- Email: [email protected]
Usage of spot urine protein to creatinine ratios in the evaluation of preeclampsia
Research
www. AJOG.org
OBSTETRICS
Usage of spot urine protein to creatinine ratios in the evaluation of preeclampsia Thomas L. Wheeler II, MD; Dawn W. Blackhurst, DrPH; Eric H. Dellinger, MD; Patrick S. Ramsey, MD, MSPH OBJECTIVE: The objective of the study was to prospectively compare spot urine protein to creatinine (P:C) ratios with 24 hour urine collections for protein in women being evaluated for preeclampsia. STUDY DESIGN: A spot urine P:C ratio was obtained at the beginning of 24 hour urine collections from 126 patients admitted to evaluate for preeclampsia. Correlation between the spot P:C ratio with the 24 hour urine collections was calculated. Receiver operator characteristic curves were constructed to determine best P:C cutoffs for 300 mg and 5000 mg protein per 24 hours.
RESULTS: Random spot P:C ratios were strongly correlated with 24 hour urine protein levels (Pearson r ⫽ 0.88). The optimal P:C cut-offs were 0.21 (300 mg per 24 hours) and 3.0 (5000 mg per 24 hours). A P:C ratio of less than 0.21 (300 mg per 24 hours) had a negative predictive value (NPV) of 83.3% and a P:C ratio of less than 3.0 (5000 mg per 24 hours) had 100% NPV. CONCLUSION: Urine spot P:C ratio correlated well with 24 hour urine collections for protein but was not justified as a substitute for timed collections.
Key words: creatinine, measurement, preeclampsia, protein
Cite this article as: Wheeler TL, Blackhurst DW, Dellinger EH, et al. Usage of spot urine protein to creatinine ratios in the evaluation of preeclampsia. Am J Obstet Gynecol 2007;196;465.e1-465.e4.
A
ssessment of urinary protein is an important component of the work-up for preeclampsia. In-patient 24 hour urine collections for protein, the gold standard, are cumbersome and time consuming, typically requiring a 2 night stay to get results. One simpler alternative, a random spot urine protein, is hypothetically subject to inaccuracies sec-
From the Department of Obstetrics and Gynecology, University of Alabama at Birmingham (Drs Wheeler and Ramsey), Birmingham, AL; and Department of Obstetrics and Gynecology, Greenville Hospital Systems University Medical Center (Dr. Blackhurst and Dellinger), Greenville, SC. Received July 5, 2006; revised September 6, 2006; accepted October 24, 2006. Reprint requests: Thomas L. Wheeler II, MD, Fellow, Female Pelvic Medicine and Reconstructive Pelvic Surgery, University of Alabama at Birmingham, Division of Women’s Pelvic Medicine and Reconstructive Surgery, 619 19 Street South, NHB 219, Birmingham, AL 35249-7333; [email protected]. 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2006.10.892
See Journal Club, page 491
ondary to time being the denominator for this value.1 Likewise, a spot urinary dipstick measurement of protein has not been shown to be reliable.2,3 However, random spot sampling for a urine protein to creatinine (P:C) ratio has been investigated as an alternative. With results in hours and easier collection, a spot P:C ratio is a more efficient test than a 24 hour collection for proteinuria assessment. Because the excretion of protein and creatinine are expressed per time, the ratio cancels out the time factor and allows for an estimation of 24 hour protein excretion if stable renal function is assumed.1 This ratio also normalizes urinary protein concentration to that of creatinine. In fact, in the literature on the medical general nonpregnant population, the spot P:C ratio performs well at assessing proteinuria, correlating with 24-hour protein excretion in the following populations: systemic lupus erythematosus, glomerular disease, renal transplant, and ages older than 5 years.4-8 The spot P:C ratio has been evaluated for use in pregnancy with mixed results.9-21 Some investigators have supported its use by demonstrating high correlations between spot P:Cs and 24 hour collections,9-17 whereas others have
stressed caution in its application, especially because renal function can destabilize within hours in preeclampsia.18-21 Recently a systematic review of spot P:C ratio use in the work-up of preeclampsia demonstrated strong evidence, based on the cumulative negative likelihood ratio of 0.14, that the spot P:C ratio identified patients unlikely to have significant proteinuria on a 24-hour collection. However, the variability in cut off values between studies did not allow for a recommendation of a critical threshold P:C ratio for 300 mg per 24 hours. The objectives of this study were to determine the strength of correlation between these 2 approaches to urinary protein assessment and to determine the optimal P:C ratios that best correspond to critical values on a 24 hour urine.
