidney disease; O/E and O-E, observed vs expected ratio and differences, respectively; time, time 1 Method A: Baseline risk adjustment without considering competing risk 2 Method B: Baseline risk adjustment + adjustment of linear predictor magnitude without considering competing risk 3 Method C: Baseline risk adjustment considering competing risk 4 Method D: Baseline risk adjustment + adjustment of linear predictor magnitude considering competing risk The calibration plots for Method D show a reasonable alignment with the ideal 45-degree line, with minor overestimation at higher risk levels (Fig. 5 ). It's important to note that this overestimation at the upper tail involves fewer data points, which may impact the curve's stability. In contrast, Methods A, B, and C show more pronounced deviations, especially at the extremes. Thus, while Method D is not perfectly calibrated, it offers the most balanced performance among the recalibrated models. Clinical utility of original model KFRE and recalibrated versions Figure 6 presents the decision curve analysis comparing the net benefits of the original and recalibrated KFRE models across 2-year and 5-year horizons. Method D demonstrated the highest net benefit across most thresholds; however, the difference was modest compared to the original KFRE model. For example, at a 5% threshold over 5 years, Method D had a net benefit of 0.0261, nearly identical to the original model's 0.0255. All models, including the original KFRE and its recalibrated versions, outperformed alternative referral strategies based on the NICE 2014 guidelines and Peruvian National Guidelines across a range of thresholds. This suggests that the KFRE models, despite calibration issues, effectively balance harm and benefit by accurately identifying more patients for referral while minimizing unnecessary ones. At the 2-year horizon, the net benefit of both the original KFRE and the recalibrated models was slightly superior to the "refer none" strategy at threshold probabilities of 20% and 30%. For instance, at a 20% threshold over 2 years, Method D had a net benefit of 0.0016, while the original model had a net benefit of 0.0015. However, at higher threshold probabilities (> 40%), the net benefit of these models became lower than the "refer none" strategy, indicating limited utility in these scenarios. In summary, Methods D and B showed the highest clinical utility, consistently demonstrating the highest net benefits at both 2- and 5-year horizons. However, these differences were relatively small, indicating that the original KFRE model remains a viable option for clinical decision-making. Sensitivity Analysis A sensitivity analysis was performed without applying winsorization to the extreme values of the ACR. Figure S3 and S4 show the distribution of KFRE’s predicted risks and four variables, respectively in the original dataset without winsorization of ACR. The results remained similar to the primary analysis ( Table S9 , Figures S5 , S6 and S7 ), indicating that the winsorization of ACR values did not significantly impact the predictive performance or calibration of the KFRE model. DISCUSSION Main findings This study externally validated the KFRE multivariable model in a national cohort of EsSalud patients with CKD stages 3–4, recalibrated KFRE for this population, and assessed the clinical utility of the original and recalibrated versions. While KFRE showed strong discrimination for predicting kidney failure at both 2 and 5 years, it exhibited poor calibration. Two recalibrated models (Methods B and D) improved calibration-in-the-large, yet all recalibrated models struggled with weak or moderate calibration. The recalibrated models maintained a similar pattern to the original KFRE, overestimating risk for high-risk individuals and underestimating it for low-risk individuals. Despite these calibration issues, the original and recalibrated KFRE versions offered a net benefit compared to the strategy of referring no patients. Their net benefit also surpassed that of the Peruvian National Guidelines and NICE 2014 guidelines across various thresholds. Thus, while the original KFRE and its recalibrated versions may not perfectly predict individual risk, they remain useful tools for guiding nephrology referrals in Peru. When assessing early referral for renal replacement therapy preparation, the net benefit of KFRE at 2 years turned negative at higher thresholds, indicating that unnecessary referrals (false positives) may outweigh correct ones (true positives). However, for decisions involving long-term referral at 5 years, the KFRE models showed a positive net benefit across thresholds of 3–10%, indicating their utility in effectively identifying high-risk patients who would benefit from early nephrology intervention. Comparison with previous literature The 4-variable KFRE model has been externally validated in over 30 countries across all continents 42 : North America 12 – 15 , 43 – 46 , Europe 16 – 18 , Asia 19 – 23 , Oceania 24 , 30 , and more recently, Latin America 31 – 33 . However, apart from the cohorts from Chile and Brazil in the 1990s used to recalibrate the initial KFRE model and derive a specific equation for non-North American countries 25 , evidence on external validity in Latin America was absent for nearly two decades. Previous research in Peru showed KFRE's good discrimination but highlighted poor calibration, a finding we confirmed in a larger national cohort 31 . Other studies in Latin America, such as those in Colombia and Uruguay, reported good discrimination and calibration but lacked adequate validation methodologies, limiting the comparability of their results with our findings 32 , 33 . These methodological issues underscore the urgent need to generate high-quality evidence on the applicability of KFRE in clinical practice in Latin America, ensuring that well-sound methods are used to properly validate prognostic models. In contrast, the literature from non-North American regions, excluding Latin America, is more extensive. Globally, studies consistently demonstrate KFRE's high discrimination (> 0.80) 12–28 , but have identified moderate calibration issues, particularly with overprediction in high-risk groups. Differences in predictor profiles and incidence rates do not fully account for the poor calibration observed in Peru. For example, the renal failure incidence rate in our study closely matches that of the original model (see Table S8 ). However, our population's characteristics, including higher albuminuria levels, higher diabetes prevalence, and exclusion of G5 stages, may contribute to the observed miscalibration. Importantly, our decision to exclude stage G5 patients is well supported by both the intended clinical use of the KFRE model and prevailing practices in the literature. The KFRE is primarily designed for use in outpatient settings where earlier stages of CKD (G3–4) are managed, and its application in stage G5 is inherently less useful, as these patients are already recognized as high risk and often follow different management pathways (e.g., conservative or palliative care). Notably, Tangri’s original derivation and validation cohorts included a very low proportion of G5 patients (approximately 5.3%) 25 , and several subsequent external validation studies have similarly focused on patients with CKD stages G3–4 27,47 . This selective validation approach is common, valid, and reasonable, as it ensures that the model’s calibration and performance are most relevant to the population where it is intended to be applied 34 , 35 , 48 . It is also important to note that although one might argue that the exclusion of stage G5 patients could explain part of the miscalibration, evidence from studies such as Ramspek et al. 16 indicates that calibration issues persist even when G5 patients are included. Ramspek et al. 16 attribute these issues not solely to the inclusion of patients with advanced CKD, but to a combination of factors—including heterogeneity in clinical profiles and the failure to account for the competing risk of death. In fact, Tangri's original model did not account for this competing risk, leading to an overprediction of renal failure risk in settings where the incidence of death is high 25 . In our population, incidence of pre-dialysis death is 19.7% at five years, nearly four times the incidence of renal failure for the same period, making it likely that the Cox model would overestimate renal failure in these high-risk group. This explanation aligns with our observations and further supports our decision to exclude stage G5 patients, thereby validating the model in the specific outpatient context where it is intended to be applied. A key finding of this study is that the KFRE model retains clinical utility despite its miscalibration. This can be understood through the net benefit framework, which emphasizes that a model may still be valuable if the miscalibrations occur infrequently or in less critical subpopulations. In this study, the overall miscalibration was most pronounced in patients at the extremes of risk, who represent a smaller fraction of the population. Most patients had reasonably calibrated risk predictions, contributing to a higher net benefit in clinical decision-making. Therefore, even with calibration issues, the KFRE model remains a robust tool for predicting kidney failure and guiding appropriate referrals, particularly in resource-limited settings where optimizing healthcare allocation is crucial. Importantly, this study is the first in Latin America to rigorously evaluate the clinical utility of the KFRE model, highlighting its potential role in enhancing CKD management strategies in the region. Strengths and limitations of this study A major strength of this study is its large sample size of 30,031 CKD patients, providing robust statistical power. Additionally, the use of VISARE surveillance data captures patients at primary healthcare centers nationwide, who are often identified through screening programs targeting individuals with diabetes, hypertension, or those over 55 years of age. This real-world data from a nationwide screening program at the primary care level reflects a population with characteristics that are both relevant and underrepresented in the literature, providing valuable insights into CKD management. However, we excluded 22,627 patients due to missing eGFR and other critical variables, with most missing data attributed to albumin-to-creatinine ratio (ACR) measurements (60.7%). This raises concerns about the usability of KFRE in settings where ACR testing is limited, emphasizing the need to strengthen laboratory capacities, particularly outside of Lima, the capital of the country. In many regions, the lack of ACR data is primarily due to shortages of supplies, which limits the full implementation of risk prediction models like KFRE. To address missing data in our study, multiple imputation was performed under the assumption that data were missing at random (MAR). We included auxiliary variables such as age, sex, comorbidities, and lab values to make the MAR assumption more plausible and enhance the accuracy of the imputations. Nevertheless, the validity of this assumption cannot be guaranteed, which remains a limitation. The geographical bias, with data predominantly from Lima, also underscores the need for further research in less-represented regions. Measurement errors in laboratory data, particularly outside of Lima, further highlight the necessity for improved healthcare infrastructure, including consistent access to ACR testing, to better support CKD risk prediction and management. Implications for clinical practice and health systems The KFRE model's superior net benefit for predicting kidney failure at 5 years offers a key opportunity for early referral for secondary prevention, allowing nephrology interventions to slow disease progression. For 2-year predictions, the model supports planning for renal replacement therapy or conservative treatment. Recalibrated KFRE versions outperformed current national referral strategies and NICE 2014 guidelines across various threshold probabilities. This aligns with NICE 2024's updated recommendation of a KFRE score > 5% or an ACR > 70 for nephrology referral. In settings with limited nephrology resources, the KFRE's ability to identify high-risk patients can optimize referrals and prevent unnecessary ones. The use of a 3–5% risk threshold over 5 years has shown to be effective in various healthcare settings. Retrospective studies in Canada and the UK found that these thresholds reduced late referrals for patients progressing to kidney failure. Similarly, prospective evaluations noted shorter nephrology wait times, particularly for high-risk individuals. However, the model's utility at 2 years requires caution. Net benefit analysis suggests an advantage at thresholds of 20–30%, but this diminishes at thresholds above 35–40%, where unnecessary referrals outweigh the benefits. For patients with higher risk probabilities, additional tests may be required to avoid premature dialysis preparations. Lower thresholds like > 20% can help optimize sensitivity, aiding in early dialysis planning or transplant referral. Future research Improving the KFRE model's calibration, especially for 2-year predictions, may require substantial updates, including re-estimating regression coefficients or modifying predictors. While promising, this approach risks overfitting and model instability. Therefore, using the current KFRE model remains more practical despite its limitations. Future studies should explore KFRE's predictive performance in subgroups like children, young adults, and those with diabetes, as emphasized by NICE 2024 guidelines. Assessing simplified versions of the model, such as the 3-variable KFRE or proxies like urine dipsticks, could also enhance its clinical applicability. In addition, cost-effectiveness analyses should be conducted to evaluate the economic impact of implementing KFRE in clinical practice. Implementation studies examining the integration of KFRE into clinical workflows and trials that evaluate its impact on patient outcomes are needed to further establish its role in diverse healthcare settings. Conclusion Despite calibration issues, the KFRE model, especially in its recalibrated forms, remains a valuable tool for guiding nephrology referrals in Peru, providing a higher net benefit than current national guidelines. Its use can enhance early identification of high-risk patients, improving healthcare resource allocation in resource-limited settings. Further research is needed to refine the model, explore its utility in diverse