Skip to main content
NCBI home page
Springer logoLink to Springer
. 2025 Sep 22;41(1):203–216. doi: 10.1007/s00467-025-06942-8

Comparison of different equations for estimating the glomerular filtration rate in pediatric kidney transplant recipients

Paphawadee Sukboonthong 1, Julaporn Pooliam 2, Maturin Jantongsree 3, Achra Sumboonnanonda 4, Anirut Pattaragarn 4, Suroj Supavekin 4, Nuntawan Piyaphanee 4, Kraisoon Lomjansook 4, Yarnarin Thunsiribuddhichai 4, Intraparch Tinnabut 5, Nuttiporn Khueankong 5, Thanaporn Chaiyapak 4,✉
PMCID: PMC12686102  PMID: 40983736

Abstract

Background

Accurate glomerular filtration rate estimation (eGFR) is essential for managing pediatric kidney transplant recipients. Given the physiology of pediatric patients receiving adult-donor kidneys, identifying the most appropriate plasma creatinine (PCr)-based formula—pediatric or adult-specific—is crucial.

Methods

This cross-sectional study included pediatric kidney transplant recipients (age 1–18 years) who received adult-donor kidneys. We compared agreement thresholds of various pediatric and adult PCr-based GFR equations with CKiD 2012 combined PCr‒cystatin C (PCr-CystC) equation via intraclass correlation coefficients (ICCs), concordance correlation coefficients (CCCs), total deviation index (TDI), P30 performance metric (P30), Bland–Altman plots, and receiver-operating characteristic (ROC) analysis. Correlation between CKiD under 25 (U25) PCr–CystC and reference CKiD 2012 equation was also evaluated.

Results

One hundred twenty samples were collected from 23 recipients (mean age = 14.2 ± 3.4 years) and donors (mean age = 31.7 ± 10.0 years). Schwartz–Lyon equation demonstrated the highest performance with the reference (ICC = 0.913, CCC = 0.911, TDI = 14.0 mL/min/1.73 m2, P30 = 99.2%). U25 (ICC = 0.922, CCC = 0.882, P30 = 93.3%), full age spectrum (FAS)-height (ICC = 0.897, CCC = 0.877, P30 = 96.7%), and Bedside Schwartz equations (ICC = 0.850, CCC = 0.819, P30 = 89.2%) showed comparable performance. Bland–Altman plots revealed proportional bias (p < 0.05), leading to ROC analysis, which identified eGFR < 70 mL/min/1.73 m2 for Schwartz–Lyon, U25, and FAS-height, and < 60 mL/min/1.73 m2 for Bedside Schwartz as optimal agreement thresholds, beyond which each equation showed increased bias. Subgroup analyses also showed better performance in patients aged 10–18 years. Additionally, U25 PCr-CystC equation showed excellent agreement with the reference (ICC = 0.993, CCC = 0.990, P30 = 100%).

Conclusions

Schwartz–Lyon equation demonstrated the highest performance among PCr-based equations with the reference in pediatric kidney transplant recipients, particularly when eGFR was < 70 mL/min/1.73 m2 and in patients aged 10–18 years. U25 PCr-CystC equation showed best overall agreement with the reference and should be preferred where CystC measurement is feasible.

Graphical abstract

A higher resolution version of the Graphical abstract is available as Supplementary information. graphic file with name 467_2025_6942_Figa_HTML.jpg

Supplementary information

The online version contains supplementary material available at 10.1007/s00467-025-06942-8.

Keywords: Kidney transplantation, Pediatrics, Estimated glomerular filtration rate, Schwartz–Lyon equation, CKiD under 25, Bedside Schwartz equation

Introduction

Compared with chronic dialysis, kidney transplantation is the preferred treatment for pediatric patients with stage 5 chronic kidney disease (CKD), offering superior survival outcomes and quality of life [1, 2]. Accurate estimation of the glomerular filtration rate (eGFR) is essential in managing pediatric kidney transplant recipients, as it can be used to guide medication dosing, monitor allograft rejection, and assess long-term outcomes. Traditionally, GFR measurement using inulin clearance or iohexol plasma clearance has been considered the gold standard. However, these methods are costly, time-consuming, and require specialized laboratory facilities, limiting their widespread use in routine clinical practice [3, 4]. Consequently, endogenous biomarkers such as plasma creatinine (PCr) are commonly used for the eGFR. However, PCr is influenced by muscle mass, diet, and hydration status, often leading to overestimation or underestimation of the GFR in pediatric patients [5].

Although several PCr-based eGFR equations have been developed, most are population-specific. Pediatric-specific equations, such as the Bedside Schwartz, Schwartz–Lyon, and Chronic Kidney Disease in Children (CKiD) under 25 (U25) equations, are commonly used in children, whereas adult-specific equations, including Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), Modification of Diet in Renal Disease (MDRD), and Cockcroft–Gault, are standard in adults [6]. Recently, full-age spectrum (FAS) equations—such as FAS-age and FAS-height by Pottel et al., the European Kidney Function Consortium (EKFC), and the Lund–Malmö equations—have been introduced to bridge this gap and are applicable across both pediatric and adult populations. However, most of these equations have not been directly validated in pediatric kidney transplant recipients or systematically compared across pediatric and adult age ranges. In pediatric kidney transplant recipients, who typically receive kidneys from adult donors, the optimal equation remains unclear. While some previous pediatric studies have assessed the GFR in pediatric kidney transplant recipients, none have directly compared pediatric- and adult-specific PCr-based eGFR equations [7–10]. Most studies have focused solely on pediatric eGFR equations, leaving a knowledge gap regarding the most appropriate formula for this unique group of patients.

Cystatin C (CystC), a low-molecular–weight protein, is unaffected by muscle mass, sex, and inflammation, making it a promising alternative for eGFR, particularly in pediatric populations above 12 months of age [4, 11–15]. However, it is less accessible and more expensive, limiting its routine use in healthcare centers [16]. The recently updated Kidney Disease: Improving Global Outcomes (KDIGO) 2024 guidelines emphasize the use of the combined PCr-CystC-based eGFR when the PCr-based eGFR alone is less reliable, given its greater accuracy across different populations [17]. A previous study identified the CKiD 2012 combined PCr-CystC-based equation as the most accurate eGFR in pediatric kidney transplant recipients, outperforming both PCr-based and CystC-based equations alone, particularly for measured GFR (mGFR) values below 90 mL/min/1.73 m2 [7]. Additionally, another pediatric study showed that the Schwartz combined PCr-CystC-based equation performed better in pediatric kidney transplant recipients than equations based on PCr or CystC alone [8].

Therefore, the aim of this study was to compare the performance of pediatric and adult PCr-based GFR equations against the CKiD 2012 combined PCr-CystC equation in pediatric kidney transplant recipients receiving adult donor kidneys.

Methods

Study population

This study included pediatric kidney transplant recipients aged 1–18 years at Siriraj Hospital between April 2024 and November 2024. Eligible patients had undergone kidney transplantation at least one month prior to recruitment, with donors aged over 18 years. The exclusion criteria were as follows: (i) acute kidney injury (AKI), defined as a serum creatinine increase > 0.3 mg/dL within 48 h, > 1.5 times baseline within 7 days, or a urine output < 0.5 mL/kg/h for 6 h [18]; (ii) allograft failure, i.e., requiring peritoneal dialysis or hemodialysis; and (iii) conditions affecting blood urea nitrogen (BUN) levels, including moderate to severe dehydration, gastrointestinal hemorrhage, heart failure, severe infection, liver disease (e.g., cirrhosis, liver failure), use of diuretics, or systemic steroid use > 0.5 mg/kg/day.

Measurement of Pcr and CystC

PCr was measured using an enzymatic method, and CystC was quantified using a particle-enhanced immunoturbidimetric assay. Both methods were applied throughout the study to ensure reliability. PCr, CystC, and BUN samples were collected simultaneously to maintain standardized procedures.

