International Journal of Nephrology

International Journal of Nephrology / 2015 / Article

Research Article | Open Access

Volume 2015 |Article ID 156484 | 8 pages | https://doi.org/10.1155/2015/156484

Spot Urine Estimations Are Equivalent to 24-Hour Urine Assessments of Urine Protein Excretion for Predicting Clinical Outcomes

Academic Editor: Francesca Mallamaci
Received01 Sep 2014
Revised04 Dec 2014
Accepted04 Dec 2014
Published08 Jan 2015

Abstract

Background. The use of spot urine protein to creatinine ratios in estimating 24 hr urine protein excretion rates for diagnosing and managing chronic kidney disease (CKD) predated the standardization of creatinine assays. The comparative predictive performance of spot urine ratios and 24 hr urine collections (of albumin or protein) for the clinical outcomes of CKD progression, end-stage renal disease (ESRD), and mortality in Asians is unclear. We compared 4 methods of assessing urine protein excretion in a multiethnic population of CKD patients. Methods. Patients with CKD () provided 24 hr urine collections followed by spot urine samples the next morning. We created multiple linear regression models to assess the factors associated with GFR decline (median follow-up: 37 months, IQR 26–41) and constructed Cox proportional-hazards models for predicting the combined outcome of ESRD and death. Results. The linear regression models showed that 24 hr urine protein excretion was most predictive of GFR decline but all other methods were similar. For the combined outcomes of ESRD and death, the proportional hazards models had similar predictive performance. Conclusions. We showed that all methods of assessments were comparable for clinical end-points, and any method can be used in clinical practice or research.

1. Background

The estimation of 24-hour urine protein excretion or 24-hour urine albumin excretion using urine protein to creatinine ratio (UPCR) and urine albumin to creatinine ratio (UACR), respectively, is well established in clinical practice and promulgated by practice guidelines [1]. The original studies were, however, performed prior to the standardization of creatinine assays [24]. The standardized creatinine values may differ by 5–20% from values obtained by methods not calibrated or traceable to the isotope dilution mass spectrometry (IDMS) standard [5]. As an example, if standardized creatinine is 20% higher than previously measured, UPCR or UACR may be 17% lower on estimation. Conversely if standardized creatinine is 10% lower, then UPCR or UACR may be 11% higher on estimation. Therefore, it is clear that creatinine standardization has important implications for the proteinuria estimation equations and their application in clinical practice. The measurement of proteinuria and albuminuria has not been standardized [6]. Moreover, the different assay methods may result in differences in the concentrations obtained. Research studies and clinical practice may favor UPCR over UACR for reason of cost, and there are no reliable methods to convert ratios [7]. However, the new iteration of the clinical practice guidelines for the identification and classification of chronic kidney disease incorporated urine albumin to creatinine ratio as part of risk stratification [8, 9]. In this study, we assessed the correlation of early morning spot urine tests to 24-hour urine protein and albumin excretion, provided conversion equations developed from a population of patients with a variety of chronic kidney disease, and compared their predictive effects for GFR decline, end-stage renal disease (ESRD), and mortality.

2. Methods

We used data from the Asian Kidney Disease Study as previously described [10]. Briefly, stable CKD patients were recruited from outpatient clinics for a study on glomerular filtration rates (GFR). They collected 24 hr urine collections and presented the next day for a GFR measurement and also provided early morning urine and blood samples. The 24 hr urine collection and early morning spot urine were tested for protein, albumin, and creatinine concentrations. We performed assays on the Siemens Advia 2400 (http://www.siemens.com/). Urine protein was measured using a pyrogallol-based assay calibrated to the manufacturer’s internal standard. Urine albumin was measured using a PEG-enhanced immunoturbidimetric assay and was also calibrated to an internal standard. Creatinine was measured by an enzymatic method (creatininase) in a central laboratory accredited by the College of American Pathologists and the assay was calibrated with manufacturer-provided materials traceable to standardized creatinine (National Institute for Standards and Technology Standard Reference Material 967) measured by isotope dilution mass spectrometry (as recommended by the National Kidney Disease Education Program, http://www.nkdep.nih.gov/). To extract the last known serum creatinine, the hospital clinical laboratory database was reviewed. All participants were cross-referenced with our ESRD database for date of dialysis initiation or death. All longitudinal follow-up data for this study were correct till January 15, 2012. This study was approved by the National Healthcare Group, Domain-Specific Review Board (D/07/524 and 2007/00225).