M ATERIALS AND M ETHODS Institutional review board approval was obtained from Greenville Hospital Systems University Medical Center (Greenville, SC). Spot P:C samples were gathered from 154 patients meeting inpatient admission criteria for the evaluation of preeclampsia, which was in general new-onset persistent hypertension, worsening hypertension, or proteinuria between December 2000 and July 2002
MAY 2007 American Journal of Obstetrics & Gynecology
465.e1
Research
Obstetrics
www.AJOG.org C OMMENT
TABLE
Test performance of spot P:C ratio assessment for 24 hour urine protein results 24 hour urine collection quantitative protein
Optimal spot P:C (per ROC analysis)
300 mg
0.21
86.8
77.6
81.9
83.3
1000 mg
0.46
87.5
82.4
53.8
96.6
2000 mg
0.82
100
94.8
62.5
5000 mg
3.0
100
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
100
..............................................................................................................................................................................................................................................
at the Greenville Hospital Systems University Medical Center. Following American College of Obstetricians and Gynecologists guidelines, new-onset hypertension was a systolic blood pressure greater than 140 mm Hg or a diastolic blood pressure greater than 90 mm Hg after 20 weeks’ gestation in a previously normotensive patient, whereas worsening hypertension was an increase in blood pressure from baseline taken before 20 weeks’ gestation.23 Samples were obtained at the beginning of the 24 hour urine collection. No firstmorning voids were used. Women who had bacteriuria on microscopy or who were on more than 24 hours’ bed rest, because of a potential poor correlation between spot P:C and 24 hour urine collections for protein after prolonged recumbency, were excluded.1,15 Urinary protein was determined by the Biuret method. 24 Urinary creatinine was determined by the 2-point rate method,25 and aliquots were analyzed by a Johnson & Johnson Vitros 250 (Johnson & Johnson Clinical Diagnostics Inc, Rochester, NY). The strength of correlation (Pearson correlation coefficient, r) between the spot P:C ratio and 24 hour urine collection protein amount was calculated by generating a least squares regression line. Using protein values of 300, 1000, 2000, and 5000 mg on 24-hour urine collections, receiver operator characteristic (ROC) curves were constructed, and the best P:C cut-off points were calculated. The best P:C ratio cut-off was identified as the point on the left shoulder of the ROC curve that maximizes sensitivity and specificity. As a measure of accuracy, the area under the ROC curves were calculated with 1 representing a perfect test 465.e2
100
100
100
and 0.5 a worthless test (ie, as accurate as flipping a fair coin). Therefore, as the area under the curve approaches 1, the accuracy improves. Sensitivity, specificity, and negative (NPV) and positive predictive values (PPV) were calculated for each cut-off value. ROC curves were constructed, and data were analyzed using MedCalc software, version 7 (MedCalc Software, Mariakerke, Belgium).
R ESULTS Urine collections were completed on 81.8% (126 of 154) of women. The other 18.2% (28 of 154) of women were delivered before the completion of their 24hour urine collection, which was then not processed. For the 126 patients completing the assessment, the mean age was 26.6 ⫾ 5.8 years, and the mean estimated gestational age was 34.0 ⫾ 3.3 weeks, with 27% blacks, 72% whites, and 1% Hispanics. Parity was 56% nulliparous, 30% primiparous, 12% para 2, and 2% para 3. The test performance for the spot P:C ratio for the varying degrees of proteinuria noted on 24 hour urine collection are shown in the Table. The spot P:C ratio correlated with 24 hour urine protein collection results (r ⫽ 0.88). The optimal P:C ratio cut-offs, based on the ROC analysis, were 0.21 for 300 mg per 24 hours, 0.46 for 1000 mg per 24 hours, 0.82 for 2000 mg per 24 hours, and 3.0 for 5000 mg per 24 hours. The areas under the receiver operator curves for 300, 1000, 2000, and 5000 mg of protein on a 24 hour urine collection were 0.86, 0.91, 0.98, and 1.0, respectively (Figure).