Estimation of eGFR

eGFR values were calculated using 12 PCr-based equations: pediatric-specific equations (Bedside Schwartz, Schwartz–Lyon, U25, and Counahan), 4 adult-specific equations (CKD-EPI, MDRD, Cockcroft–Gault, and Baracskay), and 4 full-age spectrum equations (FAS-age, FAS-height, EKFC, and Lund Malmö). These equations were compared with the CKiD 2012 combined PCr-CystC equation. The equations used were as follows:

  • (i)

    Bedside Schwartz [19]: eGFR = K × height (cm)/PCr, where K = 0.413

  • (ii)

    Schwartz‒Lyon [20]: eGFR = K × height (cm)/PCr, where K = 0.413 for boys > 13 years and K = 0.367 for others

  • (iii)
    U25 [21] = K × height (m)/PCr, where K is calculated as the following:
    • Male, K is calculated as age 1 to < 12 years old = 39.0 × 1.008(age−12), age 12 to < 18 years old = 39.0 × 1.045(age−12), and age 18–25 years old = 50.8
    •  Female, K is calculated as age 1 to < 12 years old = 36.1 × 1.008(age−12), age 12 to < 18 years old = 36.1 × 1.023(age−12), and age 18–25 years old = 41.4
  • (iv)
    CKD-EPI [22]: eGFR = 142 × (PCr/A)B × 0.9938.age × (1.012 if female)
    • Female: A = 0.7, B =  − 0.241 if PCr ≤ 0.7; A = 0.7, B =  − 1.2 if PCr > 0.7
    • Male: A = 0.9, B =  − 0.302 if PCr ≤ 0.9; A = 0.9, B =  − 1.2 if PCr ≥ 0.9
  • (v)

    MDRD [23]: eGFR = 175 × PCr−1.154 × age.−0.203 × 1.212 (if the patient is black) × 0.742 (if female)

  • (vi)

    Cockcroft-Gault [6]: eGFR = (140-age) × (weight, kg) × (0.85 if female)/(72 × PCr)

  • (vii)

    FAS-age [24]: eGFR = 107.3/(PCr/Q), with Q matching age

  • (viii)

    FAS-height [24]: eGFR = 107.3/(PCr/Q), with Q matching height

  • (ix)

    EKFC [25]: eGFR = 107.3 × (PCr/Q)−0.322 if PCr/Q < 1 or eGFR = 107.3 × (PCr/Q)−1.132 if PCr/Q ≥ 1, with Q values correspond to mean PCr for age and sex specific populations

  • (x)
    Lund-Malmö [26]: PCr (µmol/L) = PCr (mg/dL) × 88.4
    • Female: If PCr < 150 µmol/L, eGFR = exp (2.5 + 0.0121 × (150 − PCr) − 0.0158 × age + 0.438 × ln (age))
      • If PCr ≥ 150 µmol/L, eGFR = exp (2.5–0.926 × ln (PCr) − 0.0158 × age + 0.438 × In (age))
    • Male: If PCr < 180 µmol/L, eGFR = exp (2.56 + 0.00968 × (180 − PCr) − 0.0158 × age + 0.438 × ln (age))
      • If PCr ≥ 180 µmol/L, eGFR = exp (2.56 − 0.926 × ln (PCr) − 0.0158 × age + 0.438 × ln (age))
  • (xi)

    Counahan [6]: eGFR = 0.43 × height/PCr

  • (xii)

    Baracskay [27]: eGFR = 1/2 (100/PCr) + 88 − age

  • (xiii)

    CKiD 2012 combined PCr-CystC [13]: eGFR = 39.8 × (height/PCr)0.456 × (1.8/CystC)0.418 × (30/BUN)0.079 × (1.076male) (1.00female) × (height/1.4).0.179

  • (xiv)
    U25 combined PCr-CystC [21]: eGFR = (PCr-based U25 eGFR + CystC-based U25 eGFR)/2
    • CystC-based U25 eGFR = K × (1/CystC), where K is calculated as the following:
      • Male, K is calculated as age 1 to < 15 years old = 87.2 × 1.011(age−15), age 15 to < 18 years old = 87.2 × 0.96(age−15), and age 18 to 25 years old = 77.1
      • Female, K is calculated as age 1 to < 12 years old = 79.9 × 1.004(age−12), age 12 to < 18 years old = 79.9 × 0.974(age−12), and age 18–25 years old = 68.3

Statistical analysis

Descriptive statistics were used to summarize patient demographics. Continuous data were presented as the means with standard deviations (SDs) for normally distributed variables or medians with interquartile ranges (IQRs) for non-normally distributed variables. Categorical data are expressed as numbers and percentages. Mean differences in eGFR were analyzed using paired Student’s t tests for normally distributed data or Mann‒Whitney U tests for nonnormally distributed data. A p value < 0.05 was considered to indicate statistical significance.

The performance of the 12 PCr-based eGFR equations was assessed against the CKiD 2012 combined PCr-CystC equation using intraclass correlation coefficients (ICCs), concordance correlation coefficient (CCC), total deviation index (TDI), 10% of the reference (P10 performance metric, P10), and 30% of the reference (P30 performance metric, P30) [17]. The ICC values were interpreted as follows [28]: > 0.9 (excellent), 0.75–0.9 (good), 0.5–0.75 (moderate), and < 0.5 (poor). The CCC values were interpreted as follows [29]: > 0.99 (almost perfect), 0.95–0.99 (substantial), 0.90–0.95 (moderate), < 0.9 (poor). Pearson’s correlation coefficients (r) were interpreted as follows [30]: 0.90–1.00 (very strong correlation), 0.70–0.89 (strong correlation), 0.40–0.69 (moderate correlation), 0.10–0.39 (weak correlation), and 0.00–0.10 (negligible correlation). P10 and P30 performance metric represent the proportions of eGFR estimates falling within 10% and 30%, respectively, of the reference eGFR; higher values of P10 and P30 performance metric indicate greater performance. Agreement between the PCr-based eGFR equation and the CKiD 2012 combined PCr-CystC equation was evaluated via Bland‒Altman plots, which calculate bias as the mean difference between equations. Proportional bias was assessed using linear regression analysis of Bland–Altman plots, and if significant bias was detected (p < 0.05), receiver-operating characteristic (ROC) curves were used to determine the eGFR thresholds beyond which the agreement between methods deteriorated.

Statistical analyses were conducted using MedCalc version 19.6.4 (MedCalc Software Ltd., Ostend, Belgium) for Bland‒Altman plots, CCC, and TDI and IBM® SPSS® Statistics version 29.0 (IBM Corp, Armonk, NY, USA) for all other analyses.

Results

Demographic and clinical characteristics

Twenty-three pediatric kidney transplant recipients aged 1–18 years at Siriraj Hospital were included in the study. All met the inclusion criteria, resulting in 120 eGFR measurements. Most patients underwent multiple assessments during different blood work periods, with a median of 4.0 (4.0) GFR measurements per patient.

The baseline demographic and clinical characteristics are summarized in Table 1. The mean age was 14.2 years (3.4), with a mean body weight of 40.2 kg (16.3) and a mean height of 142.5 cm (19.9). The main causes of CKD were glomerular diseases (56.5%) and congenital anomalies of the kidney and urinary tract (30.4%). All patients received their first kidney transplant, 52.2% from cadaveric donors, with a mean age of 32.1 years (10.6). Immunosuppressive therapies received were prednisolone and tacrolimus for all patients, supplemented with either mycophenolate mofetil (74.0%) or everolimus (26.0%). Of the 23 patients included, 17 (73.9%) demonstrated stable graft function at 6 months post-transplantation. Among the six remaining patients, the underlying conditions were antibody-mediated rejection (ABMR, n = 2), recurrent IgA nephropathy (n = 1), recurrent focal segmental glomerulosclerosis (FSGS, n = 1), and CMV infection requiring ganciclovir treatment (n = 2). Importantly, despite these events, all six patients had stable graft function without acute kidney injury (AKI) at the time of study enrollment.

Table 1.