3. Statistics

Where appropriate, variables were naturally Log-transformed before linear regression to correct for nonnormal distribution and nonconstant variability of observed points around the regression line. We used Bland-Altman analysis of agreement to assess UACR and UPCR in predicting 24 hr urine albumin excretion and 24 hr urine protein excretion, respectively. For comparisons with earlier studies, we used non-SI units in calculating the urine protein or albumin to creatinine ratios [3, 4]. Conversion equations between UACR, UPCR, 24 hr urine protein excretion, and 24 hr urine albumin excretion were developed. To assess these equations, an external dataset of 45 participants with similarly collected 24 hr urine sample followed by an early morning spot urine sample the next day was used as a validation test dataset. The predicted variable was compared to the measured values by Pearson correlation and Wilcoxon signed rank test. Because clinical practice and research analysts commonly use UACR and UPCR to estimate the respective 24 hr urine excretion rates above the clinically significant thresholds of albumin >300 mg/day and protein >0.5 g/day, we also analyzed the measured (main derivation dataset, ) and equation-predicted UACR and UPCR (validation dataset, ) for their predictive abilities in terms of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) [11, 12].

We estimated GFR using the CKD-EPI equation [13]. To compare the predictive performance of each method of urine protein or albumin assessment with GFR decline, we developed models using linear regression. We used stepwise linear regression by value threshold for entry or exit from the models in a mixed direction with no rules to select the variables for models predicting GFR decline. The initial variables screened were age, gender, ethnicity, initial serum creatinine, initial serum urea, smoking history, medical history of diabetes, hypertension, coronary artery disease, peripheral artery disease, systolic blood pressure, diastolic blood pressure, body mass index, serum albumin, UACR, UPCR, 24 hr urine protein excretion, and 24 hr urine albumin excretion. The urine estimation or measurement method in the best model was substituted in turn for three additional models for comparisons. Variables with values and ethnicity were included in the final models. We assessed the linear regression models by reviewing the , AIC (Akaike information criterion), and BIC (Bayesian information criterion) [14, 15]. Since always will increase with an increasing number of predictor variables in a model, we select our model using the AIC and BIC which penalizes models for having large number of predictor variables. AIC might favor more complex models and overfit, while BIC may select a parsimonious model and underfit. For our model assessments, the lowest AIC and BIC criteria indicate the best model. Similarly, we also assessed the predictive performance of urine protein estimations for the combined end-point of ESRD and death using Cox-proportional hazards modeling. Analyses were performed on JMP 10 (Cary, NC, USA) and R (http://www.r-project.org/).

4. Results

There were 232 patients with characteristics as shown in Table 1. Four patients had 24-hour urine protein excretion >3.5 g. The correlation of spot urine estimation ratios (UPCR and UACR to 24 hr urine protein and albumin excretion, resp.) is shown in Figure 1. UACR appears to predict 24 hr urine albumin excretion better as the slope is closer to 1. The prediction equations are