American Journal of Obstetrics & Gynecology MAY 2007
Most studies have repeated good correlations between spot P:C ratios and 24 hour urine collections for protein with correlation coefficients (r) ranging between 0.80 and 0.97.11-17,20 Furthermore, spot urine P:C ratios correlated with 24 hour collections, even when the sample population included hypertensive and renal disorders.13 A spot P:C ratio has also been proposed to not be confounded by maternal age, gestational age, or parity (10,16,18,20) and has been correlated with adverse maternal and fetal outcomes.26 However, 2 recent studies determined lower correlation coefficients of r ⫽ 0.56 and r ⫽ 0.41 and cautioned against substituting spot P:C ratio for timed collections.18,19 Additionally, its use may be limited in women with pregestational diabetes and in ambulatory patients in the second and third trimesters.20,21 Along with the concerns about the strength of correlation, finding a spot P:C ratio value with acceptable PPVs and NPVs has been a challenge. ROC curves have been used to select the P:C ratio point that maximizes sensitivity and specificity. For example a P:C cut-off such as 0.19 has been described, but false negatives still occur with this method.11,19 To address this concern and limit false negatives, a P:C point with maximal sensitivity has been chosen, like 0.14,11 at the expense of specificity, which then results in increased misdiagnoses of preeclampsia with obvious ramifications to management. In fact, the recent systematic review analyzing spot P:C ratios in preeclampsia supports the use of a spot P:C ratio in ruling out significant proteinuria.23 However, no consensus for specific P:C cut-off values has been obtained. These uncertainties have added to the resistance of substituting it for timed collections when evaluating preeclampsia, in which renal function can decline within hours, unlike the usual time course of renal decline in the general nonpregnant medical population. Our study found a strong correlation (r ⫽ 0.88) between the spot P:C ratio and 24 hour urine protein results. Based on
Obstetrics
www.AJOG.org
FIGURE
Receiver operator characteristic curves for 24 hour urine protein results are shown
A, For 300 mg, the P:C ratio cut-off is greater than 0.21. B, For 1000 mg, the P:C ratio cut-off is greater than 0.46. C, For 2000 mg, the P:C ratio cut-off is greater than 0.82. D, For 5000 mg, the P:C ratio cut-off is 3.0 or greater.
our ROC analysis, the optimal P:C ratio of 0.21 was identified for 300 mg per 24 hours and 3.0 for 5000 mg per 24 hours. The areas under the ROC curves appeared to show that these cut-off spot P:C ratios were accurate in predicting the 24 hour urine collection results. However, 13.2% (9 of 68) of women with significant proteinuria (more than 300 mg per 24 hours) were missed by this P:C ratio cut off (ie, false-negative rate). We concluded the following: (1) spot P:C ratio is strongly correlated (r ⫽ 0.88) with 24 hour urine collections for protein, (2) the NPV for P:C ratios 0.21 of 83.3% is too low to be a substitute for 24 hour collections, and (3) despite a 100% NPV and PPV, the sample size for P:C ratio greater than 3.0 (5000 mg per 24 hours) of n ⫽ 5 was too small to justify eliminating the 24 hour collection for this value. Weaknesses of our study include the small sampling of women with greater than 5000 mg protein per 24 hours. This study also included patients with renal disease, chronic hypertension, and dia-
betes, in whom preexisting proteinuria could exist, but did not subgroup them for analysis. Hence, the impact of these conditions and preexisting proteinuria on the strength of correlation between the spot P:C and 24 hour urine collections for protein is unknown in this study subpopulation. We suggest that a clinical study with a larger sample size to determine the best P:C cut point for 5000 mg is needed. It is also important to clarify the impact of comorbid disease and/or preexisting proteinuria on the correlation between spot P:C and 24 hour urine collection for protein.22 f REFERENCES 1. Ginsberg JM, Chang BS, Matarese RA, Garella S. Use of single voided urine sample to estimate quantitative proteinuria. N Engl J Med 1982; 309:1543-6. 2. Phelan LK, Brown MA, Davis GK, Mangos G. A prospective study of the impact of automated dipstick urinalysis on the diagnosis of preeclampsia. Hypertens Pregnancy 2004;23: 135-42.