Demographic and clinical characteristics of 23 pediatric transplant recipients

Characteristics Number (n = 23)
Male, n (%) 13 (56.5)
Age, years (SD) 14.2 (3.4)
Cause of CKD, n (%)
  Glomerular disease 13 (56.5)
  CAKUT 7 (30.4)
  Unknown 1 (4.3)
  Others 2 (8.7)
Weight, kg (SD) 40.2 (16.3)
Height, cm (SD) 142.5 (19.9)
Cadaveric donor, n (%) 12 (52.2)
Donor age, years (SD) 32.1 (10.6)
Native nephrectomy, n (%) 2 (8.7)
First kidney transplant, n (%) 23 (100)
Creatinine, mg/dL (SD) 1.2 (0.6)
Cystatin C, mg/dL 1.8 (0.6)
Immunosuppressive drugs, n (%)
  Prednisolone 23 (100)
  Tacrolimus 23 (100)
  Mycophenolate mofetil 17 (74)
  Everolimus 6 (26)
Open in a new tab

CAKUT congenital anomalies of the kidney and urinary tract, CKD chronic kidney disease, SD standard deviation

Formula performance of 12 Pcr-based eGFR equations compared with the CKiD 2012 combined Pcr-CystC equation

The performance of the 12 PCr-based eGFR equations compared with the CKiD 2012 combined PCr-CystC equation is summarized in Tables 2 and 3. A total of 12 PCr-based eGFR equations were compared against the CKiD 2012 combined equation. Most equations overestimated the eGFR compared with the reference, with the smallest bias observed for Schwartz–Lyon (mean difference, 1.2 mL/min/1.73 m2; 95% CI, − 0.1, 2.5) and the largest for Baracskay (mean difference, 72.0; 95% CI, 68.7, 75.3). The Schwartz–Lyon equation showed the highest agreement with the reference (ICC = 0.913, CCC = 0.911, TDI = 14.0 mL/min/1.73 m2, P30 = 99.2%). The U25 (ICC = 0.922, CCC = 0.882, TDI = 20.0 mL/min/1.73 m2, P30 = 93.3%), FAS–height (ICC = 0.897, CCC = 0.877, TDI = 18.5 mL/min/1.73 m2, P30 = 96.7%), and Bedside Schwartz (ICC = 0.850, CCC = 0.819, TDI = 23.8 mL/min/1.73 m2, P30 = 89.2%) equations showed comparable performance. In contrast, MDRD and Baracskay exhibited poor concordance (CCC = 0.212 and 0.099, respectively) and low P30 performance metric (P30 = 26.7 and 0, respectively). Subgroup analyses by age and eGFR strata consistently supported the superior performance of the Schwartz–Lyon equation, particularly in patients aged 10–18 years and those with GFR < 75 mL/min/1.73 m2. In addition, the U25, FAS–height, and Bedside Schwartz equations showed comparable performance and also performed well in this subgroup.

Table 2.

Correlation and performance metric of the creatinine-based eGFR equations compared with the CKiD 2012 combined plasma creatinine–cystatin C equation (n = 120)