subgroups, and integrate it effectively into clinical practice. Abbreviations ACR Albumin-to-Creatinine Ratio C-index Concordance Index CKD Chronic Kidney Disease CKD-EPI Chronic Kidney Disease Epidemiology Collaboration CNSR National Renal Health Center DCA Decision Curve Analysis eGFR Estimated Glomerular Filtration Rate EsSalud Social Health Insurance of Peru ICD-10 International Classification of Diseases, 10th Revision KFRE Kidney Failure Risk Equation KDIGO Kidney Disease:Improving Global Outcomes MAR Missing At Random NICE National Institute for Health and Care Excellence O/E ratio Observed-to-Expected Ratio RRT Renal Replacement Therapy TRIPOD Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis UMERC Informatics application used to provide information to VISARE VISARE Renal Health Surveillance Declarations ETHICS APPROVAL AND CONSENT TO PARTICIPATE This study was approved by the Research Ethics Committee of the Edgardo Rebagliati National Hospital (Code No. 519-GRPR-ESSALUD-2023) and was conducted in accordance with the principles of the Declaration of Helsinki. A waiver of informed consent was granted by the committee because the study involved routinely collected secondary data. Given the minimal risks associated with this type of research, this waiver was deemed a reasonable justification and was approved accordingly. CONSENT FOR PUBLICATION Not apply. AVAILABILITY OF DATA AND MATERIALS The analysis code essential for the replication of study findings can be accessed at this link: https://github.com/psotob91/kfre-ckd-nationwide-essalud-peru. As per the privacy policies of EsSalud, the minimal data set is not open to public access. However, we are amenable to providing the anonymised data upon receipt of a reasonable request directed to the corresponding author ( [email protected] ). COMPETING INTERESTS JBZ, RCG, EPT, AVPV, and LCAG are full-time employees of EsSalud, serving as nephrologists at the Nephrology Department, Hospital Edgardo Rebagliati Martins, and the National Center of Renal Health in Lima, Peru. PSB is an associated researcher at the Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI), EsSalud, and has received consultancy fees from EsSalud. EJCP serves as an associated researcher and Deputy Manager at IETSI, EsSalud. DDO was a full-time employee of EsSalud during the initial phases of the study. The authors affirm that their respective affiliations with EsSalud have not influenced any aspect of the study, including its design, data collection, analysis, interpretation, or manuscript preparation. Furthermore, none of the authors have any financial or personal relationships that could inappropriately influence (bias) the work reported in this manuscript. The authors declare that there are no other competing interests or potential conflicts of interest related to the content of this study. FUNDING This study was funded by the Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI) of EsSalud. AUTHORS' CONTRIBUTIONS JBZ served as the principal investigators, conceived the study concept, developed the proposal, contributed to the study design, coordinated the project, and participated in drafting the manuscript. PSB served as the co-principal investigator, conceived the study concept, developed the proposal, contributed to the study design, coordinated the project, was responsible for cleaning the raw data, conducting the analysis, and preparing the manuscript. EC co-authored the proposal, contributed to the study design, oversaw data acquisition, was responsible for cleaning the raw data, and assisted in drafting the manuscript. RCG, DZDO, and LCAG co-authored the proposal and contributed to the study design, oversaw data acquisition, and assisted in drafting the manuscript. co-authored the proposal, contributed to the study design and oversaw data acquisition. JBZ, PSB, EC and LCAG serve as guarantors, ensuring the integrity of the study. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. ACKNOWLEDGEMENTS Not applicable. No specific acknowledgements are necessary for this study. CLINICAL TRIAL NUMBER Not applicable. This study is not a clinical trial. References Francis A, Harhay MN, Ong ACM, et al. Chronic kidney disease and the global public health agenda: an international consensus. Nat Rev Nephrol. 2024;20(7):473–85. 10.1038/s41581-024-00820-6 . Hill NR, Fatoba ST, Oke JL, et al. Global Prevalence of Chronic Kidney Disease - A Systematic Review and Meta-Analysis. PLoS ONE. 2016;11(7):e0158765. 10.1371/journal.pone.0158765 . Stevens PE, Ahmed SB, Carrero JJ, et al. KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int. 2024;105(4):S117–314. 10.1016/j.kint.2023.10.018 . Recommendations, August. 25, 2021. Accessed August 6, 2024. https://www.nice.org.uk/guidance/ng203/chapter/Recommendations#risk-assessment-referral-criteria-and-shared-care Bravo-Zúñiga J, Hinostoza-Sayas J, Goicochea Lugo S, et al. Guía de práctica clínica para el tamizaje, diagnóstico y manejo de la enfermedad renal crónica en estadios 1 al 3 en el Seguro Social del Perú (EsSalud). Acta Médica Peruana. 2020;37(4):518–31. 10.35663/amp.2020.374.1843 . Bravo-Zúñiga J, Garcia-Gomero D, Dolores-Maldonado G, et al. Guía de práctica clínica para el manejo de la Enfermedad Renal Crónica estadíos 3b, 4 y 5 en el Seguro Social de Salud del Perú (EsSalud). Revista del Cuerpo Médico Hosp Nac Almanzor Aguinaga Asenjo. 2023;16(1):102–11. 10.35434/rcmhnaaa.2023.161.1709 . Chronic Kidney Disease in Adults: Assessment and Management. National Institute for Health and Care Excellence (NICE). 2015. Accessed September 16, 2024. http://www.ncbi.nlm.nih.gov/books/NBK555204/ Inserra F, Torres ML, Alles A, et al. Especial atención sobre situaciones clínicas específicas. Acta Bioquímica Clínica Latinoam. 2022;56(1):43–74. Evaluación de la función renal para la detección y seguimiento de la enfermedad renal crónica. Documento multidisciplinario de consenso 2021. Collins GS, Dhiman P, Ma J, et al. Evaluation of clinical prediction models (part 1): from development to external validation. BMJ. 2024;384:e074819. 10.1136/bmj-2023-074819 . Riley RD, Archer L, Snell KIE, et al. Evaluation of clinical prediction models (part 2): how to undertake an external validation study. BMJ. 2024;384:e074820. 10.1136/bmj-2023-074820 . Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD Statement. BMC Med. 2015;13(1):1. 10.1186/s12916-014-0241-z . Thanabalasingam SJ, Iliescu EA, Norman PA, et al. Independent External Validation and Comparison of Death and Kidney Replacement Therapy Prediction Models in Advanced CKD. Kidney Med. 2022;4(4):100440. 10.1016/j.xkme.2022.100440 . Hundemer GL, Tangri N, Sood MM, et al. Performance of the Kidney Failure Risk Equation by Disease Etiology in Advanced CKD. Clin J Am Soc Nephrol. 2020;15(10):1424–32. 10.2215/CJN.03940320 . Peeters MJ, van Zuilen AD, van den Brand JAJG, et al. Validation of the kidney failure risk equation in European CKD patients. Nephrol Dial Transpl. 2013;28(7):1773–9. 10.1093/ndt/gft063 . Whitlock RH, Chartier M, Komenda P, et al. Validation of the Kidney Failure Risk Equation in Manitoba. Can J Kidney Health Dis. 2017;4:2054358117705372. 10.1177/2054358117705372 . Ramspek CL, Evans M, Wanner C, et al. Kidney Failure Prediction Models: A Comprehensive External Validation Study in Patients with Advanced CKD. J Am Soc Nephrol. 2021;32(5):1174–86. 10.1681/ASN.2020071077 . Major RW, Shepherd D, Medcalf JF, Xu G, Gray LJ, Brunskill NJ. The Kidney Failure Risk Equation for prediction of end stage renal disease in UK primary care: An external validation and clinical impact projection cohort study. PLoS Med. 2019;16(11):e1002955. 10.1371/journal.pmed.1002955 . Marks A, Fluck N, Prescott GJ, et al. Looking to the future: predicting renal replacement outcomes in a large community cohort with chronic kidney disease. Nephrol Dialysis Transplantation. 2015;30(9):1507–17. 10.1093/ndt/gfv089 . Kwek JL, Pang HQJ, Li H, et al. 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. Singap Med J. 2022;63(6):313–8. 10.11622/smedj.2020170 . Kang MW, Tangri N, Kim YC, et al. An independent validation of the kidney failure risk equation in an Asian population. Sci Rep. 2020;10(1):12920. 10.1038/s41598-020-69715-3 . Pan L, Wang J, Deng Y, et al. External Validation of the Kidney Failure Risk Equation Among Urban Community-Based Chinese Patients With CKD. Kidney Med. 2024;6(5):100817. 10.1016/j.xkme.2024.100817 . Maher F, Teece L, Major RW, et al. Using the kidney failure risk equation to predict end-stage kidney disease in CKD patients of South Asian ethnicity: an external validation study. Diagn Progn Res. 2023;7(1):22. 10.1186/s41512-023-00157-x . Wang Y, Nguyen FNHL, Allen JC, Lew JQL, Tan NC, Jafar TH. Validation of the kidney failure risk equation for end-stage kidney disease in Southeast Asia. BMC Nephrol. 2019;20(1):451. 10.1186/s12882-019-1643-0 . Jahan S, Hale J, Malacova E, Hurst C, Kark A, Mallett A. Real world evaluation of kidney failure risk equations in predicting progression from chronic kidney disease to kidney failure in an Australian cohort. J Nephrol. 2024;37(1):231–7. 10.1007/s40620-023-01680-2 . Tangri N, Grams ME, Levey AS, et al. Multinational Assessment of Accuracy of Equations for Predicting Risk of Kidney Failure: A Meta-analysis. JAMA. 2016;315(2):164–74. 10.1001/jama.2015.18202 . Gallego-Valcarce E, Shabaka A, Tato-Ribera AM, et al. External validation of the KFRE and Grams prediction models for kidney failure and death in a Spanish cohort of patients with advanced chronic kidney disease. J Nephrol. 2024;37(2):429–37. 10.1007/s40620-023-01819-1 . Lennartz CS, Pickering JW, Seiler-Mußler S, et al. External Validation of the Kidney Failure Risk Equation and Re-Calibration with Addition of Ultrasound Parameters. Clin J Am Soc Nephrol. 2016;11(4):609–15. 10.2215/CJN.08110715 . da Silva BM, Charreu J, Duarte I, Outerelo C, Gameiro J. Validation of the kidney failure risk equation in a Portuguese cohort. Nefrologia (Engl Ed). 2023;43(4):467–73. 10.1016/j.nefroe.2022.03.007 . Ali I, Donne RL, Kalra PA. A validation study of the kidney failure risk equation in advanced chronic kidney disease according to disease aetiology with evaluation of discrimination, calibration and clinical utility. BMC Nephrol. 2021;22(1):194. 10.1186/s12882-021-02402-1 . Irish GL, Cuthbertson L, Kitsos A, Saunder T, Clayton PA, Jose MD. The kidney failure risk equation predicts kidney failure: Validation in an Australian cohort. Nephrol (Carlton). 2023;28(6):328–35. 10.1111/nep.14160 . Bravo-Zúñiga J, Chávez-Gómez R, Soto-Becerra P. 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. BMJ Open. 2024;14(1):e076217. 10.1136/bmjopen-2023-076217 . Larrarte C, Vesga J, Ardila F, Aldana A, Perea D, Sanabria M. Validation of the Kidney Failure Risk Equation in the Colombian Population. Int J Nephrol. 2024;2024:1282664. 10.1155/2024/1282664 . González D, Sola L, Rios P, WCN24-1069 VALIDATION OF THE KIDNEY FAILURE RISK EQUATION (KFRE) AND ITS APPLICATION IN PATIENTS WITH CHRONIC KIDNEY DISEASE FOLLOWED BY THE RENAL HEALTH-CARE PROGRAM OF URUGUAY, et al. Kidney Int Rep. 2024;9(4):S86–7. 10.1016/j.ekir.2024.02.184 . Moons KGM, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162(1):W1–73. 10.7326/M14-0698 . Steyerberg EW. Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating. 2nd ed. Springer; 2020. Harrell FE. Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis . Second edition. Springer; 2015. Austin PC. Multiple imputation with competing risk outcomes. Comput Stat . Published online June. 2024;26. 10.1007/s00180-024-01518-w . van Geloven N, Giardiello D, Bonneville EF, et al. Validation of prediction models in the presence of competing risks: a guide through modern methods. BMJ. 2022;377:e069249. 10.1136/bmj-2021-069249 . Ramspek CL, Teece L, Snell KIE, et al. Lessons learnt when accounting for competing events in the external validation of time-to-event prognostic models. Int J Epidemiol. 2022;51(2):615–25. 10.1093/ije/dyab256 . Vickers AJ, van Calster B, Steyerberg EW. A simple, step-by-step guide to interpreting decision curve analysis. Diagn Prognostic Res. 2019;3(1):18. 10.1186/s41512-019-0064-7 . van Buuren S. Flexible Imputation of Missing Data. CRC; 2012. Ooi YG, Sarvanandan T, Hee NKY, et al. Risk Prediction and Management of Chronic Kidney Disease in People Living with Type 2 Diabetes Mellitus. Diabetes Metab J. 2024;48(2):196–207. 10.4093/dmj.2023.0244 . Potok OA, Nguyen HA, Abdelmalek JA, Beben T, Woodell TB, Rifkin DE. Patients,’ Nephrologists,’ and Predicted Estimations of ESKD Risk Compared with 2-Year Incidence of ESKD. Clin J Am Soc Nephrol. 2019;14(2):206–12. 10.2215/CJN.07970718 . Grams ME, Sang Y, Ballew SH, et al. Predicting timing of clinical outcomes in patients with chronic kidney disease and severely decreased glomerular filtration rate. Kidney Int. 2018;93(6):1442–51. 10.1016/j.kint.2018.01.009 . Naranjo FS, Sang Y, Ballew SH, et al. Estimating Kidney Failure Risk Using Electronic Medical Records. Kidney360. 2021;2(3):415–24. 10.34067/KID.0005592020 . Bundy JD, Mills KT, Anderson AH, et al. Prediction of End-Stage Kidney Disease Using Estimated Glomerular Filtration Rate With and Without Race: A Prospective Cohort Study. Ann Intern Med. 2022;175(3):305–13. 10.7326/M21-2928 . Grams ME, Li L, Greene TH, Tin A, Sang Y, Kao WHL, et al. Estimating time to ESRD using kidney failure risk equations: results from the African American Study of Kidney Disease and Hypertension (AASK). Am J Kidney Dis Off J Natl Kidney Found. 