Research
3. Meyer NL, Mercer BM, Friedman SA, Sibai BM. Urinary dipstick protein: a poor predictor of absent or severe proteinuria. Am J Obstet Gynecol 1994;170:137-41. 4. Christopher-Stine L, Petri M, Astor, BC, Fine D. Urine protein-to-creatinine ratio is a reliable measure of proteinuria in lupus nephritis. J Rheumatol 2004;31:1557-9. 5. Chitalia VC, Kothari J, Wells EJ, et al. Costbenefit analysis and prediction of 24-hour proteinuria from the spot urine protein-creatinine ratio. Clin Nephrol 2001;55:436-47. 6. Xin G, Wang M, Jiao L, Xu G, Wang H. Protein-to-creatinine ratio in spot urine samples as a predictor of quantification of proteinuria. Clinica Chimica Acta 2004;350:35-9. 7. Torng S, Rigatto C, Rush DN, Nickerson P, Jeffery JR. The urine protein to creatinine ration (P/C) as a predictor of 24-hour urine protein excretion in renal transplant patients. Transplantation 2001;72:1453-6. 8. Tamura S, Shimizu T, Kawakatsu H, Tateishi S. Correlation between 24-hour urinary protein excretion and protein/creatinine ratio in the first voided morning urine samples. Jpn J Nephrol 2004;46:26-34. 9. Yamasmit W, Wongkitisophon K, Charoenvidhya D, Uerpairojkit B, Chaithongwongwatthana S. Correlation between random urinary protein-to creatinine ratio and quantitation of 24-hour proteinuria in preeclampsia. J Med Assoc Thai 2003;86:69-73. 10. Neithardt AB, Dooley SL, Borensztajn J. Predicition of 24-hour protein excretion in pregnancy with a single voided urine protein-to-creatinine ratio. Am J Obstet Gynecol 2002;186: 883-6. 11. Rodriguez-Thompson D, Lieberman E. Use of random urinary protein-to-creatinine ratio for the diagnosis of significant proteinuria during pregnancy. Am J Obstet Gynecol 2001; 185:808-11. 12. Ramos JG, Martins-Costa SH, Mathias MM, Guerin YL, Barros EG. Urinary protein/creatinine ratio in hypertensive pregnant women. Hypertens Pregnancy 1999;18:209-18. 13. Robert M, Sepandj F, Liston R, Dooley K. Random protein-creatinine ratio for the quantization of proteinuria in pregnancy. Obstet Gynecol 1997;90:893-5. 14. Saudan PJ, Brown MA, Farrrell T, Shaw L. Improved methods of assessing proteinuria in hypertensive pregnancy. Br J Obstet Gynaecol 1997;104:1159-64. 15. Young RA, Buchanan RJ, Kinch RA. Use of the protein/creatinine ratio of a single voided urine specimen in the evaluation of suspected pregnancy-induced hypertension. J Fam Pract 1996;42:385-9. 16. Quadri KH, Bernardini J, Greenberg A, Laifer S, Syed A, Holley JL. Assessment of renal function during pregnancy using a random urine protein to creatinine ratio and Cockcroft-Gault formula. Am J Kidney Dis 1994;24:416-20. 17. Jaschevatzky OE, Rosenberg RP, Shalit A, Zonder HB, Grunstein S. Protein/creatinine ra-
MAY 2007 American Journal of Obstetrics & Gynecology
465.e3
Research
Obstetrics