Measurements CKiD 2012 Bedside Schwartz Schwartz-Lyon U25 CKD-EPI MDRD Cockcroft-Gault
All patients (n = 120)
  eGFR, mL/min/1.73 m2, mean (SD) 53.0 (14.9) 58.3 (22.9) 54.2 (19.4) 58.3 (19.8) 66.6 (24.4) 97.2 (60.8) 62.4 (26.5)
  Mean difference (95% CI) Reference 5.3 (3.4, 7.2) 1.2 (− 0.1, 2.5) 5.3 (4.0, 6.5) 13.6 (11.2, 16.0) 44.2 (34.8, 53.5) 9.4 (6.4, 12.5)
  ICC (95% CI) Reference 0.850 (0.791, 0.893) 0.913 (0.878, 0.939) 0.922 (0.890, 0.945) 0.786 (0.70, 0.846) 0.312 (0.148, 0.470) 0.694 (0.588, 0.776)
  CCC (95%CI) Reference 0.819 (0.753, 0.885) 0.911 (0.853, 0.969) 0.882 (0.831, 0.932) 0.641 (0.557, 0.724) 0.212 (0.082, 0.342) 0.632 (0.528, 0.737)
  r Reference 0.930 0.946 0.955 0.884 0.689 0.812
  TDI (95%CI), mL/min/1.73 m2 Reference 23.8 (20.2 to 33.4) 14.0 (12.1 to 21.0) 20.0 (16.2–24.0) 37.1 (30.2 to 43.2) 172.6 (92.8, 243.1) 45.4 (31.0, 54.5)
  P10 performance metric (95% CI) Reference 40.0 (31.7, 48.9) 55.0 (46.1, 63.6) 67.5 (58.7, 75.2) 26.7 (19.6, 35.2) 3.3 (1.3, 8.2) 20.0 (13.8, 28.0)
  P30 performance metric (95% CI) Reference 89.2 (82.3, 93.6) 99.2 (95.4, 99.9) 93.3 (87.4, 96.6) 54.2 (45.3, 62.8) 26.7 (19.6, 35.2) 60.8 (51.9, 69.1)
Age < 10 years (n = 26)
  eGFR, mL/min/1.73 m2, mean (SD) 64.6 (8.0) 83.2 (15.6) 73.9 (13.9) 73.0 (15.0) 74.4 (11.1) 185.2 (70.3) 67.4 (17.7)
  Mean difference (95% CI) Reference 18.6 (8.5) 9.3 (6.9) 8.3 (8.4) 9.8 (12.6) 120.6 (64.3) 2.8 (15.6)
  ICC (95% CI) Reference 0.765 (0.542, 0.887) 0.816 (0.631, 0.913) 0.759 (0.531, 0.884) 0.149 (0.246, 0.501) 0.174 (− 0.221, 0.520) 0.351 (− 0.035, 0.645)
  CCC (95%CI) Reference 0.360 (− 0.031, 0.656) 0.609 (0.291, 0.806) 0.611 (0.294, 0.808) 0.098 (− 0.300, 0.468) 0.045 (− 0.349, 0.425) 0.344 (− 0.050, 0.645)
  r Reference 0.943 0.943 0.914 0.157 0.775 0.467
  TDI (95%CI), mL/min/1.73 m2 Reference 35.9 (24.6, 37.1) 22.8 (14.0, 23.8) 25.4 (15.6, 26.2) 28.3 (25.2, 30.2) 249.2 (196.3, 264.0) 32.1 (22.2, 34.0)
  P10 performance metric (95% CI) Reference 7.7 (2.1, 24.1) 38.5 (22.4, 57.5) 46.2 (28.8, 64.5) 53.8 (35.4, 71.3) 0 (0, 12.9) 26.9 (13.7, 46.1)
  P30 performance metric (95% CI) Reference 57.7 (39.0, 74.5) 96.2 (81.1, 99.3) 92.3 (75.9, 97.9) 65.4 (46.2, 80.6) 0 (0, 12.9) 80.8 (62.1, 91.5)
Age 10–15 years (n = 38)
  eGFR, mL/min/1.73 m2, mean (SD) 44.3 (16.6) 48.0 (23.9) 44.5 (19.9) 46.5 (20.0) 53.8 (28.9) 72.1 (29.2) 47.8 (30.2)
  Mean difference (95% CI) Reference 3.7 (1.3) 0.13 (4.7) 2.1 (4.6) 9.5 (13.0) 27.8 (14.5) 3.5 (15.7)
  ICC (95% CI) Reference 0.920 (0.852, 0.958) 0.967 (0.937, 0.983) 0.969 (0.942, 0.984) 0.848 (0.726, 0.918) 0.813 (0.668, 0.898) 0.792 (0.635, 0.886)
  CCC (95%CI) Reference 0.906 (0.825, 0.950) 0.967 (0.937, 0.983) 0.963 (0.929, 0.981) 0.785 (0.621, 0.883) 0.482 (0.192, 0.695) 0.784 (0.620, 0.883)
  r Reference 0.983 0.983 0.986 0.983 0.946 0.939
  TDI (95%CI), mL/min/1.73 m2 Reference 20.2 (12.6, 23.1) 9.4 (6.1 to 12.9) 11.7 (6.3, 15.3) 31.4 (23.8, 34.4) 49.5 (43.0, 72.0) 41.9 (14.6, 45.3)
  P10 performance metric (95% CI) Reference 47.4 (32.5, 62.7) 65.8 (49.9, 78.8) 78.9 (63.7, 88.9) 26.3 (15.0, 42.1) 0 (0, 9.2) 18.4 (9.2, 33.4)
  P30 performance metric (95% CI) Reference 94.7 (82.7, 98.5) 100.0 (90.8, 100.0) 100.0 (90.8, 100.0) 65.8 (49.9, 78.8) 2.6 (0.5, 13.5) 57.9 (42.2, 72.2)
Age > 15 years (n = 56)
  eGFR, mL/min/1.73 m2, mean (SD) 53.5 (12.2) 53.7 (15.7) 51.7 (14.2) 59.5 (16.6) 71.7 (22.3) 73.3 (22.3) 70.1 (23.3)
  Mean difference (95% CI) Reference 0.2 (7.2)  − 1.8 (5.9) 6.0 (6.8) 18.2 (12.3) 19.8 (14.3) 16.6 (15.4)
  ICC (95% CI) Reference 0.870 (0.788, 0.922) 0.902 (0.838, 0.941) 0.891 (0.821, 0.935) 0.767 (0.633, 0.857) 0.681 (0.511, 0.800) 0.657 (0.479, 0.784)
  CCC (95%CI) Reference 0.870 (0.787, 0.922) 0.894 (0.825, 0.937) 0.821 (0.712, 0.892) 0.507 (0.281 to 0.679) 0.423 (0.180, 0.617) 0.470 (0.237, 0.653)
  r Reference 0.898 0.913 0.934 0.911 0.809 0.801
  TDI (95%CI), mL/min/1.73 m2 Reference 14.0 (10.5, 15.9) 11.7 (8.3, 14.0) 20.0 (16.2, 24.1) 42.2 (33.8, 44.0) 54.7 (32.8, 60.9) 51.3 (27.8, 55.6)
  P10 performance metric (95% CI) Reference 50.0 (37.3, 62.7) 55.4 (42.4, 67.6) 69.5 (56.7, 80.1) 14.3 (74.2, 25.7) 7.1 (2.8, 17.0) 17.9 (10.0, 29.8)
  P30 performance metric (95% CI) Reference 100.0 (93.6, 100.0) 100.0 (93.6, 100.0) 89.3 (78.5, 95.0) 41.1 (29.2, 54.1) 55.4 (42.4, 67.6) 53.6 (40.7, 66.0)
CKiD eGFR < 75 mL/min/1.73 m2 (n = 114)
  eGFR, mL/min/1.73 m2, mean (SD) 51.7 (14.0) 55.8 (20.4) 52.1 (17.4) 56.2 (17.9) 65.4 (24.0) 90.3 (48.4) 61.1 (26.1)
  Mean difference (95% CI) Reference 4.1 (9.3) 0.5 (6.3) 4.6 (6.2) 13.7 (12.8) 38.7 (40.3) 9.4 (16.5)
  ICC (95% CI) Reference 0.859 (0.802, 0.900) 0.919 (9.885, 0.944) 0.927 (0.896, 0.949) 0.788 (0.707, 0.849) 0.360 (0.189, 0.510) 0.688 (0.577, 0.773)
  CCC (95%CI) Reference 0.835 (0.770, 0.883) 0.919 (0.884, 0.943) 0.891 (0.846, 0.924) 0.634 (0.510, 0.733) 0.226 (0.044, 0.394) 0.625 (0.498, 0.725)
  r Reference 0.921 0.941 0.956 0.907 0.675 0.825
  TDI (95%CI), mL/min/1.73 m2 Reference 22.0 (17.2, 24.2) 13.2 (11.1, 14.4) 17.7 (15.5, 20.2) 36.9 (30.2, 41.9) 96.1 (88.5, 197.2) 43.7 (29.9, 50.3)
  P10 performance metric (95% CI) Reference 42.1 (33.0, 51.2) 57.0 (47.8, 65.7) 71.1 (62.2, 78.6) 26.3 (19.1, 35.1) 3.5 (1.4, 8.7) 16.7 (10.9, 24.6)
  P30 performance metric (95% CI) Reference 91.2 (84.6, 95.2) 100.0 (96.7, 100.0) 94.7 (89.0, 97.6) 52.6 (43.5, 61.6) 28.1 (20.7, 36.9) 59.6 (50.5, 68.2)
CKiD eGFR ≥ 75 mL/min/1.73 m2 (n = 6)
  eGFR, mL/min/1.73 m2, mean (SD) 78.6 (3.5) 105.8 (12.4) 94.0 (11.0) 97.6 (9.6) 90.3 (18.3) 227.3 (116.7) 88.8 (21.7)
  Mean difference (95% CI) Reference 27.2 (9.5) 15.4 (8.2) 19.0 (6.9) 11.7 (20.7) 148.7 (114.3) 10.2 (23.5)
  ICC (95% CI) Reference 0.444 (− 0.467, 0.898) 0.489 (− 0.420, 0.908) 0.539 (− 0.363, 0.920)  − 0.235 (− 0.840, 0.632) 0.041 (− 0.736, 0.772)  − 0.141 (− 0.810, 0.686)
  CCC (95%CI) Reference 0.081 (− 0.782, 0.838) 0.176 (− 0.742, 0.864) 0.121 (− 0.766, 0.849)  − 0.168 (− 0.862, 0.745) 0.016 (− 0.806, 0.817)  − 0.117 (− 0.848, 0.768)
  r Reference 0.837 0.837 0.828  − 0.629 0.679  − 0.444
  TDI (95%CI), mL/min/1.73 m2 Reference 37.0 (30.2, 37.1) 23.7 (18.2, 23.8) 26.2 (21.1, 26.2) 38.4 (18.2, 43.9) 260.8 (194.6, 264.0) 44.5 (4.4, 57.6)
  P10 performance metric (95% CI) Reference 0 (0, 39.0) 16.7 (3.0, 56.4) 0 (0, 39.0) 33.3 (9.7, 70.0) 0 (0, 39.0) 83.3 (43.7, 97.0)
  P30 performance metric (95% CI) Reference 50.0 (18.8, 81.2) 83.3 (43.7, 97.0) 66.7 (30.0, 90.3) 83.3 (43.7, 97.0) 0 (0, 39.0) 83.3 (43.7, 97.0)
Open in a new tab

95% CI 95% confidence interval, eGFR estimated glomerular filtration rate, CKiD Chronic Kidney Disease in Children, U25 the CKiD under 25, CKD-EPI Chronic Kidney Disease Epidemiology Collaboration, MDRD Modification of Diet in Renal Disease, ICC intraclass correlation coefficient, CCC concordance correlation coefficient, r Pearson’s correlation coefficient, TDI total deviation index, SD standard deviation, P10 performance metric proportion within 10% of the reference, P30 performance metric proportion within 30% of the reference

Table 3.

Correlation and accuracy of the creatinine-based eGFR equations compared with the CKiD 2012 combined plasma creatinine–cystatin C equation (n = 120)