2015;65:394–402. Ramspek CL, Jager KJ, Dekker FW, Zoccali C, van Diepen M. External validation of prognostic models: what, why, how, when and where? Clin. Kidney J. 2021;14:49–58. Additional Declarations Competing interest reported. JBZ, RCG, EPT, AVPV, and LCAG are full-time employees of EsSalud, serving as nephrologists at the Nephrology Department, Hospital Edgardo Rebagliati Martins, and the National Center of Renal Health in Lima, Peru. PSB is an associated researcher at the Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI), EsSalud, and has received consultancy fees from EsSalud. EJCP serves as an associated researcher and Deputy Manager at IETSI, EsSalud. DDO was a full-time employee of EsSalud during the initial phases of the study. The authors affirm that their respective affiliations with EsSalud have not influenced any aspect of the study, including its design, data collection, analysis, interpretation, or manuscript preparation. Furthermore, none of the authors have any financial or personal relationships that could inappropriately influence (bias) the work reported in this manuscript. The authors declare that there are no other competing interests or potential conflicts of interest related to the content of this study. Supplementary Files suppmaterialv20240908.docx completedTRIPODchecklist.pdf Cite Share Download PDF Status: Published Journal Publication published 05 Dec, 2025 Read the published version in BMC Nephrology → Version 1 posted Editorial decision: Revision requested 19 Jun, 2025 Reviews received at journal 16 Jun, 2025 Reviewers agreed at journal 16 Jun, 2025 Reviewers invited by journal 13 Jun, 2025 Submission checks completed at journal 11 Jun, 2025 First submitted to journal 09 Jun, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5520011","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":470831984,"identity":"c66286fd-89d8-4fea-a7b4-a4fb3163f1d1","order_by":0,"name":"Jessica Ivonne Bravo-Zúñiga","email":"","orcid":"","institution":"Hospital Nacional Edgardo Rebagliati Martins, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Jessica","middleName":"Ivonne","lastName":"Bravo-Zúñiga","suffix":""},{"id":470831985,"identity":"29ef5597-d341-4ba7-b6a8-ba419d3623bf","order_by":1,"name":"Percy Soto-Becerra","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIiWNgGAWjYHAD5gMQ+gCxGngY2BJI1sJjQJwWc/beh595GOrk7Nl7vkkX7mCQ47uRwPbgAx4tlj3HjaV5GA4b8/Cc3SY98wyDseSNBHbDGXi0GNxIY2PmYTiQ2CORu02at40hcQPQFqAheLTcfwbSUgfUkvMMpKUerOUPXlvYQFqYQVrYQFoSDEBa8HnfsieNWXKOAdAvZ44ZW89skzCceeZhu2EPHi3m7McYP7ypqJNjb29+eLuwzUae73jysQc/8DkMiJlgMcLMwCABpBjb8LkLrIURZiYzlGbDq2UUjIJRMApGHAAADAVDnVZowF8AAAAASUVORK5CYII=","orcid":"","institution":"Universidad Privada del Norte","correspondingAuthor":true,"prefix":"","firstName":"Percy","middleName":"","lastName":"Soto-Becerra","suffix":""},{"id":470831986,"identity":"be7d3c63-0232-449b-8acf-4dca5398ae05","order_by":2,"name":"Edgar Juan Coila-Paricahua","email":"","orcid":"","institution":"Instituto de Evaluación de Tecnologías en Salud e Investigación - IETSI, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Edgar","middleName":"Juan","lastName":"Coila-Paricahua","suffix":""},{"id":470831987,"identity":"e925c817-7c8d-4890-8f28-e87e1e067f25","order_by":3,"name":"Ricardo Chávez-Gómez","email":"","orcid":"","institution":"Hospital Nacional Edgardo Rebagliati Martins, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Ricardo","middleName":"","lastName":"Chávez-Gómez","suffix":""},{"id":470831988,"identity":"f790f228-dd40-48ee-aaaf-4d8cb37f2c09","order_by":4,"name":"Eduardo Pérez-Tejada","email":"","orcid":"","institution":"National Center of Renal Health, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Eduardo","middleName":"","lastName":"Pérez-Tejada","suffix":""},{"id":470831989,"identity":"19502680-b5bb-4b9e-aee0-91460a8c066d","order_by":5,"name":"Anselma Victoria Pardo-Villafranca","email":"","orcid":"","institution":"National Center of Renal Health, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Anselma","middleName":"Victoria","lastName":"Pardo-Villafranca","suffix":""},{"id":470831990,"identity":"b25dce35-214b-4581-ab1d-e9c0a1a9e5b5","order_by":6,"name":"Lizbeth Carmen Arce-Gallo","email":"","orcid":"","institution":"National Center of Renal Health, EsSalud","correspondingAuthor":false,"prefix":"","firstName":"Lizbeth","middleName":"Carmen","lastName":"Arce-Gallo","suffix":""},{"id":470831991,"identity":"24c03440-2160-4ade-8b5d-52d0b6d46570","order_by":7,"name":"Daysi Díaz-Obregon","email":"","orcid":"","institution":"Universidad Peruana Cayetano Heredia","correspondingAuthor":false,"prefix":"","firstName":"Daysi","middleName":"","lastName":"Díaz-Obregon","suffix":""}],"badges":[],"createdAt":"2024-11-25 11:53:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5520011/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5520011/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12882-025-04357-z","type":"published","date":"2025-12-05T15:57:22+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":84682033,"identity":"eb1e55ef-e922-4a34-9001-34e1af9fae87","added_by":"auto","created_at":"2025-06-16 08:30:55","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":994054,"visible":true,"origin":"","legend":"\u003cp\u003eRegional distribution from EsSalud’s patients included in the analysis at national level\u003c/p\u003e","description":"","filename":"Figure1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/757b4fa54dd085cbb874dfa5.jpeg"},{"id":84682034,"identity":"c781eac5-6c62-4e61-9f2d-53ce90cb3e82","added_by":"auto","created_at":"2025-06-16 08:30:55","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2458706,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of study participant inclusion\u003c/p\u003e","description":"","filename":"Figure2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/8eb7e5371305b5509fbb7fab.jpeg"},{"id":84680334,"identity":"4def4b94-d53c-40ab-b07c-56e84d1f2d8c","added_by":"auto","created_at":"2025-06-16 08:14:55","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":888098,"visible":true,"origin":"","legend":"\u003cp\u003eCumulative incidence function curves for Kidney Failure and Death in patients with CKD G3-4 included in the study\u003c/p\u003e","description":"","filename":"Figure3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/6e7eba4d43b227659effb387.jpeg"},{"id":84680337,"identity":"28a12c71-067f-49bf-9890-a58302bdf40e","added_by":"auto","created_at":"2025-06-16 08:14:55","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":741182,"visible":true,"origin":"","legend":"\u003cp\u003eCalibration curves of the original KFRE model. The predicted risk by the KFRE model is shown on the x-axis, and the observed risk on the y-axis. The observed risk was estimated using cumulative incidence function curves to account for the competing risk of death without kidney failure. CKD: chronic kidney disease\u003c/p\u003e","description":"","filename":"Figure4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/9613e479180223034eb28e1a.jpeg"},{"id":84680930,"identity":"49b597fb-6fea-45e2-bb83-4693ca38c9e3","added_by":"auto","created_at":"2025-06-16 08:22:56","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1347944,"visible":true,"origin":"","legend":"\u003cp\u003eCalibration plots for the recalibrated KFRE models showing observed outcome proportions against predicted risks. Recalibrated models using (A) method A at 2 years, (B) method A at 5 years, (C) method B at 2 years, (D) method B at 5 years, (E) method C at 2 years, (F) method C at 5 years, (G) method D at 2 years, and (H) method D at 5 years. The red dashed line represents the ideal calibration line where predicted risks perfectly match observed proportions. Blue points indicate the deciles of predicted risk, and the grey line represents a smoothed calibration curve.\u003c/p\u003e","description":"","filename":"Figure5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/bb42965455f329bdc34b33c7.jpeg"},{"id":84680346,"identity":"1ad62523-e9e2-4c39-947d-9c8721766e66","added_by":"auto","created_at":"2025-06-16 08:14:56","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1232119,"visible":true,"origin":"","legend":"\u003cp\u003eDecision curve analysis (DCA) for the original and recalibrated KFRE models, and alternative nephrology referral guidelines (NICE 2014 and Peruvian National Guidelines). Net benefit is plotted against the threshold probability for the (A) 2-year and (B\u003cstrong\u003e)\u003c/strong\u003e5-year horizon. The lines indicate the performance, in terms of net benefit, of different strategies: the original KFRE model, recalibrated KFRE models (Methods A, B, C, and D), Peruvian National Guidelines, NICE 2014 Guidelines, and the strategies of referring all or none. The net benefit values show how each strategy performs in balancing the correct identification of high-risk patients against minimising unnecessary referrals.\u003c/p\u003e","description":"","filename":"Figure6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/3303de71b64c5b59f9c0fb2d.jpeg"},{"id":97725083,"identity":"d0f62345-8468-446b-aeb0-375c102860f0","added_by":"auto","created_at":"2025-12-08 16:14:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7863301,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/3ee57ade-66c8-4792-b6ed-dad457945e01.pdf"},{"id":84680922,"identity":"07125f6b-98b5-4177-8ee4-bb272a1e03c8","added_by":"auto","created_at":"2025-06-16 08:22:55","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1417679,"visible":true,"origin":"","legend":"","description":"","filename":"suppmaterialv20240908.docx","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/badea80a4345063a961ce393.docx"},{"id":84682036,"identity":"2b70d108-85f1-417b-b392-247d94e5de1b","added_by":"auto","created_at":"2025-06-16 08:30:56","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":129602,"visible":true,"origin":"","legend":"","description":"","filename":"completedTRIPODchecklist.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5520011/v1/8c5e840dd9e22686dc44d32d.pdf"}],"financialInterests":"Competing interest reported. JBZ, RCG, EPT, AVPV, and LCAG are full-time employees of EsSalud, serving as nephrologists at the Nephrology Department, Hospital Edgardo Rebagliati Martins, and the National Center of Renal Health in Lima, Peru. PSB is an associated researcher at the Instituto de Evaluación de Tecnologías en Salud e Investigación (IETSI), EsSalud, and has received consultancy fees from EsSalud. EJCP serves as an associated researcher and Deputy Manager at IETSI, EsSalud. DDO was a full-time employee of EsSalud during the initial phases of the study. The authors affirm that their respective affiliations with EsSalud have not influenced any aspect of the study, including its design, data collection, analysis, interpretation, or manuscript preparation. Furthermore, none of the authors have any financial or personal relationships that could inappropriately influence (bias) the work reported in this manuscript. The authors declare that there are no other competing interests or potential conflicts of interest related to the content of this study.","formattedTitle":"External validation, recalibration, and clinical utility of the kidney failure risk equation in patients with advanced CKD: a nationwide retrospective cohort analysis in Peru","fulltext":[{"header":"BACKGROUND","content":"\u003cp\u003eChronic kidney disease (CKD) is a growing global health challenge, affecting about 10% of the population and often leading to kidney failure, requiring costly interventions like dialysis or transplantation\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Recent guidelines, including NICE 2021 and KDIGO 2024\u003csup\u003e3,4\u003c/sup\u003e, emphasize that accurate, individualized risk prediction models can help reduce the burden of CKD by identifying high-risk patients who may benefit from early interventions. This enables healthcare providers to make informed decisions about when to refer patients to nephrology, allowing for the implementation of nephroprotective strategies to slow disease progression and facilitating timely planning for renal replacement therapy (RRT), ultimately improving patient outcomes and optimizing healthcare resources.\u003c/p\u003e \u003cp\u003eWhile international guidelines recommend the use of the Kidney Failure Risk Equation (KFRE) for predicting kidney failure\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, its implementation in Peru remains limited\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. The Peruvian Clinical Practice Guideline instead relies on a combination of estimated glomerular filtration rate (eGFR) and albuminuria for nephrology referrals, similar to the earlier approach of NICE 2014 guidelines\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Despite endorsements from the Latin American Nephrology Society to recommend KFRE\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, concerns have persisted due to the lack of region-specific evidence supporting its utility. This highlights the importance of externally validating prognostic models like KFRE within the specific settings where they are intended to be used\u003csup\u003e\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e, particularly in the Peruvian and broader Latin American context.\u003c/p\u003e \u003cp\u003eAlthough KFRE has shown strong prognostic performance in international contexts\u003csup\u003e\u003cspan additionalcitationids=\"CR13 CR14 CR15 CR16 CR17 CR18 CR19 CR20 CR21 CR22 CR23 CR24 CR25 CR26 CR27\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, recent studies have identified miscalibration in various countries\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e,\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. In Peru, a recent study found KFRE to perform well in a cohort from Lima, the country's capital and largest urban center, but did not undertake recalibration, raising questions about its nationwide applicability\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Moreover, region-specific evidence in Latin America is scarce, with no studies evaluating the clinical utility of KFRE\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. Addressing this gap is essential to provide evidence that supports the use of KFRE in the Peruvian context and informs tailored CKD management strategies. Therefore, this study aims to perform an external validation and recalibration of the KFRE to predict kidney failure, as well as assess its clinical utility in guiding decisions for referral and planning for RRT in a nationwide cohort of patients with CKD stages G3-4 in Peru.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design, population and data source\u003c/h2\u003e \u003cp\u003eWe conducted a retrospective cohort study following the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) guidelines\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e to externally validate, recalibrate and assess the clinical utility of the KFRE model for predicting the risk of kidney failure at 2-year and 5-year horizons in patients diagnosed with chronic kidney disease (CKD) stages 3\u0026ndash;4 (eGFR\u0026thinsp;\u0026gt;\u0026thinsp;15 and \u0026lt;\u0026thinsp;60 ml/min per 1.73 m\u0026sup2; ) between 2013 and 2022 across all 45 care networks of EsSalud (see TRIPOD checklist in online \u003cb\u003eSupplementary Material\u003c/b\u003e). The data were collected and managed by the National Renal Health Center as part of the national renal health surveillance system (VISARE, by its Spanish acronym). For health centers within the Rebagliati Network, the largest in Lima, data were directly sourced from the UMERC informatics application designed specifically for this network to provide information to VISARE, as previously described\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Additionally, data from the Kaelin Hospital were provided directly by the hospital itself due to its special status as a public-private partnership, which maintains its own renal health surveillance and does not routinely report to the CNSR.