Age (years)58.4 ± 12.8
Male (, %)120 (51.7)
Ethnicity (, %)
 All232 (100)
 Chinese94 (40.5)
 Malay74 (31.9)
 Indian56 (24.1)
 Others8 (0.03)
Height (m)1.59 ± 0.09
Weight (kg)70.3 ± 15.9
Body mass index (kg/m2)27.6 ± 5.5
Body surface area (m2)1.72 ± 0.21
Measured GFR (mL/min/1.73 m2)51.7 ± 27.5
Serum creatinine (µmol/L)153 ± 92
Serum protein (g/L)72.2 ± 5.7
Serum albumin (g/L)41.8 ± 3.2
Serum urea (mmol/L)8.35 ± 6.35
24 hr urine volume (L)1.76 ± 0.78
24 hr urine protein (g)0.6 ± 0.9
24 hr urine albumin (mg)383.7 ± 685.9
24 hr urine creatinine (mmol)8.2 ± 3.6
Spot urine protein (g/L)0.64 ± 1.04
Spot urine albumin (mg/L)413.3 ± 721.5
Spot urine creatinine (mmol/L)6.9 ± 4.6
UPCR (mg/mg)1.03 ± 1.87
UACR (mg/mg)0.68 ± 1.3
Diabetes (, %)119 (51)
Hypertension (, %)192 (83)
Cause of CKD (, %)
 Hypertension115 (49.6)
 Diabetic nephropathy54 (23.3)
 Glomerular disease38 (16.4)
 Polycystic kidney disease6 (2.6)
 Obstructive kidney disease4 (1.7)
 Other or unknown cause15 (6.5)

Data shown as mean ± SD or frequency (percentage).

UPCR has poorer correlation to 24 hr urine protein excretion (, 95% CI 0.75–0.84) than UACR has to 24 hr urine albumin excretion (, 95% CI 0.82–0.89). UACR is highly correlated to UPCR (, 95% CI 0.99-0.99) (Figure 2). 24 hr urine albumin excretion is also highly correlated to 24 hr urine protein excretion (, 95% CI 0.99-0.99). UACR is less correlated with 24 hr urine protein excretion (, 95% CI 0.72–0.83) than UPCR is with 24 hr urine albumin excretion (, 95% CI 0.76–0.85). The sensitivity and specificity of UACR of between 30 and 300 mg/mg for predicting 24 hr urine albumin excretion of between 30 and 300 mg (microalbuminuria) are 0.69 and 0.77, respectively. The PPV and NPV are 0.63 and 0.81, respectively. The sensitivity and specificity of UACR of >300 mg/mg for predicting 24 hr urine albumin excretion of >300 mg (macroalbuminuria) are 0.94 and 0.84, respectively. The PPV and NPV are 0.70 and 0.97, respectively. The sensitivity and specificity of UPCR of >0.5 mg/mg for predicting 24 hr urine protein excretion of >0.5 g/day are 0.90 and 0.81, respectively. The PPV and NPV are 0.66 and 0.95, respectively.

The various conversion equations derived from the main study data () are in Table 2. Using the external validation dataset () for assessing the performance of the conversion equations, predicted spot urine ratios were highly correlated but significantly differed from measured spot urine values. UPCR correlated with “predicted 24 hr urine protein excretion” (, 95% CI 0.41–0.78) better than UACR with “predicted 24 hr urine albumin excretion” (, 95% CI 0.31–0.73). The sensitivity and specificity of “predicted-UACR,” calculated from UPCR, of >300 mg/mg for predicting 24 hr urine albumin excretion of >300 mg (macroalbuminuria) are 0.059 and 1.0, respectively. The PPV and NPV are 1.0 and 0.64, respectively. The sensitivity and specificity of “predicted-UACR,” calculated from 24 hr urine protein excretion, of >300 mg/mg for predicting 24 hr urine albumin excretion of >300 mg (macroalbuminuria) are 1.0 and 0.89, respectively. The PPV and NPV are 0.85 and 1.0, respectively. The sensitivity and specificity of “predicted-UPCR,” calculated from UACR, of >0.5 mg/mg for predicting 24 hr urine protein excretion of >0.5 g/day are 0.1 and 1.0, respectively. The PPV and NPV are 1.0 and 0.58, respectively.