Measurements CKiD 2012 FAS-age FAS-height EKFC Lund_malmö Counahan Baracskay
All patients (n = 120)
  eGFR, mL/min/1.73 m2, mean (SD) 53.0 (14.9) 68.6 (21.5) 56.9 (21.4) 63.8 (21.7) 67.2 (22.5) 60.7 (23.8) 125.0 (26.8)
  Mean difference (95% CI) Reference 15.6 (9.4) 3.9 (8.4) 10.8 (9.8) 14.2 (10/6) 7.7 (5.6, 9.7) 72.0 (68.7, 75.3)
  ICC (95% CI) Reference 0.869 (0.818, 0.907) 0.897 (0.855, 0.927) 0.861 (0.807, 0.901) 0.846 (0.786, 0.890) 0.836 (0.773, 0.883) 0.644 (0.526, 0.738)
  CCC (95% CI) Reference 0.641 (0.575, 0.707) 0.877 (0.825, 0.929) 0.738 (0.669, 0.806) 0.663 (0.593, 0.732) 0.777 (0.712, 0.843) 0.099 (0, 0.215)
  r Reference 0.929 0.956 0.924 0.920 0.930 0.759
  TDI (95%CI), mL/min/1.73 m2 Reference 33.3 (29.5, 37.3) 18.5 (15.2, 24.4) 29.7 (25.2 to 34.3) 31.1 (30.5, 33.5) 27.6 (23.9, 38.0) 107.6 (100.9, 122.2)
  P10 performance metric (95% CI) Reference 10.8 (6.4, 17.7) 36.7 (28.6, 45.6) 31.7 (24.0, 40.5) 14.2 (9.0, 21.5) 44.2 (35.6, 53.1) 0 (0, 0)
  P30 performance metric (95% CI) Reference 51.7 (42.9, 60.4) 96.7 (91.7, 98.7) 76.7 (68.3, 83.3) 61.7 (52.7, 69.9) 80.0 (72.0, 86.2) 0 (0, 0)
Age < 10 years (n = 26)
  eGFR, mL/min/1.73 m2, mean (SD) 64.6 (8.0) 84.5 (14.3) 77.0 (14.4) 81.6 (13.9) 91.6 (9.3) 86.5 (16.3) 165.5 (18.4)
  Mean difference (95% CI) Reference 0.927 0.940 0.919 0.890 21.9 (9.2) 100.9 (11.8)
  ICC (95% CI) Reference 0.791 (0.586, 0.900) 0.798 (0.600, 0.904) 0.794 (0.592, 0.902) 0.880 (0.751, 0.944) 0.746 (0.511, 0.878) 0.653 (0.362, 0.828)
  CCC (95% CI) Reference 0.319 (–0.078, 0.629) 0.510 (0.153, 0.749) 0.375 (− 0.014, 0.666) 0.151 (− 0.251, 0.509) 0.304 (− 0.095, 0.618) 0.025 (− 0.366, 0.408)
  r Reference 0.898 0.925 0.883 0.891 0.944 0.891
  TDI (95%CI), mL/min/1.73 m2 Reference 32.6 (24.2, 34.0) 26.8 (17.5, 27.9) 28.7 (22.1, 29.9) 33.3 (30.9, 34.6) 40.6 (28.3, 42.1) 123.1 (111.5, 124.1)
  P10 performance metric (95% CI) Reference 3.8 (0.7, 18.9) 15.4 (6.2, 33.5) 3.8 (0.7, 18.9) 0 (0, 12.9) 0 (0, 12.9) 0 (0, 12.9)
  P30 performance metric (95% CI) Reference 34.6 (19.4, 53.8) 88.5 (71.0, 96.0) 61.5 (42.5, 77.6) 7.7 (2.1, 24.1) 38.5 (22.4, 57.5) 0 (0, 12.9)
Age 10–15 years (n = 38)
  eGFR, mL/min/1.73 m2, mean (SD) 44.3 (16.6) 57.9 (22.1) 46.5 (23.5) 53.6 (22.6) 56.3 (22.4) 50.1 (24.9) 115.5 (19.4)
  Mean difference (95% CI) Reference 13.5 (9.0) 2.1 (7.8) 9.3 (9.2) 11.9 (8.2) 5.7 (9.1) 71.1 (7.3)
  ICC (95% CI) Reference 0.895 (0.807, 0.944) 0.927 (0.865, 0.962) 0.892 (0.803, 0.943) 0.914 (0.840, 0.954) 0.907 (0.829, 0.951) 0.919 (0.851, 0.957)
  CCC (95%CI) Reference 0.722 (0.523–0.846) 0.922 (0.855, 0.959) 0.804 (0.651, 0.894) 0.772 (0.601, 0.876) 0.875 (0.771, 0.934) 0.105 (–0.222, 0.411)
  r Reference 0.932 0.985 0.935 0.955 0.983 0.931
  TDI (95%CI), mL/min/1.73 m2 Reference 31.7 (20.0, 43.2) 18.5 (10.9, 21.6) 27.4 (15.1, 41.2) 30.3 (17.0, 36.4) 24.2 (15.6 to 26.9) 87.0 (75.2, 94.7)
  P10 performance metric (95% CI) Reference 15.8 (7.4, 30.4) 31.6 (19.1, 47.5) 23.7 (13.0, 39.2) 10.5 (4.2, 24.1) 57.9 (42.2, 72.2) 0 (0, 9.2)
  P30 performance metric (95% CI) Reference 42.1 (27.9, 57.8) 97.4 (86.5, 99.5) 84.2 (69.6, 92.6) 73.7 (58.0, 85.0) 84.2 (69.6, 92.6) 0 (0, 9.2)
Age > 15 years (n = 56)
  eGFR, mL/min/1.73 m2, mean (SD) 53.5 (12.2) 68.5 (19.4) 54.6 (15.8) 62.4 (19.2) 63.3 (18.5) 55.9 (16.4) 112.6 (12.2)
  Mean difference (95% CI) Reference 15.0 (10.0) 1.2 (6.5) 8.9 (10.2) 9.9 (9.5) 2.4 (7.6) 59.1 (7.7)
  ICC (95% CI) Reference 0.808 (0.693, 0.882) 0.895 (0.827, 0.937) 0.799 (0.680, 0.877) 0.817 (0.707, 0.880) 0.861 (0.773, 0.916) 0.800 (0.681, 0.878)
  CCC (95%CI) Reference 0.565 (0.355, 0.721) 0.892 (0.821, 0.935) 0.693 (0.526, 0.809) 0.683 (0.512, 0.802) 0.849 (0.754, 0.909) 0.063 (− 0.204, 0.320)
  r Reference 0.898 0.925 0.883 0.891 0.898 0.800
  TDI (95%CI), mL/min/1.73 m2 Reference 35.6 (27.5, 37.4) 12.5 (9.2, 14.1) 30.0 (21.4, 34.4) 30.6 (20.6, 31.7) 17.2 (12.5, 19.6) 71.1 (69.8, 72.8)
  P10 performance metric (95% CI) Reference 10.7 (5.0, 21.5) 50.0 (37.3, 62.7) 50.0 (37.3, 62.7) 23.2 (14.1, 35.8) 55.4 (42.4, 67.6) 0 (0, 6.4)
  P30 performance metric (95% CI) Reference 66.1 (53.0, 77.1) 100.0 (93.6, 100.0) 78.6 (66.2, 87.3) 78.6 (66.2, 87.3) 96.4 (87.9, 99.0) 0 (0, 6.4)
CKiD eGFR < 75 mL/min/1.73 m2 (n = 114)
  eGFR, mL/min/1.73 m2, mean (SD) 51.7 (14.0) 66.6 (19.8) 54.6 (19.3) 61.3 (20.1) 65.3 (21.5) 58.1 (21.1) 122.3 (23.8)
  Mean difference (95% CI) Reference 14.8 (13.2, 16.5) 3.0 (1.6, 4.3) 10.0 (8.2, 11.7) 13.7 (11.7, 15.6) 6.4 (10.0) 70.7 (16.7)
  ICC (95% CI) Reference 0.863 (0.807, 0.903) 0.904 (0.864, 0.933) 0.854 (0.795, 0.897) 0.831 (0.765, 0.881) 0.846 (0.785, 0.891) 0.633 (0.509, 0.732)
  CCC (95% CI) Reference 0.628 (0.503, 0.728) 0.890 (0.845, 0.923) 0.732 (0.633, 0.807) 0.647 (0.525, 0.742) 0.795 (0.716, 0.854) 0.084 (− 0.101, 0.264)
  r Reference 0.916 0.951 0.910 0.909 0.921 0.726
  TDI (95%CI), mL/min/1.73 m2 Reference 32.7 (25.8, 37.3) 16.8 (3.8, 18.5) 28.8 (22.1, 34.3) 31.1 (29.3, 33.5) 25.1 (20.5, 28.0) 102.3 (97.5, 111.8)
  P10 performance metric (95% CI) Reference 11.4 (6.8, 18.5) 38.6 (30.2, 47.8) 33.3 (25.4, 42.4) 14.9 (9.5, 22.6) 46.5 (37.6, 55.6) 0 (0, 39.3)
  P30 performance metric (95% CI) Reference 54.4 (45.3, 63.2) 99.1 (95.2, 99.8) 78.9 (70.6, 85.4) 62.3 (53.1, 70.6) 83.3 (75.4, 89.1) 50 (18.8, 81.2)
CKiD eGFR ≥ 75 mL/min/1.73 m2 (n = 6)
  eGFR, mL/min/1.73 m2, mean (SD) 78.6 (3.5) 109.0 (6.7) 100.2 (9.0) 104.6 (5.6) 102.7 (9.0) 110.2 (12.9) 175.9 (30.9)
  Mean difference (95% CI) Reference 30.3 (4.8) 21.6 (6.1) 26.0 (4.4) 24.1 (6.9) 31.6 (10.1) 97.2 (28.5)
  ICC (95% CI) Reference 0.594 (− 0.291, 0.931) 0.608 (− 0.270, 0.934) 0.555 (− 0.343, 0.923) 0.493 (− 0.416, 0.909) 0.428 (− 0.482, 0.894) 0.158 (− 0.677, 0.815)
  CCC (95%CI) Reference 0.035 (− 0.799, 0.823) 0.102 (− 0.774, 0.844) 0.034 (− 0.800, 0.823) 0.068 (− 0.787, 0.834) 0.065 (− 0.788, 0.833) 0.015 (− 0.807, 0.817)
  r Reference 0.718 0.892 0.612 0.722 0.836 0.697
  TDI (95%CI), mL/min/1.73 m2 Reference 34.9 (32.6, 35.3) 27.8 (23.0, 27.9) 29.9 (28.2, 29.9) 31.4 (26.0, 31.5) 42.0 (34.6, 42.1) 123.9 (110.0, 124.1)
  P10 performance metric (95% CI) Reference 0 (0, 39.0) 0 (0, 39.0) 0 (0, 39.0) 0 (0, 39.0) 0 (0, 39.0) 0 (0, 39.0)
  P30 performance metric (95% CI) Reference 0 (0, 39.0) 50.0 (18.8, 81.2) 33.3 (9.7, 70.0) 50.0 (18.8, 81.2) 3.0 (56.4) 0 (0, 39.0)
Open in a new tab