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eValidation model and predictors\u003c/h3\u003e\n\u003cp\u003eWe validated the 4-variable KFRE, which includes age (scaled to 10 years), eGFR (ml/min per 1.73 m\u0026sup2; using the CKD-EPI formula), sex, and urinary albumin/creatinine ratio (ACR) in mg/g \u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Standardized laboratory protocols were followed across health facilities to measure these predictors (see \u003cb\u003eSupplementary Methods\u003c/b\u003e for details on KFRE equations, coding of predictors, CKD-EPI formula and laboratory considerations).\u003c/p\u003e\n\u003ch3\u003eOutcome variable\u003c/h3\u003e\n\u003cp\u003eThe primary outcome was the time to kidney failure, defined as the date of the first hemodialysis or peritoneal dialysis based on administrative data from the National Center for Renal Health or through specific ICD-10 diagnoses (e.g., N18.5, N18.6, Z99.2, Z49.1, Z49.2, Z94.0). This computational phenotyping approach has been validated in previous studies and is further detailed in the \u003cb\u003eSupplementary Methods\u003c/b\u003e. Death was considered a competing risk, with mortality data obtained from both the National Death System of Peru and the National Registry of Identification and Civil Status, which together cover more than 90% of all deaths in the country. For more information on the outcome definitions and data sources, please refer to the \u003cb\u003esupplementary material\u003c/b\u003e.\u003c/p\u003e\n\u003ch3\u003eFollow-up time\u003c/h3\u003e\n\u003cp\u003ePatients were followed until kidney failure, death or the administrative censorship date (December 31, 2022), whichever occurred first. Censoring occurred at loss to follow-up or study end.\u003c/p\u003e\n\u003ch3\u003eSample size\u003c/h3\u003e\n\u003cp\u003eDue to the comprehensive nature of the national dataset, a specific sample size calculation was unnecessary. All patients meeting inclusion criteria were analyzed (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Table S4). Since the number of events per EsSalud health network was significantly low\u0026mdash;often fewer than 100 events, and in many instances, nearly zero\u0026mdash;it was decided to evaluate the performance of KFRE across the entire country without considering regions as separate clusters.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eInitial data analyses were performed to identify extreme values, inconsistencies, and missing data. Winsorization of albumin-creatinine ratio (ACR) values at the 1.5th and 98.5th percentiles was conducted\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Missing data were managed using multiple imputation via Additive Regression, Bootstrapping, and Predictive Mean Matching, creating 100 imputed datasets\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e. To ensure that the imputation model was congenial (at least semicompatible) with the competing risk models used in the substantive analysis, interactions between all predictors and the cumulative baseline hazards for kidney failure and death were included in the imputation process\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e. This step aimed to improve the accuracy of the imputations and their compatibility with the intended analyses (see Supplementary Methods for a detailed description).\u003c/p\u003e \u003cp\u003eFor the external validation of the KFRE, model discrimination was assessed using the concordance index (C-index) at 2 and 5 years. In line with the TRIPOD guidelines, model performance was evaluated through both discrimination and calibration assessments, incorporating the consideration of competing risks based on recent methodological recommendations\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Calibration was evaluated through assessment of calibration in-the-large: observed-to-expected (O/E) ratios, difference and calibration intercepts; weak calibration: slopes calibration; and moderate calibration via calibration plots. All these measures were obtained accounting for competing risk\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. The calibration curves were obtained using LOESS smoothing applied to the cumulative baseline hazard predicted by the Fine-Gray model for subdistribution hazards\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. See Supplementary Methods for further details.\u003c/p\u003e \u003cp\u003eRecalibration was performed using four methods\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e: Methods A and B used the traditional Cox model without accounting for competing risks, as originally proposed in the KFRE model by Tangri et al\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Method A adjusted baseline risk only, while Method B adjusted both baseline risk and the magnitude of the linear predictor. Methods C and D, in contrast, employed the Cause-Specific Cox model to account for competing risks. Method C adjusted the baseline risk, and Method D combined baseline risk adjustment with linear predictor adjustment while considering competing risks.\u003c/p\u003e \u003cp\u003eClinical utility was assessed using decision curve analysis\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e by comparing the net benefit of the original and recalibrated KFRE models against the Peruvian National Guidelines\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e and NICE 2014 Guidelines for nephrology referral\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. For each prediction horizon (2 and 5 years), we evaluated the utility of using KFRE to guide decisions for referral and planning for renal replacement therapy. Predefined reasonable decision thresholds (based on existing literature) were used to identify patients who might benefit from referral. For long-term management and nephroprotection to halt or reverse CKD progression, a 5-year horizon was used, with thresholds typically ranging between 3\u0026ndash;5%\u003csup\u003e5,6\u003c/sup\u003e. For planning renal replacement therapy, the 2-year horizon was considered, with decision thresholds set at higher probabilities, usually around 20\u0026ndash;40%\u003csup\u003e5,6\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe final estimates and standard errors were pooled across the imputed datasets using Rubin's rules, and 95% confidence intervals were calculated based on these standard errors\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. All statistical analyses were performed using R version 4.3. The reproducible code used for this analysis is available in an open GitHub repository (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/psotob91/kfre-ckd-nationwide-essalud-peru\u003c/span\u003e\u003cspan address=\"https://github.com/psotob91/kfre-ckd-nationwide-essalud-peru\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eLanguage Editing Assistance\u003c/strong\u003e \u003cp\u003eThe authors utilized ChatGPT-4o, a Large Language Model (LLM) developed by OpenAI, to review and improve the English grammar and style of this manuscript. The AI tool was employed solely for language editing purposes and was not used to generate or create any content.\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eStudy Population\u003c/h2\u003e \u003cp\u003eOut of 152,084 patients screened between January 1, 2013, and December 30, 2022, in all EsSalud facilities nationwide under the VISARE program, 30,031 met the selection criteria for CKD stages G3-4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Only 38.4% (11,540) had complete data for the four variables required for KFRE estimation. After multiple imputations, all eligible individuals were included in the analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the key sociodemographic and clinical characteristics of the study population after imputation. Of the total, 56.4% were women, and ages ranged from 18 to 109 years, with a median age of 73.8 years. The prevalence of diabetes mellitus (41.5%) and hypertension (75.8%) was high, with most participants classified as stage G3a (61.6%) or G3b (27.3%). A detailed comparison of population characteristics with and without imputation is provided and according to outcome are showed in \u003cb\u003eTable S5\u003c/b\u003e and \u003cb\u003eS6\u003c/b\u003e, respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline Characteristics (at first recorded evaluation in VISARE) of patients with CKD G3-4 Included in the Analysis\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eN\u0026thinsp;=\u0026thinsp;30,031\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e13,097 (43.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16,934 (56.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAge (years)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e73.8 (11.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e75.0 (67.0\u0026ndash;82.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e18.0\u0026ndash;109.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eEsSalud Network\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetropolitan Lima\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14,784 (49.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther Regions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15,247 (50.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHypertension\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7,254 (24.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e22,777 (75.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiabetes Mellitus\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e17,567 (58.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e12,464 (41.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePersistent Albuminuria Categories\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eA1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14, (48.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eA2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,095 (33.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eA3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5,274 (17.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eeGFR Categories\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eG3a\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e18,491 (61.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eG3b\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8,201 (27.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eG4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3,339 (11.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCKD KDIGO Classification\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0 (0.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModerately increased risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9,403 (31.3%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,169 (33.9%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVery high risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10,459 (34.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eSerum Creatinine (mg/dL)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.4 (0.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.3 (1.2\u0026ndash;1.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9\u0026ndash;4.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eeGFR using CKD-EPI (ml/min/1.73m\u0026sup2;)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e45.9 (10.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e48.5 (39.4\u0026ndash;54.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15.0\u0026ndash;60.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAlbumin-Creatinine Ratio (mg/g)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e802.1 (3,534.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e32.0 (8.1\u0026ndash;160.3)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.6\u0026ndash;27,817.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eUrine Albumin (mg/dl)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e35.8 (161.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.6 (0.4\u0026ndash;8.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.0\u0026ndash;1,365.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eUrine Creatinine (mg/dL)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e60.8 (50.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian (Q1 - Q3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e49.2 (27.0\u0026ndash;85.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin - Max\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.1\u0026ndash;221.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDeath at 2 years*\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e27,640 (92.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2,391 (8.0%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOutcomes at 2 years\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlive w/o Kidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e27,227 (90.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e793 (2.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeath w/o Kidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2,011 (6.7%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDeath at 5 years*\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24,261 (80.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5,770 (19.2%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eOutcomes at 5 years\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlive w/o Kidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e23,579 (78.5%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,308 (4.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDeath w/o Kidney Failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5,144 (17.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003e*Death after o before kidney failure.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003eSD: standard deviation, IQR: first quartile and third quartile, ACR, urine albumin to creatinine ratio; CKD, chronic kidney disease; eGFR, glomerular filtration rate estimated by CKD Epidemiology Collaboration formula\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003c/p\u003e \u003cp\u003eCumulative incidences of kidney failure at 2 and 5 years were 2.73% (95% CI: 2.55%-2.92%) and 4.76% (95% CI: 4.51%-5.02%), respectively (\u003cb\u003eTable \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e\u003c/b\u003e). The cumulative incidences of death without kidney failure were notably higher, with values of 6.96% (95% CI: 6.67%-7.26%) at 2 years and 19.71% (95% CI: 19.22%-20.2%) at 5 years. The detailed cumulative incidence curve for both kidney failure and death are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. 