(a) Conversion predicting equations

Equations value

Log ⁡24 hr urine protein excretion (g) = −0.617019 + 0.7150918 × Log UPCR (mg/mg)<0.001
Log ⁡24 hr urine albumin excretion (g) = −0.800153 + 0.8257142 × Log UACR (mg/mg)<0.001
Log UACR (mg/mg) = −0.656352 + 1.3881178 × Log UPCR (mg/mg)<0.001
Log UPCR (mg/mg) = 0.3216439 + 0.6394674 × Log UACR (mg/mg)<0.001
Log UACR (mg/mg) = −0.270587 + 1.2870223 × Log ⁡24 hr urine protein excretion (g)<0.001

(b) Performance of conversion equations using an external validation dataset ( = 45)

PredictorPredicted variablePredicted valueMeasured value value Correlation (95% CI)

UPCR (mg/mg)24 hr urine protein excretion (g)0.74 (0.67–0.82)0.38 (0.14–0.92)0.00510.63 (0.41–0.78)
UACR (mg/mg)24 hr urine albumin excretion (g)0.91 (0.87–0.94)0.21 (0.02–0.58)<0.0010.55 (0.31–0.73)
UPCR (mg/mg)UACR (mg/mg)0.01 (0.00–0.03)0.03 (0.00–0.08)<0.0010.97 (0.95–0.98)
UACR (mg/mg)UPCR (mg/mg)0.15 (0.04–0.29)0.05 (0.02–0.13)<0.0010.95 (0.91–0.97)
24 hr urine protein excretion (g)UACR (mg/mg)0.22 (0.06–0.69)0.03 (0.00–0.08)<0.0010.81 (0.68–0.89)

Reported as median (25th–75th percentile); value of the difference between predicted and measured values; Pearson correlation (95% confidence interval).

There were 225 patients with available follow-up serum creatinine to determine estimated GFR decline. By stepwise linear regression, we developed 4 final models to compare the performance of UACR, UPCR, and 24 hr urine protein or albumin excretion (Table 3). In all models, the method of urine protein or albumin assessment and the initial serum urea were significant. The model containing 24 hr urine protein excretion predicted GFR decline best. All the models had good predictive performance for GFR decline, with the best performance in the model that included 24 hr urine protein excretion, followed in order by 24 hr urine albumin excretion, UPCR, and UACR.


95% CIEstimateVariables valueAICBIC

Model 10.100.028 to 0.172−0.211Log 24 hr urine protein excretion<0.001524.97545.65
0.311Log serum urea0.001

Model 20.0950.025 to 0.166−0.127Log 24 hr urine albumin excretion<0.001526.13546.81
0.291Log serum urea0.002

Model 30.0710.009 to 0.133−0.148Log UPCR0.001532.38553.06
0.289Log serum urea0.003

Model 40.0660.005 to 0.126−0.092Log UACR0.002533.65554.33
0.275Log serum urea0.0072

All models adjusted for ethnicity.

The median follow-up was 37 months (IQR 26–41). There were 19 patients who reached ESRD (9/19, 47% women) and 9 (3/9, 33% women) who died during the follow-up period (4 had ESRD before dying). Patients who reached ESRD were of similar age (59.8 ± 10.3 versus 58.8 ± 13.0 years) but had a higher serum creatinine (311 ± 113 versus 138 ± 75 μmol/L) and 24 hr urine albumin (1515 ± 1251 versus 283 ± 504 mg). We created 4 Cox proportional hazard models to compare the performance of the various proteinuria assessments for predicting the combined end-point of ESRD and death (Table 4). All models, which included the standard adjusters (age, gender, and ethnicity), were significant for all methods of proteinuria assessment (all ) and Log- serum urea (all ).


Model valueTerm valueRisk ratioLower 95% CIUpper 95% CI
−Log likelihood

Model 1−79.45<0.001Log 24 hr urine protein<0.0013.872.376.31
Log serum urea<0.00117.626.2749.54
Age0.2250.970.931.02
Gender0.3101.660.624.41
Malay ethnicity0.1552.080.765.72
Indian and others ethnicity0.5721.430.414.93

Model 2−80.03<0.001Log 24 hr urine albumin<0.0012.901.874.51
Log serum urea<0.00116.946.2545.89
Age0.280.980.941.02
Gender0.311.650.624.40
Malay ethnicity0.162.070.755.69
Indian and others ethnicity0.631.360.394.72