95% CI 95% confidence interval, eGFR estimated glomerular filtration rate, FAS the full age spectrum, EKFC European Kidney Function Consortium, ICC intraclass correlation, coefficient, CCC concordance correlation coefficient, r Pearson’s correlation coefficient, TDI total deviation index, SD standard deviation, P10 performance metric proportion within 10% of the reference, P30 performance metric proportion within 30% of the reference

Bland‒Altman plots were generated to assess the bias between the CKiD 2012 combined PCr–CystC formula and four PCr-based eGFR equations: Schwartz–Lyon, U25, FAS-height, and Bedside Schwartz (Fig. 1). The plots indicated minimal bias and narrow 95% limits of agreement, suggesting low variability between these eGFR estimation methods. Additionally, Bland–Altman plots revealed proportional bias (p < 0.05) for all equations, prompting ROC analysis to determine clinically relevant agreement thresholds. The ROC analysis revealed that eGFR values < 70 mL/min/1.73 m2 for Schwartz–Lyon, U25, and FAS-height and < 60 mL/min/1.73 m2 for Bedside Schwartz were optimal thresholds where agreement with the reference standard remained acceptable. Beyond these thresholds, both equations showed increased bias, indicating reduced reliability in higher eGFR ranges.

Fig. 1.

Fig. 1

Open in a new tab

Bland‒Altman plots showing the bias and limits of agreement of PCr based and CKiD2012 combined PCr-CystC formula. The Y-axis represents the differences, calculated as (PCr based eGFR minus CKiD2012), for each data point. The X-axis represents the CKiD2012 measurements for each data point. The solid lines represent the bias; the dotted lines represent the 95% limits of agreements (mean difference ± 1.96SD). a Schwartz-Lyon. b U25. c FAS-height. d Bedside Schwartz. eGFR, estimated glomerular filtration rate; PCr, plasma creatinine; U25, CKiD under 25; FAS, full age spectrum

Correlation between the CKiD 2012 and U25 combined PCr-CysC equations

The U25 combined PCr-CystC equation demonstrated excellent agreement with the reference CKiD 2012 combined equation (Table 4). The mean eGFR values were similar (54.0 vs. 53.0 mL/min/1.73 m2), with a minimal mean difference of 1.0 mL/min/1.73 m2 (95% CI, 0.7, 1.3). The agreement between the two equations was very high, with an ICC of 0.993 (95% CI, 0.990, 0.995), CCC of 0.990 (95% CI, 0.969, 1.0), and a TDI of 3.9 mL/min/1.73 m2 (95% CI, 3.2, 5.2). The Pearson’s correlation coefficient was 0.993. Regarding performance metric, the U25 equation showed excellent performance with a P10 of 99.2% and a P30 of 100%.

Table 4.

Correlation between the CKiD 2012 and the U25 plasma creatinine-cystatin C equations (n = 120)

Measurements CKiD 2012 combined plasma creatitine-cystatin C U25 combined plasma creatinine-cystatin C
eGFR, mL/min/1.73 m2, mean (SD) 53.0 (14.9) 54.0 (15.1)
Mean difference (95% CI) Reference 1.0 (0.7, 1.3)
ICC (95% CI) Reference 0.993 (0.990, 0.995)
CCC (95% CI) Reference 0.990 (0.969, 1.0)
TDI (95% CI), mL/min/1.73 m2 Reference 3.9 (3.2, 5.2)
r Reference 0.993
P10 performance metric (95% CI) Reference 99.2 (95.4, 99.9)
P30 performance metric (95% CI) Reference 100 (1.0, 1.0)
Open in a new tab

95% CI 95% confidence interval, eGFR estimated glomerular filtration rate, CKiD Chronic Kidney Disease in Children, U25 the CKiD under 25, CKD-EPI Chronic Kidney Disease Epidemiology Collaboration, MDRD Modification of Diet in Renal Disease, ICC intraclass correlation coefficient, CCC concordance correlation coefficient, r Pearson’s correlation coefficient, TDI total deviation index, SD standard deviation, P10 performance metric proportion within 10% of the reference, P30 performance metric proportion within 30% of the reference

Discussion

This study comprehensively evaluated the performance of 12 PCr-based eGFR equations in pediatric kidney transplant recipients, using the CKiD 2012 combined PCr–CystC equation as the reference standard. Among the tested equations, the Schwartz–Lyon formula showed the highest overall performance, with excellent ICC (0.913), CCC (0.911), TDI (14.0 mL/min/1.73 m2), and the highest P30 performance metric (99.2%). Its reliability was particularly notable in patients with eGFR < 70 mL/min/1.73 m2 and those aged 10–18 years. The U25, FAS–height, and Bedside Schwartz equations demonstrated comparable performance and serve as reasonable alternatives.

Recent guidelines for renal function monitoring in solid organ transplant recipients highlight the importance of selecting an appropriate GFR estimation method tailored to the patient’s physiological characteristics [17]. The KDIGO 2024 guidelines reinforce the role of the combined PCr-CystC-based formula in clinical decision-making, particularly when the PCr-based formula alone is less reliable [17]. Pizzo et al. assessed mGFR in 45 pediatric kidney transplant recipients using several equations, including the CKiD 2012 and U25 combined PCr–CystC formulas, and found both performed similarly and ranked among the best estimators [31]. In our study, which used the CKiD 2012 combined equation as the reference, the U25 combined equation also demonstrated excellent agreement (ICC = 0.993, CCC = 0.990, TDI = 3.9 mL/min/1.73 m2, P30 = 100%), supporting its potential as an alternative equation when CystC is available.

Our study showed that the Schwartz–Lyon equation had the best overall performance, particularly in patients with eGFR < 70 mL/min/1.73 m2 and those aged 10–18 years. The U25, FAS–height, and Bedside Schwartz also performed comparably and support previous findings on the value of pediatric-specific formulas for estimating GFR in this population [7, 9, 10, 19]. Compared with previous pediatric studies, the findings align closely with those of de Souza et al. [7], who reported that the Schwartz–Lyon equation demonstrated good performance with respect to mGFR, particularly at lower GFR levels (mGFR < 60 mL/min/1.73 m2), with high P30 accuracy rates (97–98%) and better reliability than other PCr-based equations. Additionally, Papez et al. evaluated eGFR estimation in a Hispanic-dominant pediatric kidney transplant population and reported that the Bedside Schwartz equation demonstrated the highest correlations with mGFR using iothalamate clearance [9]. A previous study by Tsampalieros et al. compared the Bedside Schwartz eGFR with the CKiD 2009 combined with the PCr-CysC formula in pediatric kidney transplant recipients and reported that the Bedside Schwartz formula overestimated the GFR, with a mean bias of 1.09 mL/min/1.73 m2 [32]. However, they did not include the Schwartz–Lyon equation. A previous Canadian study by Alkandari et al. found that the modified Schwartz and FAS-age equations demonstrated better bias and P30 accuracy in pediatric kidney transplant recipients [10], although FAS-height was not assessed. In contrast, our study demonstrated superior performance of FAS-height over FAS-age. This discrepancy may reflect population-specific growth patterns. In our Thai cohort, post-transplant growth retardation is common, making age a less reliable proxy for body size. As FAS-height incorporates actual stature, it may better capture renal function in this population. Conversely, the Canadian cohort may have had growth patterns more aligned with the original European derivation cohort of FAS-age, contributing to its better performance in that setting.