2-year and 5-year KFRE\u0026rsquo;s predicted risk is shown in \u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e. The distribution of four KFRE equation variables is shown in \u003cb\u003eFigure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003ePredictive Performance of the Original 4-Variable KFRE Equation\u003c/h2\u003e \u003cp\u003eKFRE showed good discrimination across all time horizons, with C-indices of 0.88 (95% CI: 0.86\u0026ndash;0.89) at 2 years and 0.85 (95% CI: 0.84\u0026ndash;0.87) at 5 years (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). However, calibration was poor across all time horizons.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eExternal Validation Metrics for Predictive Performance of the Original 4-Variable KFRE Model\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eValidation aspect and performance measure\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eTime horizon\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003et\u0026thinsp;=\u0026thinsp;2 years\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003et\u0026thinsp;=\u0026thinsp;5 years\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage predicted risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.48%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.48%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage observed risk (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.73% (2.54\u0026ndash;2.92%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.76% (4.51\u0026ndash;5.02%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO/E ratio (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.84 (1.7 to 1.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.06 (1 to 1.13)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO-E difference (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.25% (1.05\u0026ndash;1.44%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.29% (0\u0026ndash;0.58%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration intercept (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.02 (-0.13 to 0.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.47 (-0.59 to -0.35)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration slope (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.58 (0.53 to 0.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.58 (0.54 to 0.62)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiscrimination\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC-index up to t years (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.88 (0.86 to 0.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.85 (0.84 to 0.87)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e%, percentage; C-index, truncated agreement index; CKD, chronic kidney disease; O/E and O-E, observed vs expected ratio and differences, respectively; t, time\u003c/p\u003e \u003cp\u003eAt 2 years, the model underestimated the risk of kidney failure, as indicated by an observed-to-expected (O/E) ratio of 1.84 (95% CI: 1.7\u0026ndash;1.99). Although the calibration intercept was not statistically different from zero (0.02, 95% CI: -0.13 to 0.18), the pronounced O/E ratio suggests that the model underestimates the actual risk on average. At 5 years, the O/E ratio was closer to 1 at 1.06 (95% CI: 1 to 1.13), which might initially suggest an alignment between predicted and observed risks. However, the calibration intercept was significantly negative (-0.47, 95% CI: -0.59 to -0.35), reinforcing the presence of systematic underestimation of the mean risk.\u003c/p\u003e \u003cp\u003eIn both time horizons, the calibration slope was below 1 (0.58 for both 2 and 5 years), indicating overly extreme predictions \u0026mdash; underestimating risks for low-risk individuals and overestimating for high-risk individuals. Calibration curves in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e further illustrate these trends. Overall, these results highlight that KFRE has poor mean, weak and moderate calibration in Peruvian population of EsSalud.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eRecalibration of the KFRE Model\u003c/h2\u003e \u003cp\u003eRecalibration was performed using four methods (A-D). Recalibrated equations has shown in Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eOriginal and recalibrated equations\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTime horizon\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEquations\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eOriginal model\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9832}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9365}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMethod A: Baseline risk adjustment without considering competing risk\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9688}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9363}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMethod B: Baseline risk adjustment\u0026thinsp;+\u0026thinsp;adjustment of linear predictor magnitude without considering competing risk\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9550}^{{e}^{0.7126042\\times\\:(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9130}^{{e}^{0.7126042\\times\\:(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMethod C: Baseline risk adjustment considering competing risk\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9699}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9425}^{{e}^{(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMethod D: Baseline risk adjustment\u0026thinsp;+\u0026thinsp;adjustment of linear predictor magnitude considering competing risk\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9572}^{{e}^{0.7126042\\times\\:(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5 years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:1-{0.9240}^{{e}^{0.7126042\\times\\:(-0.2201\\times\\:\\left(\\frac{age}{10}-7.036\\right)+0.2467\\times\\:\\left(sex-0.5642\\right)-0.5567\\times\\:\\left(\\frac{eGFR}{5}-7.222\\right)+0.4510\\times\\:\\left(\\text{log}\\left(ACR\\right)-5.137\\right))}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eeGFR, estimated glomerular filtration rate; urine albumin/creatinine ratio (ACR)\u003c/p\u003e \u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, Method D had the best calibration-in-the-large at both 2 and 5 years, with O/E ratios close to 1 (1.02 at 2 years; 1.04 at 5 years) and calibration intercepts not significantly different from zero. Despite this, the calibration slope (0.81 for both time horizons) indicates some underestimation in low-risk individuals and overestimation in high-risk ones, failing in achieve good weak calibration.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eValidation metrics for predictive performance of recalibrated KFRE models\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValidation aspect and performance measure\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMethod A\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMethod B\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMethod C\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMethod D\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003et\u0026thinsp;=\u0026thinsp;2 years\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage predicted risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.52%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.82%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.45%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.69%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage observed risk (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.73% (2.54\u0026ndash;2.92%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.73% (2.54\u0026ndash;2.92%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.73% (2.54\u0026ndash;2.92%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.73% (2.54\u0026ndash;2.92%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO/E ratio (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.08 (1.01 to 1.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.97 (0.9 to 1.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.11 (1.04 to 1.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.02 (0.95 to 1.09)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO-E difference (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.21% (0.02\u0026ndash;0.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.09% (-0.28\u0026ndash;0.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.28% (0.09\u0026ndash;0.47%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.04% (-0.15\u0026ndash;0.23%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration intercept (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.6 (-0.74 to -0.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.1 (-0.2 to 0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.57 (-0.71 to -0.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.05 (-0.15 to 0.05)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration slope (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.58 (0.53 to 0.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.82 (0.75 to 0.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.58 (0.53 to 0.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.82 (0.75 to 0.88)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiscrimination\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC-index up to t years (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.88 (0.86 to 0.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.88 (0.86 to 0.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.88 (0.86 to 0.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.88 (0.86 to 0.89)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003et\u0026thinsp;=\u0026thinsp;5 years\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCalibration\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage predicted risk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.48%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.19%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.14%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.59%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage observed risk (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.76% (4.51\u0026ndash;5.02%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.76% (4.51\u0026ndash;5.02%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.76% (4.51\u0026ndash;5.02%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.76% (4.51\u0026ndash;5.02%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO/E ratio (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.06 (1.01 to 1.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.92 (0.87 to 0.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.15 (1.09 to 1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.04 (0.98 to 1.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eO-E difference (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.28% (0.02\u0026ndash;0.54%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.42% (-0.68% to -0.17%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.62% (0.37\u0026ndash;0.88%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.18% (-0.08\u0026ndash;0.43%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration intercept (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.47 (-0.57 to -0.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.16 (-0.23 to -0.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0.36 (-0.47 to -0.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-0.02 (-0.09 to 0.06)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalibration slope (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.58 (0.54 to 0.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.81 (0.76 to 0.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.58 (0.54 to 0.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.81 (0.76 to 0.86)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDiscrimination\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC-index up to t years (95% CI)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.85 (0.84 to 0.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.85 (0.84 to 0.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.85 (0.84 to 0.87)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.85 (0.84 to 0.87)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e%, percentage; C-index, truncated agreement index; CKD, chronic kidney disease; O/E and O-E, observed vs expected ratio and differences, respectively; time, time\u003c/p\u003e \u003cp\u003e \u003csup\u003e \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e \u003c/sup\u003eMethod A: Baseline risk adjustment without considering competing risk\u003c/p\u003e \u003cp\u003e \u003csup\u003e \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e \u003c/sup\u003eMethod B: Baseline risk adjustment\u0026thinsp;+\u0026thinsp;adjustment of linear predictor magnitude without considering competing risk \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003eMethod C: Baseline risk adjustment considering competing risk\u003c/p\u003e \u003cp\u003e \u003csup\u003e \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e \u003c/sup\u003eMethod D: Baseline risk adjustment\u0026thinsp;+\u0026thinsp;adjustment of linear predictor magnitude considering competing risk\u003c/p\u003e \u003cp\u003eThe calibration plots for Method D show a reasonable alignment with the ideal 45-degree line, with minor overestimation at higher risk levels (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). It's important to note that this overestimation at the upper tail involves fewer data points, which may impact the curve's stability. In contrast, Methods A, B, and C show more pronounced deviations, especially at the extremes. Thus, while Method D is not perfectly calibrated, it offers the most balanced performance among the recalibrated models.