Model 3−77.30<0.001Log UPCR<0.0013.822.406.07
Log serum urea<0.00123.527.6072.85
Age0.070.960.921.00
Gender0.013.891.3311.40
Malay ethnicity0.191.940.725.22
Indian and others ethnicity0.401.710.486.04

Model 4−79.38<0.001Log UACR<0.0012.981.934.58
Log serum urea<0.00120.677.0960.23
Age0.120.970.921.01
Gender0.033.231.159.12
Malay ethnicity0.181.960.735.24
Indian and others ethnicity0.421.690.486.05

All models included the standard terms of age, gender, and ethnicity. Risk ratios for continuous variables are per unit change in the regressor, and for categorical variables: gender (men/women), with women being the reference level.

5. Discussion

This is the first prospective study of a multiethnic Asian population with a wide variety of CKD (diabetic and nondiabetic) patients that simultaneously evaluates early morning spot urine prediction ratios to 24 hr urine collections, while accounting for the standardization of the creatinine assay. Previous studies were retrospective and did not have simultaneous collection of spot urine and 24 hr urine collections [16]. Clinical practice and research involving CKD patients are highly dependent on the use of urine protein or albumin to creatinine ratios as estimates of their respective 24 hr urine excretions. Yet, many are unaware of the implications of the use of urine ratios [17, 18]. Fundamentally, the main questions that need to be answered are as follows. Does UACR or UPCR predict 24 hr urine albumin or protein excretion? Which spot urine ratio predicts 24 hr urine excretion better? (3) In our setting, how do the ratios relate to the 24 hr urine excretions? Especially, now that we have standardized creatinine assays but not for albumin and protein assays. And, of course, (4) which parameter (ratios or 24 hr urine excretions) predicts longitudinal outcomes data better (GFR declines, ESRD, or mortality in CKD patients)?

Our study shows that UACR is correlated to 24 hr urine albumin excretion better than UPCR. We did not find as high a correlation for UPCR as the earlier studies [3, 4]. This may be partly due to creatinine calibration which results in a systematically larger ratio. The clinical practice guideline for the identification and classification of CKD incorporates UACR in addition to estimated GFR for staging CKD. Our study derived the conversion equations for clinical research or practice needing spot or 24 hr urine albumin or protein excretions, specific to the assay methods. On average, predicted spot urine ratios were reasonably correlated but were significantly different from measured values. Many clinical and research databases contain both UACR and UPCR in the same patients; analysts often apply conversion equations for the purposes of analysis. But the sensitivity of “predicted UACR” or “predicted UPCR” for identifying clinically significant 24 hr urine excretion rates is poor. Therefore, we do not recommend using the conversion equations of UACR to UPCR, and vice versa. However, the performance of “predicted-UACR” calculated from 24 hr urine protein excretion appears to be acceptable for identifying clinically significant proteinuria of >0.5 g/day. Nonetheless, in clinical practice, it is currently not recommended to interchangeably convert urine albumin and urine protein concentrations [11].

Contrary to the findings by Ruggenenti et al. in their cohort of only nondiabetic patients, we did not find that UPCR predicted GFR decline better than 24 hr urine protein excretion [4]. But, in all our models, all methods of assessing urine protein excretion rates were significant for predicting GFR decline, ESRD, and mortality. In our opinion, this supports the current practice of using spot urine tests as estimates of assessing 24 hr urine protein or albumin excretion [11, 16]. And all of the methods of assessments are significantly associated with predicting clinical outcomes in a multiethnic Asian population with different types of CKD, making our results more generalizable and supportive of clinical practice and research.