A previous study conducted by de Souza et al. analyzed the performance of different eGFR equations in pediatric kidney transplant recipients and assessed their ability to correctly classify renal dysfunction using AUC analysis. They reported that the Schwartz–Lyon and Bedside Schwartz AUCs were 0.97 and 0.96, respectively, for detecting an mGFR < 60 mL/min/1.73 m2 [7]. Consistent with our findings, this study reinforces the need for equation-specific eGFR thresholds, as identified by our ROC analysis (< 70 mL/min/1.73 m2 for Schwartz–Lyon, U25, and FAS-height and < 60 mL/min/1.73 m2 or Bedside Schwartz). Additionally, a Canadian study by Alkandari et al. found that the modified Schwartz performed best in children aged 5–15 years. In our study, the Bedside Schwartz showed better performance in those aged 10–18 years. These results highlight the importance of selecting eGFR equations with high classification accuracy, particularly in populations at risk of developing allograft dysfunction.

This study has several strengths. First, our study comprehensively compared several PCr-based eGFR equations. This comprehensive comparison provides substantial advantages over studies that evaluate only a limited number of equations. Second, this study focused on a unique clinical population of pediatric kidney transplant recipients receiving adult donor kidneys. This focus provides valuable insights into the performance of eGFR equations tailored to the unique physiological and clinical characteristics of these patients, thereby filling a critical gap in nephrology research. Third, the application of advanced statistical analyses, including Bland‒Altman plots, ICCs, and performance metrics such as P10 and P30 performance metrics, ensures a thorough assessment of each eGFR formula’s reliability and agreement with the reference standard. These methods are widely recognized for their effectiveness in method comparison studies, enhancing the credibility of the results. Fourth, ROC analysis was used to define eGFR thresholds at which each equation maintains optimal accuracy, and subgroup analyses by age were performed. This approach enhances clinical applicability by offering more precise guidance for renal function assessment in pediatric kidney transplant recipients.

There are some limitations in this study. First, although the sample size calculation indicated the need for at least 120 samples to demonstrate a significant difference [7], this study included only 23 patients because of the limited number of transplant recipients at our center during the study period. However, each individual measurement remained independent and did not exert any influence on the others. Second, there is a lack of direct mGFR assessments to confirm the absolute accuracy of the eGFR equations. While eGFR equations are widely used and practical for clinical application, they may not always precisely reflect the true GFR. Consequently, our study relies on comparative analyses between different eGFR equations rather than direct GFR measurements. Third, while corticosteroids may affect CystC levels, a prior study in pediatric kidney transplant recipients revealed no significant correlation between CystC levels and prednisolone use [33]. Furthermore, this study included only patients receiving low-dose prednisolone (< 0.5 mg/kg/day), which may have minimized any potential impact on CystC levels. Fourth, although timing post-transplant may influence the performance of eGFR equations, our dataset included only four eGFR measurements obtained within the first 3 months after transplantation. Due to this limited sample size, we were unable to perform a dedicated subgroup analysis for this early post-kidney transplant period. Fifth, the CKiD 2012 equation was developed in children with mostly stage 3 or higher CKD and may not fully reflect transplant recipients with near-normal GFR [13]. In our study, most equations performed better in the < 75 mL/min/1.73 m2 group. Similarly, a recent study noted reduced accuracy of eGFR equations at higher GFR levels, highlighting the need for improved models for children with preserved kidney function [34]. Future research should involve larger, multicenter studies—including patients in the early post-kidney transplant period and incorporating direct GFR measurements to increase the validity and reliability of kidney function assessment methods in pediatric kidney transplant recipients.

In conclusion, the Schwartz–Lyon equation demonstrated the highest performance in estimating the eGFR in pediatric kidney transplant recipients, particularly when eGFR was < 70 mL/min/1.73 m2 and in patients aged 10–18 years, and should be considered the preferred equation in this population when CystC is not available. Notably, the U25 combined PCr-CystC equation showed the best overall agreement of the reference CKiD 2012 combined equation and may be preferred in clinical settings when CystC measurement is feasible. The U25, FAS–height, and Bedside Schwartz equations demonstrated comparable performance and serve as reasonable alternatives. These findings reinforce the importance of selecting appropriate eGFR estimation methods tailored to the physiology of pediatric kidney transplant patients.

Supplementary information

Below is the link to the electronic supplementary material.

Graphical abstract (PPTX 258 KB) (258.7KB, pptx)

Acknowledgements

The authors express their sincere gratitude to all the study participants and extend special thanks to Miss Putcharin Thaingsompong and Miss Apichaya Yoomeesuk of the Department of Pediatrics for their assistance with the document.

Author contributions

P.S. designed the study, collected the data, performed the analysis, interpreted the data, drafted the initial manuscript. J.P. performed the analysis, interpreted the data. M.J. invited participants, collected the data, assisted with financial documentation. A.S. interpreted the data, revised the work critically for important intellectual content. S.S. interpreted the data, revised the work critically for important intellectual content. A.P. interpreted the data, revised the work critically for important intellectual content. N.P. interpreted the data, revised the work critically for important intellectual content. K.L. interpreted the data, revised the work critically for important intellectual content. Y.T. interpreted the data, revised the work critically for important intellectual content. I.T. arranged for cystatin C measurements. N.K. arranged for cystatin C measurements. T.C. made substantial contributions to the conception of the study, designed the study, interpreted the data, revised the work critically for important intellectual content. All authors approved the final manuscript as submitted.

Funding

Open access funding provided by Mahidol University. This study is supported by the Siriraj Research Development Fund (Managed by Routine to Research: R2R), Grant Number (IO) R016735025, Siriraj Foundation and the Faculty of Medicine Siriraj Hospital, Mahidol University, R016735025, Thanaporn Chaiyapak, Siriraj Foundation.

Declarations

Ethics approval

This study was approved by the Institutional Review Board of Siriraj Hospital (COA no. Si 587/2023).