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eClinical utility of original model KFRE and recalibrated versions\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents the decision curve analysis comparing the net benefits of the original and recalibrated KFRE models across 2-year and 5-year horizons. Method D demonstrated the highest net benefit across most thresholds; however, the difference was modest compared to the original KFRE model. For example, at a 5% threshold over 5 years, Method D had a net benefit of 0.0261, nearly identical to the original model's 0.0255.\u003c/p\u003e \u003cp\u003e All models, including the original KFRE and its recalibrated versions, outperformed alternative referral strategies based on the NICE 2014 guidelines and Peruvian National Guidelines across a range of thresholds. This suggests that the KFRE models, despite calibration issues, effectively balance harm and benefit by accurately identifying more patients for referral while minimizing unnecessary ones.\u003c/p\u003e \u003cp\u003eAt the 2-year horizon, the net benefit of both the original KFRE and the recalibrated models was slightly superior to the \"refer none\" strategy at threshold probabilities of 20% and 30%. For instance, at a 20% threshold over 2 years, Method D had a net benefit of 0.0016, while the original model had a net benefit of 0.0015. However, at higher threshold probabilities (\u0026gt;\u0026thinsp;40%), the net benefit of these models became lower than the \"refer none\" strategy, indicating limited utility in these scenarios.\u003c/p\u003e \u003cp\u003eIn summary, Methods D and B showed the highest clinical utility, consistently demonstrating the highest net benefits at both 2- and 5-year horizons. However, these differences were relatively small, indicating that the original KFRE model remains a viable option for clinical decision-making.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSensitivity Analysis\u003c/h2\u003e \u003cp\u003eA sensitivity analysis was performed without applying winsorization to the extreme values of the ACR. \u003cb\u003eFigure S3\u003c/b\u003e and \u003cb\u003eS4\u003c/b\u003e show the distribution of KFRE\u0026rsquo;s predicted risks and four variables, respectively in the original dataset without winsorization of ACR. The results remained similar to the primary analysis (\u003cb\u003eTable S9\u003c/b\u003e, \u003cb\u003eFigures S5\u003c/b\u003e, \u003cb\u003eS6\u003c/b\u003e and \u003cb\u003eS7\u003c/b\u003e), indicating that the winsorization of ACR values did not significantly impact the predictive performance or calibration of the KFRE model.\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eMain findings\u003c/h2\u003e \u003cp\u003eThis study externally validated the KFRE multivariable model in a national cohort of EsSalud patients with CKD stages 3\u0026ndash;4, recalibrated KFRE for this population, and assessed the clinical utility of the original and recalibrated versions. While KFRE showed strong discrimination for predicting kidney failure at both 2 and 5 years, it exhibited poor calibration. Two recalibrated models (Methods B and D) improved calibration-in-the-large, yet all recalibrated models struggled with weak or moderate calibration. The recalibrated models maintained a similar pattern to the original KFRE, overestimating risk for high-risk individuals and underestimating it for low-risk individuals.\u003c/p\u003e \u003cp\u003eDespite these calibration issues, the original and recalibrated KFRE versions offered a net benefit compared to the strategy of referring no patients. Their net benefit also surpassed that of the Peruvian National Guidelines and NICE 2014 guidelines across various thresholds. Thus, while the original KFRE and its recalibrated versions may not perfectly predict individual risk, they remain useful tools for guiding nephrology referrals in Peru.\u003c/p\u003e \u003cp\u003eWhen assessing early referral for renal replacement therapy preparation, the net benefit of KFRE at 2 years turned negative at higher thresholds, indicating that unnecessary referrals (false positives) may outweigh correct ones (true positives). However, for decisions involving long-term referral at 5 years, the KFRE models showed a positive net benefit across thresholds of 3\u0026ndash;10%, indicating their utility in effectively identifying high-risk patients who would benefit from early nephrology intervention.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eComparison with previous literature\u003c/h2\u003e \u003cp\u003eThe 4-variable KFRE model has been externally validated in over 30 countries across all continents\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e: North America\u003csup\u003e\u003cspan additionalcitationids=\"CR13 CR14\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan additionalcitationids=\"CR44 CR45\" citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, Europe\u003csup\u003e\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, Asia\u003csup\u003e\u003cspan additionalcitationids=\"CR20 CR21 CR22\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, Oceania\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, and more recently, Latin America\u003csup\u003e\u003cspan additionalcitationids=\"CR32\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. However, apart from the cohorts from Chile and Brazil in the 1990s used to recalibrate the initial KFRE model and derive a specific equation for non-North American countries\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, evidence on external validity in Latin America was absent for nearly two decades. Previous research in Peru showed KFRE's good discrimination but highlighted poor calibration, a finding we confirmed in a larger national cohort\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. Other studies in Latin America, such as those in Colombia and Uruguay, reported good discrimination and calibration but lacked adequate validation methodologies, limiting the comparability of their results with our findings\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. These methodological issues underscore the urgent need to generate high-quality evidence on the applicability of KFRE in clinical practice in Latin America, ensuring that well-sound methods are used to properly validate prognostic models.\u003c/p\u003e \u003cp\u003eIn contrast, the literature from non-North American regions, excluding Latin America, is more extensive. Globally, studies consistently demonstrate KFRE's high discrimination (\u0026gt;\u0026thinsp;0.80)\u003csup\u003e12\u0026ndash;28\u003c/sup\u003e, but have identified moderate calibration issues, particularly with overprediction in high-risk groups. Differences in predictor profiles and incidence rates do not fully account for the poor calibration observed in Peru. For example, the renal failure incidence rate in our study closely matches that of the original model (see \u003cb\u003eTable S8\u003c/b\u003e). However, our population's characteristics, including higher albuminuria levels, higher diabetes prevalence, and exclusion of G5 stages, may contribute to the observed miscalibration.\u003c/p\u003e \u003cp\u003eImportantly, our decision to exclude stage G5 patients is well supported by both the intended clinical use of the KFRE model and prevailing practices in the literature. The KFRE is primarily designed for use in outpatient settings where earlier stages of CKD (G3\u0026ndash;4) are managed, and its application in stage G5 is inherently less useful, as these patients are already recognized as high risk and often follow different management pathways (e.g., conservative or palliative care). Notably, Tangri\u0026rsquo;s original derivation and validation cohorts included a very low proportion of G5 patients (approximately 5.3%)\u003csup\u003e25\u003c/sup\u003e, and several subsequent external validation studies have similarly focused on patients with CKD stages G3\u0026ndash;4\u003csup\u003e27,47\u003c/sup\u003e. This selective validation approach is common, valid, and reasonable, as it ensures that the model\u0026rsquo;s calibration and performance are most relevant to the population where it is intended to be applied\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIt is also important to note that although one might argue that the exclusion of stage G5 patients could explain part of the miscalibration, evidence from studies such as Ramspek et al.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e indicates that calibration issues persist even when G5 patients are included. Ramspek et al.\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e attribute these issues not solely to the inclusion of patients with advanced CKD, but to a combination of factors\u0026mdash;including heterogeneity in clinical profiles and the failure to account for the competing risk of death. In fact, Tangri's original model did not account for this competing risk, leading to an overprediction of renal failure risk in settings where the incidence of death is high\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. In our population, incidence of pre-dialysis death is 19.7% at five years, nearly four times the incidence of renal failure for the same period, making it likely that the Cox model would overestimate renal failure in these high-risk group. This explanation aligns with our observations and further supports our decision to exclude stage G5 patients, thereby validating the model in the specific outpatient context where it is intended to be applied.\u003c/p\u003e \u003cp\u003eA key finding of this study is that the KFRE model retains clinical utility despite its miscalibration. This can be understood through the net benefit framework, which emphasizes that a model may still be valuable if the miscalibrations occur infrequently or in less critical subpopulations. In this study, the overall miscalibration was most pronounced in patients at the extremes of risk, who represent a smaller fraction of the population. Most patients had reasonably calibrated risk predictions, contributing to a higher net benefit in clinical decision-making. Therefore, even with calibration issues, the KFRE model remains a robust tool for predicting kidney failure and guiding appropriate referrals, particularly in resource-limited settings where optimizing healthcare allocation is crucial. Importantly, this study is the first in Latin America to rigorously evaluate the clinical utility of the KFRE model, highlighting its potential role in enhancing CKD management strategies in the region.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eStrengths and limitations of this study\u003c/h2\u003e \u003cp\u003eA major strength of this study is its large sample size of 30,031 CKD patients, providing robust statistical power. Additionally, the use of VISARE surveillance data captures patients at primary healthcare centers nationwide, who are often identified through screening programs targeting individuals with diabetes, hypertension, or those over 55 years of age. This real-world data from a nationwide screening program at the primary care level reflects a population with characteristics that are both relevant and underrepresented in the literature, providing valuable insights into CKD management.\u003c/p\u003e \u003cp\u003eHowever, we excluded 22,627 patients due to missing eGFR and other critical variables, with most missing data attributed to albumin-to-creatinine ratio (ACR) measurements (60.7%). This raises concerns about the usability of KFRE in settings where ACR testing is limited, emphasizing the need to strengthen laboratory capacities, particularly outside of Lima, the capital of the country. In many regions, the lack of ACR data is primarily due to shortages of supplies, which limits the full implementation of risk prediction models like KFRE. To address missing data in our study, multiple imputation was performed under the assumption that data were missing at random (MAR). We included auxiliary variables such as age, sex, comorbidities, and lab values to make the MAR assumption more plausible and enhance the accuracy of the imputations. Nevertheless, the validity of this assumption cannot be guaranteed, which remains a limitation.\u003c/p\u003e \u003cp\u003eThe geographical bias, with data predominantly from Lima, also underscores the need for further research in less-represented regions. Measurement errors in laboratory data, particularly outside of Lima, further highlight the necessity for improved healthcare infrastructure, including consistent access to ACR testing, to better support CKD risk prediction and management.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eImplications for clinical practice and health systems\u003c/h2\u003e \u003cp\u003eThe KFRE model's superior net benefit for predicting kidney failure at 5 years offers a key opportunity for early referral for secondary prevention, allowing nephrology interventions to slow disease progression. For 2-year predictions, the model supports planning for renal replacement therapy or conservative treatment. Recalibrated KFRE versions outperformed current national referral strategies and NICE 2014 guidelines across various threshold probabilities. This aligns with NICE 2024's updated recommendation of a KFRE score\u0026thinsp;\u0026gt;\u0026thinsp;5% or an ACR\u0026thinsp;\u0026gt;\u0026thinsp;70 for nephrology referral.\u003c/p\u003e \u003cp\u003eIn settings with limited nephrology resources, the KFRE's ability to identify high-risk patients can optimize referrals and prevent unnecessary ones. The use of a 3\u0026ndash;5% risk threshold over 5 years has shown to be effective in various healthcare settings. Retrospective studies in Canada and the UK found that these thresholds reduced late referrals for patients progressing to kidney failure. Similarly, prospective evaluations noted shorter nephrology wait times, particularly for high-risk individuals.\u003c/p\u003e \u003cp\u003eHowever, the model's utility at 2 years requires caution. Net benefit analysis suggests an advantage at thresholds of 20\u0026ndash;30%, but this diminishes at thresholds above 35\u0026ndash;40%, where unnecessary referrals outweigh the benefits. For patients with higher risk probabilities, additional tests may be required to avoid premature dialysis preparations. Lower thresholds like \u0026gt;\u0026thinsp;20% can help optimize sensitivity, aiding in early dialysis planning or transplant referral.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eFuture research\u003c/h2\u003e \u003cp\u003eImproving the KFRE model's calibration, especially for 2-year predictions, may require substantial updates, including re-estimating regression coefficients or modifying predictors. While promising, this approach risks overfitting and model instability. Therefore, using the current KFRE model remains more practical despite its limitations. Future studies should explore KFRE's predictive performance in subgroups like children, young adults, and those with diabetes, as emphasized by NICE 2024 guidelines. Assessing simplified versions of the model, such as the 3-variable KFRE or proxies like urine dipsticks, could also enhance its clinical applicability.