The strengths of our study include a fairly large multiethnic Asian population comprising of both diabetic and nondiabetic CKD, with systematically collected spot and 24 hr urine when compared to previous studies [3, 4]. We also used turbidimetry, a robust method for determining albumin concentrations, although some others advocate using nephelometry as the preferred albumin assay method [7]. Our study is also limited by fewer CKD patients with nephrotic-range proteinuria (24 hr urine protein excretion >3.5 g). However, it had been shown that spot urine estimates were less accurate at higher levels of proteinuria [3, 4]. The urine estimation to 24 hr measurements may be less accurate since the urine collections are self-directed. Conversely, others would also argue that the derived equations are more reflective of actual practice, and, therefore, prediction equations and longitudinal analyses will be more valid and generalizable to clinical practice. Moreover, in practice, we are generally interested in categories of proteinuria excretion, namely, <1 g/day, 1 to 3 g/day, and >3 g/day, and that, at higher levels, one should obtain a 24 hr urine collection to ascertain the parameters for initiating treatment of CKD. The sample size limits the accuracy of the multivariate regression models, and further definitive studies are required.

6. Conclusions

In summary, we appraised the use of urine spot ratios for assessing urine protein excretion rates and developed helpful conversion equations for both clinical research and practice. We showed that all methods of urine protein assessment were comparable for clinical end-points, and any method can be used in clinical practice or research.

Conflict of Interests

The authors have no conflict of interests to report.

Authors’ Contribution

Boon Wee Teo, Ping Tyug Loh, and Evan J. C. Lee conceived, supervised, and conducted the study. Boon Wee Teo and Hui Xu recruited the participants and collected the data. Boon Wee Teo, Ping Tyug Loh, Peh Joo Ho, and Kwok Pui Choi analyzed the data. Sharon Saw, Sunil Sethi, Hui Xu, and Boon Wee Teo supervised the laboratory measurements of specimens. Weng Kin Wong, Boon Wee Teo, and Ping Tyug Loh wrote the initial draft of the paper. All authors reviewed and edited the paper.

Acknowledgments

Majority funding for the Asian Kidney Disease Study was awarded by the National University of Singapore Yong Loo Lin School of Medicine Faculty Research Committee, and it is funded in part by the National Medical Research Council (block vote) and the National Kidney Foundation of Singapore. Boon Wee Teo is a recipient of the NHG-NUS Clinical Leadership in Research Award 2007 and the NMRC Overseas Fellowship Award 2009 for research studies in kidney function. Sponsors had no role in the study design, data collection, analysis, and interpretation. Other coinvestigators of our study section include Dr. Jialiang Li, Dr. Borys Shuter, and Dr. Arvind Kumar Sinha. Dr. Li supervised the initial statistical analysis. Abstracts of this study were presented in part at Singapore Society of Nephrology Annual Scientific Meeting 2012 by Dr. Weng Kin Wong.