Consent to participate

Informed consent and assent were obtained from all parents and participants, respectively.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.McDonald SP, Craig JC; Australian and New Zealand Paediatric Nephrology Association (2004) Long-term survival of children with end-stage renal disease. N Engl J Med 350:2654–2662. 10.1056/NEJMoa031643 [DOI] [PubMed] [Google Scholar]
  • 2.Chaichana U, Pooliam J, Jantongsree M, Chantaratin S, Sumboonnanonda A, Pattaragarn A, Supavekin S, Piyaphanee N, Lomjansook K, Thunsiribuddhichai Y, Tinnabut I, Chaiyapak T (2025) Quality of life in children at different stages of chronic kidney disease in a developing country. Pediatr Nephrol 40:177–187. 10.1007/s00467-024-06442-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.den Bakker E, Gemke R, Bökenkamp A (2018) Endogenous markers for kidney function in children: a review. Crit Rev Clin Lab Sci 55:163–183. 10.1080/10408363.2018.1427041 [DOI] [PubMed] [Google Scholar]
  • 4.Filler G, Sharma AP (2008) How to monitor renal function in pediatric solid organ transplant recipients. Pediatr Transplant 12:393–401. 10.1111/j.1399-3046.2007.00885.x [DOI] [PubMed] [Google Scholar]
  • 5.Mian AN, Schwartz GJ (2017) Measurement and estimation of glomerular filtration rate in childrEN. Adv Chronic Kidney Dis 24:348–356. 10.1053/j.ackd.2017.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Inker LA, Titan S (2021) Measurement and estimation of GFR for use in clinical practice: core curriculum 2021. Am J Kidney Dis 78:736–749. 10.1053/j.ajkd.2021.04.016 [DOI] [PubMed] [Google Scholar]
  • 7.de Souza V, Cochat P, Rabilloud M, Selistre L, Wagner M, Hadj-Aissa A, Dolomanova O, Ranchin B, Iwaz J, Dubourg L (2015) Accuracy of different equations in estimating GFR in pediatric kidney transplant recipients. Clin J Am Soc Nephrol 10:463–470. 10.2215/CJN.06300614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Selistre L, Roquet O, Saitovitch D, de Souza VC, Antonello IC, Ranchin B, Hadj-Aissa A, Cochat P, Dubourg L (2012) Comparison of cystatin C- and creatinine-based glomerular filtration rate formulas with inulin clearance in pediatric renal transplantation. Transplant Proc 44:2357–2359. 10.1016/j.transproceed.2012.07.051 [DOI] [PubMed] [Google Scholar]
  • 9.Papez KE, Barletta GM, Hsieh S, Joseph M, Morgenstern BZ (2013) Iothalamate versus estimated GFR in a Hispanic-dominant pediatric renal transplant population. Pediatr Nephrol 28:2369–2376. 10.1007/s00467-013-2556-y [DOI] [PubMed] [Google Scholar]
  • 10.Alkandari O, Hebert D, Langlois V, Robinson LA, Parekh RS (2017) Validation of serum creatinine-based formulae in pediatric renal transplant recipients. Pediatr Res 82:1000–1006. 10.1038/pr.2017.209 [DOI] [PubMed] [Google Scholar]
  • 11.Dharnidharka VR, Kwon C, Stevens G (2002) Serum cystatin C is superior to serum creatinine as a marker of kidney function: a meta-analysis. Am J Kidney Dis 40:221–226. 10.1053/ajkd.2002.34487 [DOI] [PubMed] [Google Scholar]
  • 12.Filler G, Lepage N (2003) Should the Schwartz formula for estimation of GFR be replaced by cystatin C formula? Pediatr Nephrol 18:981–985. 10.1007/s00467-003-1271-5 [DOI] [PubMed] [Google Scholar]
  • 13.Schwartz GJ, Schneider MF, Maier PS, Moxey-Mims M, Dharnidharka VR, Warady BA, Furth SL, Munoz A (2012) Improved equations estimating GFR in children with chronic kidney disease using an immunonephelometric determination of cystatin C. Kidney Int 82:445–453. 10.1038/ki.2012.169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Bökenkamp A, Domanetzki M, Zinck R, Schumann G, Brodehl J (1998) Reference values for cystatin C serum concentrations in children. Pediatr Nephrol 12:125–129. 10.1007/s004670050419 [DOI] [PubMed] [Google Scholar]
  • 15.Finney H, Newman DJ, Thakkar H, Fell JM, Price CP (2000) Reference ranges for plasma cystatin C and creatinine measurements in premature infants, neonates, and older children. Arch Dis Child 82:71–75. 10.1136/adc.82.1.71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Chen DC, Potok OA, Rifkin D, Estrella MM (2022) Advantages, limitations, and clinical considerations in using cystatin C to estimate GFR. Kidney360 3:1807–1814. 10.34067/kid.0003202022
  • 17.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2024) KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int 105:S117–S314. 10.1016/j.kint.2023.10.018 [DOI] [PubMed] [Google Scholar]
  • 18.Palevsky PM, Liu KD, Brophy PD, Chawla LS, Parikh CR, Thakar CV, Tolwani AJ, Waikar SS, Weisbord SD (2013) KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury. Am J Kidney Dis 61:649–672. 10.1053/j.ajkd.2013.02.349 [DOI] [PubMed] [Google Scholar]
  • 19.Schwartz GJ, Munoz A, Schneider MF, Mak RH, Kaskel F, Warady BA, Furth SL (2009) New equations to estimate GFR in children with CKD. J Am Soc Nephrol 20:629–637. 10.1681/ASN.2008030287 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.De Souza VC, Rabilloud M, Cochat P, Selistre L, Hadj-Aissa A, Kassai B, Ranchin B, Berg U, Herthelius M, Dubourg L (2012) Schwartz formula: is one k-coefficient adequate for all children? PLoS One 7:e53439. 10.1371/journal.pone.0053439 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Pierce CB, Muñoz A, Ng DK, Warady BA, Furth SL, Schwartz GJ (2021) Age- and sex-dependent clinical equations to estimate glomerular filtration rates in children and young adults with chronic kidney disease. Kidney Int 99:948–956. 10.1016/j.kint.2020.10.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J, CKD-EPI (2009) A new equation to estimate glomerular filtration rate. Ann Intern Med 150:604–612. 10.7326/0003-4819-150-9-200905050-00006
  • 23.Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, Kusek JW, Van Lente F; Chronic Kidney Disease Epidemiology Collaboration (2006) Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 145:247–254. 10.7326/0003-4819-145-4-200608150-00004 [DOI] [PubMed] [Google Scholar]
  • 24.Pottel H, Hoste L, Dubourg L et al (2016) An estimated glomerular filtration rate equation for the full age spectrum. Nephrol Dial Transplant 31:798–806. 10.1093/ndt/gfv454 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Pottel H, Björk J, Courbebaisse M et al (2021) Development and validation of a modified full age spectrum creatinine-based equation to estimate glomerular filtration rate : a cross-sectional analysis of pooled data. Ann Intern Med 174:183–191. 10.7326/m20-4366 [DOI] [PubMed] [Google Scholar]
  • 26.Nyman U, Grubb A, Larsson A, Hansson LO, Flodin M, Nordin G, Lindström V, Björk J (2014) The revised Lund-Malmö GFR estimating equation outperforms MDRD and CKD-EPI across GFR, age and BMI intervals in a large Swedish population. Clin Chem Lab Med 52:815–824. 10.1515/cclm-2013-0741 [DOI] [PubMed] [Google Scholar]
  • 27.Musso CG, Álvarez-Gregori J, Jauregui J, Macías-Núñez JF (2016) Glomerular filtration rate equations: a comprehensive review. Int Urol Nephrol 48:1105–1110. 10.1007/s11255-016-1276-1 [DOI] [PubMed] [Google Scholar]
  • 28.Koo TK, Li MY (2016) A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 15:155–163. 10.1016/j.jcm.2016.02.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Akoglu H (2018) User’s guide to correlation coefficients. Turk J Emerg Med 18:91–93. 10.1016/j.tjem.2018.08.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Schober P, Boer C, Schwarte LA (2018) Correlation coefficients: appropriate use and interpretation. Anesth Analg 126:1763–1768. 10.1213/ANE.0000000000002864 [DOI] [PubMed] [Google Scholar]
  • 31.Pizzo H, Nguyen J, Schwartz GJ, Wesseling-Perry K, Ettenger R, Chambers ET, Weng P (2024) Comparison of estimated GFR using cystatin C versus creatinine in pediatric kidney transplant recipients. Pediatr Nephrol 39:2177–2186. 10.1007/s00467-024-06316-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Tsampalieros A, Lepage N, Feber J (2011) Intraindividual variability of the modified Schwartz and novel CKiD GFR equations in pediatric renal transplant patients. Pediatr Transplant 15:760–765. 10.1111/j.1399-3046.2011.01568.x [DOI] [PubMed] [Google Scholar]
  • 33.Bökenkamp A, Domanetzki M, Zinck R, Schumann G, Byrd D, Brodehl J (1999) Cystatin C serum concentrations underestimate glomerular filtration rate in renal transplant recipients. Clin Chem 45:1866–1868 [PubMed] [Google Scholar]
  • 34.Blufpand HN, Westland R, van Wijk JA, Roelandse-Koop EA, Kaspers GJ, Bökenkamp A (2013) Height-independent estimation of glomerular filtration rate in children: an alternative to the Schwartz equation. J Pediatr 163:1722–1727. 10.1016/j.jpeds.2013.08.031 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Graphical abstract (PPTX 258 KB) (258.7KB, pptx)

Articles from Pediatric Nephrology (Berlin, Germany) are provided here courtesy of Springer

RESOURCES