\u003c/p\u003e \u003cp\u003eIn addition, cost-effectiveness analyses should be conducted to evaluate the economic impact of implementing KFRE in clinical practice. Implementation studies examining the integration of KFRE into clinical workflows and trials that evaluate its impact on patient outcomes are needed to further establish its role in diverse healthcare settings.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003e Despite calibration issues, the KFRE model, especially in its recalibrated forms, remains a valuable tool for guiding nephrology referrals in Peru, providing a higher net benefit than current national guidelines. Its use can enhance early identification of high-risk patients, improving healthcare resource allocation in resource-limited settings. Further research is needed to refine the model, explore its utility in diverse subgroups, and integrate it effectively into clinical practice.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eACR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eAlbumin-to-Creatinine Ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eC-index\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eConcordance Index\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCKD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eChronic Kidney Disease\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCKD-EPI\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eChronic Kidney Disease Epidemiology Collaboration\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCNSR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNational Renal Health Center\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDCA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eDecision Curve Analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eeGFR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eEstimated Glomerular Filtration Rate\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eEsSalud\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eSocial Health Insurance of Peru\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eICD-10\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eInternational Classification of Diseases, 10th Revision\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKFRE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKidney Failure Risk Equation\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKDIGO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKidney Disease:Improving Global Outcomes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMAR\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMissing At Random\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eNICE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eNational Institute for Health and Care Excellence\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eO/E ratio\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eObserved-to-Expected Ratio\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eRRT\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRenal Replacement Therapy\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eTRIPOD\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eTransparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eUMERC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eInformatics application used to provide information to VISARE\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eVISARE\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eRenal Health Surveillance\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eETHICS APPROVAL AND CONSENT TO PARTICIPATE\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Research Ethics Committee of the Edgardo Rebagliati National Hospital (Code No. 519-GRPR-ESSALUD-2023) and was conducted in accordance with the principles of the Declaration of Helsinki. A waiver of informed consent was granted by the committee because the study involved routinely collected secondary data. Given the minimal risks associated with this type of research, this waiver was deemed a reasonable justification and was approved accordingly.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCONSENT FOR PUBLICATION\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot apply.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAVAILABILITY OF DATA AND MATERIALS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe analysis code essential for the replication of study findings can be accessed at this link: https://github.com/psotob91/kfre-ckd-nationwide-essalud-peru. As per the privacy policies of EsSalud, the minimal data set is not open to public access. However, we are amenable to providing the anonymised data upon receipt of a reasonable request directed to the corresponding author ([email protected]).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCOMPETING INTERESTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJBZ, RCG, EPT, AVPV, and LCAG are full-time employees of EsSalud, serving as nephrologists at the Nephrology Department, Hospital Edgardo Rebagliati Martins, and the National Center of Renal Health in Lima, Peru. PSB is an associated researcher at the Instituto de Evaluaci\u0026oacute;n de Tecnolog\u0026iacute;as en Salud e Investigaci\u0026oacute;n (IETSI), EsSalud, and has received consultancy fees from EsSalud. EJCP serves as an associated researcher and Deputy Manager at IETSI, EsSalud. DDO was a full-time employee of EsSalud during the initial phases of the study. The authors affirm that their respective affiliations with EsSalud have not influenced any aspect of the study, including its design, data collection, analysis, interpretation, or manuscript preparation. Furthermore, none of the authors have any financial or personal relationships that could inappropriately influence (bias) the work reported in this manuscript. The authors declare that there are no other competing interests or potential conflicts of interest related to the content of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFUNDING\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was funded by the Instituto de Evaluaci\u0026oacute;n de Tecnolog\u0026iacute;as en Salud e Investigaci\u0026oacute;n (IETSI) of EsSalud.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAUTHORS\u0026apos; CONTRIBUTIONS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJBZ served as the principal investigators, conceived the study concept, developed the proposal, contributed to the study design, coordinated the project, and participated in drafting the manuscript. PSB served as the co-principal investigator, conceived the study concept, developed the proposal, contributed to the study design, coordinated the project, was responsible for cleaning the raw data, conducting the analysis, and preparing the manuscript. EC co-authored the proposal, contributed to the study design, oversaw data acquisition, was responsible for cleaning the raw data, and assisted in drafting the manuscript. RCG, DZDO, and LCAG co-authored the proposal and contributed to the study design, oversaw data acquisition, and assisted in drafting the manuscript. co-authored the proposal, contributed to the study design and oversaw data acquisition. JBZ, PSB, EC and LCAG serve as guarantors, ensuring the integrity of the study. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. No specific acknowledgements are necessary for this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCLINICAL TRIAL NUMBER\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable. This study is not a clinical trial.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFrancis A, Harhay MN, Ong ACM, et al. Chronic kidney disease and the global public health agenda: an international consensus. Nat Rev Nephrol. 2024;20(7):473\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41581-024-00820-6\u003c/span\u003e\u003cspan address=\"10.1038/s41581-024-00820-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHill NR, Fatoba ST, Oke JL, et al. Global Prevalence of Chronic Kidney Disease - A Systematic Review and Meta-Analysis. PLoS ONE. 2016;11(7):e0158765. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0158765\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0158765\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStevens PE, Ahmed SB, Carrero JJ, et al. KDIGO 2024 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int. 2024;105(4):S117\u0026ndash;314. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2023.10.018\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2023.10.018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRecommendations, August. 25, 2021. Accessed August 6, 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.nice.org.uk/guidance/ng203/chapter/Recommendations#risk-assessment-referral-criteria-and-shared-care\u003c/span\u003e\u003cspan address=\"https://www.nice.org.uk/guidance/ng203/chapter/Recommendations#risk-assessment-referral-criteria-and-shared-care\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBravo-Z\u0026uacute;\u0026ntilde;iga J, Hinostoza-Sayas J, Goicochea Lugo S, et al. Gu\u0026iacute;a de pr\u0026aacute;ctica cl\u0026iacute;nica para el tamizaje, diagn\u0026oacute;stico y manejo de la enfermedad renal cr\u0026oacute;nica en estadios 1 al 3 en el Seguro Social del Per\u0026uacute; (EsSalud). Acta M\u0026eacute;dica Peruana. 2020;37(4):518\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.35663/amp.2020.374.1843\u003c/span\u003e\u003cspan address=\"10.35663/amp.2020.374.1843\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBravo-Z\u0026uacute;\u0026ntilde;iga J, Garcia-Gomero D, Dolores-Maldonado G, et al. Gu\u0026iacute;a de pr\u0026aacute;ctica cl\u0026iacute;nica para el manejo de la Enfermedad Renal Cr\u0026oacute;nica estad\u0026iacute;os 3b, 4 y 5 en el Seguro Social de Salud del Per\u0026uacute; (EsSalud). Revista del Cuerpo M\u0026eacute;dico Hosp Nac Almanzor Aguinaga Asenjo. 2023;16(1):102\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.35434/rcmhnaaa.2023.161.1709\u003c/span\u003e\u003cspan address=\"10.35434/rcmhnaaa.2023.161.1709\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChronic Kidney Disease in Adults: Assessment and Management. National Institute for Health and Care Excellence (NICE). 2015. Accessed September 16, 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.ncbi.nlm.nih.gov/books/NBK555204/\u003c/span\u003e\u003cspan address=\"http://www.ncbi.nlm.nih.gov/books/NBK555204/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInserra F, Torres ML, Alles A, et al. Especial atenci\u0026oacute;n sobre situaciones cl\u0026iacute;nicas espec\u0026iacute;ficas. Acta Bioqu\u0026iacute;mica Cl\u0026iacute;nica Latinoam. 2022;56(1):43\u0026ndash;74. Evaluaci\u0026oacute;n de la funci\u0026oacute;n renal para la detecci\u0026oacute;n y seguimiento de la enfermedad renal cr\u0026oacute;nica. Documento multidisciplinario de consenso 2021.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins GS, Dhiman P, Ma J, et al. Evaluation of clinical prediction models (part 1): from development to external validation. BMJ. 2024;384:e074819. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmj-2023-074819\u003c/span\u003e\u003cspan address=\"10.1136/bmj-2023-074819\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiley RD, Archer L, Snell KIE, et al. Evaluation of clinical prediction models (part 2): how to undertake an external validation study. BMJ. 2024;384:e074820. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmj-2023-074820\u003c/span\u003e\u003cspan address=\"10.1136/bmj-2023-074820\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD Statement. BMC Med. 2015;13(1):1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12916-014-0241-z\u003c/span\u003e\u003cspan address=\"10.1186/s12916-014-0241-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThanabalasingam SJ, Iliescu EA, Norman PA, et al. Independent External Validation and Comparison of Death and Kidney Replacement Therapy Prediction Models in Advanced CKD. Kidney Med. 2022;4(4):100440. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.xkme.2022.100440\u003c/span\u003e\u003cspan address=\"10.1016/j.xkme.2022.100440\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHundemer GL, Tangri N, Sood MM, et al. Performance of the Kidney Failure Risk Equation by Disease Etiology in Advanced CKD. Clin J Am Soc Nephrol. 2020;15(10):1424\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2215/CJN.03940320\u003c/span\u003e\u003cspan address=\"10.2215/CJN.03940320\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeeters MJ, van Zuilen AD, van den Brand JAJG, et al. Validation of the kidney failure risk equation in European CKD patients. Nephrol Dial Transpl. 2013;28(7):1773\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ndt/gft063\u003c/span\u003e\u003cspan address=\"10.1093/ndt/gft063\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitlock RH, Chartier M, Komenda P, et al. Validation of the Kidney Failure Risk Equation in Manitoba. Can J Kidney Health Dis. 2017;4:2054358117705372. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/2054358117705372\u003c/span\u003e\u003cspan address=\"10.1177/2054358117705372\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamspek CL, Evans M, Wanner C, et al. Kidney Failure Prediction Models: A Comprehensive External Validation Study in Patients with Advanced CKD. J Am Soc Nephrol. 2021;32(5):1174\u0026ndash;86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1681/ASN.2020071077\u003c/span\u003e\u003cspan address=\"10.1681/ASN.2020071077\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMajor RW, Shepherd D, Medcalf JF, Xu G, Gray LJ, Brunskill NJ. The Kidney Failure Risk Equation for prediction of end stage renal disease in UK primary care: An external validation and clinical impact projection cohort study. PLoS Med. 2019;16(11):e1002955. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pmed.1002955\u003c/span\u003e\u003cspan address=\"10.1371/journal.pmed.1002955\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarks A, Fluck N, Prescott GJ, et al. Looking to the future: predicting renal replacement outcomes in a large community cohort with chronic kidney disease. Nephrol Dialysis Transplantation. 