References

  1. National Kidney Foundation, “K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification,” American Journal of Kidney Diseases, vol. 39, no. 2, supplement 1, pp. S1–S266, 2002. View at: Google Scholar
  2. G. L. Myers, W. G. Miller, J. Coresh et al., “Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program,” Clinical Chemistry, vol. 52, no. 1, pp. 5–18, 2006. View at: Publisher Site | Google Scholar
  3. R. A. Rodby, R. D. Rohde, Z. Sharon, M. A. Pohl, R. P. Bain, and E. J. Lewis, “The urine protein to creatinine ratio as a predictor of 24-hour urine protein excretion in type 1 diabetic patients with nephropathy,” American Journal of Kidney Diseases, vol. 26, no. 6, pp. 904–909, 1995. View at: Publisher Site | Google Scholar
  4. P. Ruggenenti, F. Gaspari, A. Perna, and G. Remuzzi, “Cross sectional longitudinal study of spot morning urine protein: creatinine ratio, 24 hour urine protein excretion rate, glomerular filtration rate, and end stage renal failure in chronic renal disease in patients without diabetes,” British Medical Journal, vol. 316, no. 7130, pp. 504–509, 1998. View at: Publisher Site | Google Scholar
  5. A. S. Levey, J. Coresh, T. Greene et al., “Expressing the modification of diet in renal disease study equation for estimating glomerular filtration rate with standardized serum creatinine values,” Clinical Chemistry, vol. 53, no. 4, pp. 766–772, 2007. View at: Publisher Site | Google Scholar
  6. W. G. Miller, D. E. Bruns, G. L. Hortin et al., “Current issues in measurement and reporting of urinary albumin excretion,” Clinical Chemistry, vol. 55, no. 1, pp. 24–38, 2009. View at: Publisher Site | Google Scholar
  7. P. Deurenberg and M. Deurenberg-Yap, “Validation of skinfold thickness and hand-held impedance measurements for estimation of body fat percentage among Singaporean Chinese, Malay and Indian subjects,” Asia Pacific Journal of Clinical Nutrition, vol. 11, no. 1, pp. 1–7, 2002. View at: Publisher Site | Google Scholar
  8. A. S. Levey, P. E. de Jong, J. Coresh et al., “The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report,” Kidney International, vol. 80, no. 1, pp. 17–28, 2011. View at: Publisher Site | Google Scholar
  9. “Kidney disease: improving global outcomes (KDIGO) CKD work group: KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease,” Kidney International Supplements, vol. 3, pp. 1–150, 2013. View at: Google Scholar
  10. B. W. Teo, H. Xu, D. Wang et al., “GFR estimating equations in a multiethnic asian population,” American Journal of Kidney Diseases, vol. 58, no. 1, pp. 56–63, 2011. View at: Publisher Site | Google Scholar
  11. D. W. Johnson, G. R. D. Jones, T. H. Mathew et al., “Chronic kidney disease and measurement of albuminuria or proteinuria: a position statement,” The Medical Journal of Australia, vol. 197, no. 4, pp. 224–225, 2012. View at: Publisher Site | Google Scholar
  12. A. S. Levey, K.-U. Eckardt, Y. Tsukamoto et al., “Definition and classification of chronic kidney disease: a position statement from kidney disease: improving global outcomes (KDIGO),” Kidney International, vol. 67, no. 6, pp. 2089–2100, 2005. View at: Publisher Site | Google Scholar
  13. A. S. Levey, L. A. Stevens, C. H. Schmid et al., “A new equation to estimate glomerular filtration rate,” Annals of Internal Medicine, vol. 150, no. 9, pp. 604–612, 2009. View at: Publisher Site | Google Scholar
  14. H. Akaike, “A new look at the statistical model identification,” IEEE Transactions on Automatic Control, vol. 19, no. 6, pp. 716–723, 1974. View at: Publisher Site | Google Scholar | MathSciNet
  15. G. Schwarz, “Estimating the dimension of a model,” The Annals of Statistics, vol. 6, no. 2, pp. 461–464, 1978. View at: Publisher Site | Google Scholar | MathSciNet
  16. S. Methven, M. S. MacGregor, J. P. Traynor, M. Hair, D. S. J. O'Reilly, and C. J. Deighan, “Comparison of urinary albumin and urinary total protein as predictors of patient outcomes in CKD,” American Journal of Kidney Diseases, vol. 57, no. 1, pp. 21–28, 2011. View at: Publisher Site | Google Scholar
  17. J. Fotheringham, M. J. Campbell, D. G. Fogarty, M. El Nahas, and T. Ellam, “Estimated albumin excretion rate versus urine albumin-creatinine ratio for the estimation of measured albumin excretion rate: derivation and validation of an estimated albumin excretion rate equation,” American Journal of Kidney Diseases, vol. 63, no. 3, pp. 405–414, 2014. View at: Publisher Site | Google Scholar
  18. L. A. Inker, “Albuminuria: time to focus on accuracy,” American Journal of Kidney Diseases, vol. 63, no. 3, pp. 378–381, 2014. View at: Publisher Site | Google Scholar

Copyright © 2015 Boon Wee Teo et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

11463 Views | 903 Downloads | 8 Citations
 PDF  Download Citation  Citation
 Download other formatsMore
 Order printed copiesOrder

We are committed to sharing findings related to COVID-19 as quickly and safely as possible. Any author submitting a COVID-19 paper should notify us at help@hindawi.com to ensure their research is fast-tracked and made available on a preprint server as soon as possible. We will be providing unlimited waivers of publication charges for accepted articles related to COVID-19.