tio in random urine specimens for quantization of proteinuria in preeclampsia. Obstet Gynecol 1990;75:604-6. 18. Al RA, Baykal C, Karacay O, Geyik PO, Altun S, Dolen I. Random urine protein-creatinine ratio to predict proteinuria in new-onset mild hypertension in late pregnancy. Obstet Gynecol 2004;104:367-71. 19. Durnwald C, Mercer B. A prospective comparison of total protein/creatinine ratio versus 24-hour urine protein in women with suspected preeclampsia. Am J Obstet Gynecol 2003;189:848-52. 20. Combs CA, Wheeler BC, Kitzmiller JL. Urinary protein/creatinine ratio before and during
465.e4
www.AJOG.org pregnancy in women with diabetes mellitus. Am J Obstet Gynecol 1991;165:920-3. 21. Haas DM, Sabi F, McNamara M, RiveraAlsina M. Comparing ambulatory spot urine protein/creatinine ratios and 24-h urine protein measurements in normal pregnancies. J Matern Fetal Neonatal Med 2003;14:233-6. 22. Price C, Newall R, Boyd J. Use of protein: creatinine ratio measurements on random urine samples for prediction of significant proteinuria: a systematic review. Clin Chem 2005;9: 1577-86. 23. ACOG Committee on Obstetric Practice. ACOG practice bulletin. Diagnosis and man-
American Journal of Obstetrics & Gynecology MAY 2007
agement of preeclampsia and eclampsia. Number 33, January 2002. Int J Gynecol Obstet 2002;77:67-75. 24. Ortho-Clinical Diagnostics Inc. TP slide: microslide instructions for use manual and Vitros chemistry products. Version 5. Ortho-Clinical Diagnostics Inc; 2004:1-10. 25. Ortho-Clinical Diagnostics Inc. CREA slide: microslide instructions for use manual and Vitros chemistry products. Version 5. OrthoClinical Diagnostics Inc.; 2004:1-12. 26. Chan P, Brown M, Simpsom J, Davis G. Proteinuria in pre-eclampsia: how much matters? Br J Obstet Gynaecol 2005;112:280-5.
www. AJOG.org
OBSTETRICS
Usage of spot urine protein to creatinine ratios in the evaluation of preeclampsia Thomas L. Wheeler II, MD; Dawn W. Blackhurst, DrPH; Eric H. Dellinger, MD; Patrick S. Ramsey, MD, MSPH OBJECTIVE: The objective of the study was to prospectively compare spot urine protein to creatinine (P:C) ratios with 24 hour urine collections for protein in women being evaluated for preeclampsia. STUDY DESIGN: A spot urine P:C ratio was obtained at the beginning of 24 hour urine collections from 126 patients admitted to evaluate for preeclampsia. Correlation between the spot P:C ratio with the 24 hour urine collections was calculated. Receiver operator characteristic curves were constructed to determine best P:C cutoffs for 300 mg and 5000 mg protein per 24 hours.
RESULTS: Random spot P:C ratios were strongly correlated with 24 hour urine protein levels (Pearson r ⫽ 0.88). The optimal P:C cut-offs were 0.21 (300 mg per 24 hours) and 3.0 (5000 mg per 24 hours). A P:C ratio of less than 0.21 (300 mg per 24 hours) had a negative predictive value (NPV) of 83.3% and a P:C ratio of less than 3.0 (5000 mg per 24 hours) had 100% NPV. CONCLUSION: Urine spot P:C ratio correlated well with 24 hour urine collections for protein but was not justified as a substitute for timed collections.
Key words: creatinine, measurement, preeclampsia, protein
Cite this article as: Wheeler TL, Blackhurst DW, Dellinger EH, et al. Usage of spot urine protein to creatinine ratios in the evaluation of preeclampsia. Am J Obstet Gynecol 2007;196;465.e1-465.e4.