2015;30(9):1507\u0026ndash;17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ndt/gfv089\u003c/span\u003e\u003cspan address=\"10.1093/ndt/gfv089\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKwek JL, Pang HQJ, Li H, et al. 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. Singap Med J. 2022;63(6):313\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.11622/smedj.2020170\u003c/span\u003e\u003cspan address=\"10.11622/smedj.2020170\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang MW, Tangri N, Kim YC, et al. An independent validation of the kidney failure risk equation in an Asian population. Sci Rep. 2020;10(1):12920. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-020-69715-3\u003c/span\u003e\u003cspan address=\"10.1038/s41598-020-69715-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan L, Wang J, Deng Y, et al. External Validation of the Kidney Failure Risk Equation Among Urban Community-Based Chinese Patients With CKD. Kidney Med. 2024;6(5):100817. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.xkme.2024.100817\u003c/span\u003e\u003cspan address=\"10.1016/j.xkme.2024.100817\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaher F, Teece L, Major RW, et al. Using the kidney failure risk equation to predict end-stage kidney disease in CKD patients of South Asian ethnicity: an external validation study. Diagn Progn Res. 2023;7(1):22. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s41512-023-00157-x\u003c/span\u003e\u003cspan address=\"10.1186/s41512-023-00157-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y, Nguyen FNHL, Allen JC, Lew JQL, Tan NC, Jafar TH. Validation of the kidney failure risk equation for end-stage kidney disease in Southeast Asia. BMC Nephrol. 2019;20(1):451. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12882-019-1643-0\u003c/span\u003e\u003cspan address=\"10.1186/s12882-019-1643-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJahan S, Hale J, Malacova E, Hurst C, Kark A, Mallett A. Real world evaluation of kidney failure risk equations in predicting progression from chronic kidney disease to kidney failure in an Australian cohort. J Nephrol. 2024;37(1):231\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s40620-023-01680-2\u003c/span\u003e\u003cspan address=\"10.1007/s40620-023-01680-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTangri N, Grams ME, Levey AS, et al. Multinational Assessment of Accuracy of Equations for Predicting Risk of Kidney Failure: A Meta-analysis. JAMA. 2016;315(2):164\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2015.18202\u003c/span\u003e\u003cspan address=\"10.1001/jama.2015.18202\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGallego-Valcarce E, Shabaka A, Tato-Ribera AM, et al. External validation of the KFRE and Grams prediction models for kidney failure and death in a Spanish cohort of patients with advanced chronic kidney disease. J Nephrol. 2024;37(2):429\u0026ndash;37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s40620-023-01819-1\u003c/span\u003e\u003cspan address=\"10.1007/s40620-023-01819-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLennartz CS, Pickering JW, Seiler-Mu\u0026szlig;ler S, et al. External Validation of the Kidney Failure Risk Equation and Re-Calibration with Addition of Ultrasound Parameters. Clin J Am Soc Nephrol. 2016;11(4):609\u0026ndash;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2215/CJN.08110715\u003c/span\u003e\u003cspan address=\"10.2215/CJN.08110715\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eda Silva BM, Charreu J, Duarte I, Outerelo C, Gameiro J. Validation of the kidney failure risk equation in a Portuguese cohort. Nefrologia (Engl Ed). 2023;43(4):467\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.nefroe.2022.03.007\u003c/span\u003e\u003cspan address=\"10.1016/j.nefroe.2022.03.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAli I, Donne RL, Kalra PA. A validation study of the kidney failure risk equation in advanced chronic kidney disease according to disease aetiology with evaluation of discrimination, calibration and clinical utility. BMC Nephrol. 2021;22(1):194. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12882-021-02402-1\u003c/span\u003e\u003cspan address=\"10.1186/s12882-021-02402-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIrish GL, Cuthbertson L, Kitsos A, Saunder T, Clayton PA, Jose MD. The kidney failure risk equation predicts kidney failure: Validation in an Australian cohort. Nephrol (Carlton). 2023;28(6):328\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/nep.14160\u003c/span\u003e\u003cspan address=\"10.1111/nep.14160\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBravo-Z\u0026uacute;\u0026ntilde;iga J, Ch\u0026aacute;vez-G\u0026oacute;mez R, Soto-Becerra P. 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. BMJ Open. 2024;14(1):e076217. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmjopen-2023-076217\u003c/span\u003e\u003cspan address=\"10.1136/bmjopen-2023-076217\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLarrarte C, Vesga J, Ardila F, Aldana A, Perea D, Sanabria M. Validation of the Kidney Failure Risk Equation in the Colombian Population. Int J Nephrol. 2024;2024:1282664. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2024/1282664\u003c/span\u003e\u003cspan address=\"10.1155/2024/1282664\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonz\u0026aacute;lez D, Sola L, Rios P, WCN24-1069 VALIDATION OF THE KIDNEY FAILURE RISK EQUATION (KFRE) AND ITS APPLICATION IN PATIENTS WITH CHRONIC KIDNEY DISEASE FOLLOWED BY THE RENAL HEALTH-CARE PROGRAM OF URUGUAY, et al. Kidney Int Rep. 2024;9(4):S86\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ekir.2024.02.184\u003c/span\u003e\u003cspan address=\"10.1016/j.ekir.2024.02.184\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoons KGM, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Ann Intern Med. 2015;162(1):W1\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7326/M14-0698\u003c/span\u003e\u003cspan address=\"10.7326/M14-0698\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteyerberg EW. Clinical Prediction Models: A Practical Approach to Development, Validation, and Updating. 2nd ed. Springer; 2020.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarrell FE. \u003cem\u003eRegression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis\u003c/em\u003e. Second edition. Springer; 2015.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAustin PC. Multiple imputation with competing risk outcomes. \u003cem\u003eComput Stat\u003c/em\u003e. Published online June. 2024;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00180-024-01518-w\u003c/span\u003e\u003cspan address=\"10.1007/s00180-024-01518-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Geloven N, Giardiello D, Bonneville EF, et al. Validation of prediction models in the presence of competing risks: a guide through modern methods. BMJ. 2022;377:e069249. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/bmj-2021-069249\u003c/span\u003e\u003cspan address=\"10.1136/bmj-2021-069249\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamspek CL, Teece L, Snell KIE, et al. Lessons learnt when accounting for competing events in the external validation of time-to-event prognostic models. Int J Epidemiol. 2022;51(2):615\u0026ndash;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ije/dyab256\u003c/span\u003e\u003cspan address=\"10.1093/ije/dyab256\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVickers AJ, van Calster B, Steyerberg EW. A simple, step-by-step guide to interpreting decision curve analysis. Diagn Prognostic Res. 2019;3(1):18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s41512-019-0064-7\u003c/span\u003e\u003cspan address=\"10.1186/s41512-019-0064-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Buuren S. Flexible Imputation of Missing Data. CRC; 2012.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOoi YG, Sarvanandan T, Hee NKY, et al. Risk Prediction and Management of Chronic Kidney Disease in People Living with Type 2 Diabetes Mellitus. Diabetes Metab J. 2024;48(2):196\u0026ndash;207. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4093/dmj.2023.0244\u003c/span\u003e\u003cspan address=\"10.4093/dmj.2023.0244\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePotok OA, Nguyen HA, Abdelmalek JA, Beben T, Woodell TB, Rifkin DE. Patients,\u0026rsquo; Nephrologists,\u0026rsquo; and Predicted Estimations of ESKD Risk Compared with 2-Year Incidence of ESKD. Clin J Am Soc Nephrol. 2019;14(2):206\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2215/CJN.07970718\u003c/span\u003e\u003cspan address=\"10.2215/CJN.07970718\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrams ME, Sang Y, Ballew SH, et al. Predicting timing of clinical outcomes in patients with chronic kidney disease and severely decreased glomerular filtration rate. Kidney Int. 2018;93(6):1442\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.kint.2018.01.009\u003c/span\u003e\u003cspan address=\"10.1016/j.kint.2018.01.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNaranjo FS, Sang Y, Ballew SH, et al. Estimating Kidney Failure Risk Using Electronic Medical Records. Kidney360. 2021;2(3):415\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.34067/KID.0005592020\u003c/span\u003e\u003cspan address=\"10.34067/KID.0005592020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBundy JD, Mills KT, Anderson AH, et al. Prediction of End-Stage Kidney Disease Using Estimated Glomerular Filtration Rate With and Without Race: A Prospective Cohort Study. Ann Intern Med. 2022;175(3):305\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7326/M21-2928\u003c/span\u003e\u003cspan address=\"10.7326/M21-2928\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrams ME, Li L, Greene TH, Tin A, Sang Y, Kao WHL, et al. Estimating time to ESRD using kidney failure risk equations: results from the African American Study of Kidney Disease and Hypertension (AASK). Am J Kidney Dis Off J Natl Kidney Found. 2015;65:394\u0026ndash;402.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamspek CL, Jager KJ, Dekker FW, Zoccali C, van Diepen M. External validation of prognostic models: what, why, how, when and where? Clin. Kidney J. 2021;14:49\u0026ndash;58.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-nephrology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bnep","sideBox":"Learn more about [BMC Nephrology](http://bmcnephrol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bnep/default.aspx","title":"BMC Nephrology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Chronic Kidney Disease, Kidney Failure Risk Equation, External Validation, Clinical Utility, Peru, Decision Curve Analysis","lastPublishedDoi":"10.21203/rs.3.rs-5520011/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5520011/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eThe Kidney Failure Risk Equation (KFRE) is widely used for predicting kidney failure, but its external validity in Latin America is limited. A previous study in Peru found that KFRE was miscalibrated but did not evaluate its recalibration or clinical utility.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe conducted a retrospective cohort study using data from EsSalud\u0026rsquo;s Renal Health Surveillance Program (2013\u0026ndash;2022), including 30,031 patients with chronic kidney disease (CKD) stages G3-4. Kidney failure was defined by dialysis initiation or nephrologist-confirmed end-stage renal disease. Calibration was assessed using observed-to-expected (O/E) ratios and differences, calibration slope, and intercept, while discrimination was evaluated using the concordance index (C-index). Recalibrated models were developed, and decision curve analysis (DCA) was performed to evaluate clinical utility.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThe original KFRE demonstrated good discrimination (C-index: 0.88 at 2 years, 0.85 at 5 years) but poor calibration in-the-large: O/E ratios indicated mean underestimation of risk at 2 years (O/E ratio: 1.84) and a slight mean overestimation at 5 years (O/E ratio: 1.06). Original KFRE also had poor weak (slope: 0.58) and poor moderate calibration. Recalibrated models improved calibration in-the-large, but none achieved good weak (all slope\u0026thinsp;\u0026lt;\u0026thinsp;1) and moderate calibration. However, DCA showed a higher net benefit for KFRE-based nephrology referrals (in original and recalibrated by method D) compared to Peruvian and international guidelines, especially over a 5-year horizon.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eDespite miscalibration, KFRE remains valuable for guiding nephrology referrals in Peru, with recalibrated models offering potential improvements. This is the first study in Latin America to rigorously assess the clinical utility of KFRE.\u003c/p\u003e","manuscriptTitle":"External validation, recalibration, and clinical utility of the kidney failure risk equation in patients with advanced CKD: a nationwide retrospective cohort analysis in Peru","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-16 08:14:50","doi":"10.21203/rs.3.rs-5520011/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-06-19T13:02:16+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-06-16T12:35:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"286947648573069929415572596429913407843","date":"2025-06-16T12:17:16+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-13T09:42:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-06-11T12:56:28+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Nephrology","date":"2025-06-09T06:50:45+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-nephrology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bnep","sideBox":"Learn more about [BMC Nephrology](http://bmcnephrol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bnep/default.aspx","title":"BMC Nephrology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"279ce8fa-ae68-44dc-afcf-56697ccccc24","owner":[],"postedDate":"June 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-12-08T16:13:28+00:00","versionOfRecord":{"articleIdentity":"rs-5520011","link":"https://doi.org/10.1186/s12882-025-04357-z","journal":{"identity":"bmc-nephrology","isVorOnly":false,"title":"BMC Nephrology"},"publishedOn":"2025-12-05 15:57:22","publishedOnDateReadable":"December 5th, 2025"},"versionCreatedAt":"2025-06-16 08:14:50","video":"","vorDoi":"10.1186/s12882-025-04357-z","vorDoiUrl":"https://doi.org/10.1186/s12882-025-04357-z","workflowStages":[]},"version":"v1","identity":"rs-5520011","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5520011","identity":"rs-5520011","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}