A
ssessment of urinary protein is an important component of the work-up for preeclampsia. In-patient 24 hour urine collections for protein, the gold standard, are cumbersome and time consuming, typically requiring a 2 night stay to get results. One simpler alternative, a random spot urine protein, is hypothetically subject to inaccuracies sec-
From the Department of Obstetrics and Gynecology, University of Alabama at Birmingham (Drs Wheeler and Ramsey), Birmingham, AL; and Department of Obstetrics and Gynecology, Greenville Hospital Systems University Medical Center (Dr. Blackhurst and Dellinger), Greenville, SC. Received July 5, 2006; revised September 6, 2006; accepted October 24, 2006. Reprint requests: Thomas L. Wheeler II, MD, Fellow, Female Pelvic Medicine and Reconstructive Pelvic Surgery, University of Alabama at Birmingham, Division of Women’s Pelvic Medicine and Reconstructive Surgery, 619 19 Street South, NHB 219, Birmingham, AL 35249-7333; [email protected]. 0002-9378/$32.00 © 2007 Mosby, Inc. All rights reserved. doi: 10.1016/j.ajog.2006.10.892
See Journal Club, page 491
ondary to time being the denominator for this value.1 Likewise, a spot urinary dipstick measurement of protein has not been shown to be reliable.2,3 However, random spot sampling for a urine protein to creatinine (P:C) ratio has been investigated as an alternative. With results in hours and easier collection, a spot P:C ratio is a more efficient test than a 24 hour collection for proteinuria assessment. Because the excretion of protein and creatinine are expressed per time, the ratio cancels out the time factor and allows for an estimation of 24 hour protein excretion if stable renal function is assumed.1 This ratio also normalizes urinary protein concentration to that of creatinine. In fact, in the literature on the medical general nonpregnant population, the spot P:C ratio performs well at assessing proteinuria, correlating with 24-hour protein excretion in the following populations: systemic lupus erythematosus, glomerular disease, renal transplant, and ages older than 5 years.4-8 The spot P:C ratio has been evaluated for use in pregnancy with mixed results.9-21 Some investigators have supported its use by demonstrating high correlations between spot P:Cs and 24 hour collections,9-17 whereas others have
stressed caution in its application, especially because renal function can destabilize within hours in preeclampsia.18-21 Recently a systematic review of spot P:C ratio use in the work-up of preeclampsia demonstrated strong evidence, based on the cumulative negative likelihood ratio of 0.14, that the spot P:C ratio identified patients unlikely to have significant proteinuria on a 24-hour collection. However, the variability in cut off values between studies did not allow for a recommendation of a critical threshold P:C ratio for 300 mg per 24 hours. The objectives of this study were to determine the strength of correlation between these 2 approaches to urinary protein assessment and to determine the optimal P:C ratios that best correspond to critical values on a 24 hour urine.
M ATERIALS AND M ETHODS Institutional review board approval was obtained from Greenville Hospital Systems University Medical Center (Greenville, SC). Spot P:C samples were gathered from 154 patients meeting inpatient admission criteria for the evaluation of preeclampsia, which was in general new-onset persistent hypertension, worsening hypertension, or proteinuria between December 2000 and July 2002
MAY 2007 American Journal of Obstetrics & Gynecology
465.e1
Research
Obstetrics
www.AJOG.org C OMMENT
TABLE
Test performance of spot P:C ratio assessment for 24 hour urine protein results 24 hour urine collection quantitative protein
Optimal spot P:C (per ROC analysis)
300 mg
0.21
86.8
77.6
81.9
83.3
1000 mg
0.46
87.5
82.4
53.8
96.6
2000 mg
0.82
100
94.8
62.5
5000 mg
3.0
100
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
100
..............................................................................................................................................................................................................................................
at the Greenville Hospital Systems University Medical Center. Following American College of Obstetricians and Gynecologists guidelines, new-onset hypertension was a systolic blood pressure greater than 140 mm Hg or a diastolic blood pressure greater than 90 mm Hg after 20 weeks’ gestation in a previously normotensive patient, whereas worsening hypertension was an increase in blood pressure from baseline taken before 20 weeks’ gestation.23 Samples were obtained at the beginning of the 24 hour urine collection. No firstmorning voids were used. Women who had bacteriuria on microscopy or who were on more than 24 hours’ bed rest, because of a potential poor correlation between spot P:C and 24 hour urine collections for protein after prolonged recumbency, were excluded.1,15 Urinary protein was determined by the Biuret method. 24 Urinary creatinine was determined by the 2-point rate method,25 and aliquots were analyzed by a Johnson & Johnson Vitros 250 (Johnson & Johnson Clinical Diagnostics Inc, Rochester, NY). The strength of correlation (Pearson correlation coefficient, r) between the spot P:C ratio and 24 hour urine collection protein amount was calculated by generating a least squares regression line. Using protein values of 300, 1000, 2000, and 5000 mg on 24-hour urine collections, receiver operator characteristic (ROC) curves were constructed, and the best P:C cut-off points were calculated. The best P:C ratio cut-off was identified as the point on the left shoulder of the ROC curve that maximizes sensitivity and specificity. As a measure of accuracy, the area under the ROC curves were calculated with 1 representing a perfect test 465.e2
100
100
100
and 0.5 a worthless test (ie, as accurate as flipping a fair coin). Therefore, as the area under the curve approaches 1, the accuracy improves. Sensitivity, specificity, and negative (NPV) and positive predictive values (PPV) were calculated for each cut-off value. ROC curves were constructed, and data were analyzed using MedCalc software, version 7 (MedCalc Software, Mariakerke, Belgium).
R ESULTS Urine collections were completed on 81.8% (126 of 154) of women. The other 18.2% (28 of 154) of women were delivered before the completion of their 24hour urine collection, which was then not processed. For the 126 patients completing the assessment, the mean age was 26.6 ⫾ 5.8 years, and the mean estimated gestational age was 34.0 ⫾ 3.3 weeks, with 27% blacks, 72% whites, and 1% Hispanics. Parity was 56% nulliparous, 30% primiparous, 12% para 2, and 2% para 3. The test performance for the spot P:C ratio for the varying degrees of proteinuria noted on 24 hour urine collection are shown in the Table. The spot P:C ratio correlated with 24 hour urine protein collection results (r ⫽ 0.88). The optimal P:C ratio cut-offs, based on the ROC analysis, were 0.21 for 300 mg per 24 hours, 0.46 for 1000 mg per 24 hours, 0.82 for 2000 mg per 24 hours, and 3.0 for 5000 mg per 24 hours. The areas under the receiver operator curves for 300, 1000, 2000, and 5000 mg of protein on a 24 hour urine collection were 0.86, 0.91, 0.98, and 1.0, respectively (Figure).
American Journal of Obstetrics & Gynecology MAY 2007
Most studies have repeated good correlations between spot P:C ratios and 24 hour urine collections for protein with correlation coefficients (r) ranging between 0.80 and 0.97.11-17,20 Furthermore, spot urine P:C ratios correlated with 24 hour collections, even when the sample population included hypertensive and renal disorders.13 A spot P:C ratio has also been proposed to not be confounded by maternal age, gestational age, or parity (10,16,18,20) and has been correlated with adverse maternal and fetal outcomes.26 However, 2 recent studies determined lower correlation coefficients of r ⫽ 0.56 and r ⫽ 0.41 and cautioned against substituting spot P:C ratio for timed collections.18,19 Additionally, its use may be limited in women with pregestational diabetes and in ambulatory patients in the second and third trimesters.20,21 Along with the concerns about the strength of correlation, finding a spot P:C ratio value with acceptable PPVs and NPVs has been a challenge. ROC curves have been used to select the P:C ratio point that maximizes sensitivity and specificity. For example a P:C cut-off such as 0.19 has been described, but false negatives still occur with this method.11,19 To address this concern and limit false negatives, a P:C point with maximal sensitivity has been chosen, like 0.14,11 at the expense of specificity, which then results in increased misdiagnoses of preeclampsia with obvious ramifications to management. In fact, the recent systematic review analyzing spot P:C ratios in preeclampsia supports the use of a spot P:C ratio in ruling out significant proteinuria.23 However, no consensus for specific P:C cut-off values has been obtained. These uncertainties have added to the resistance of substituting it for timed collections when evaluating preeclampsia, in which renal function can decline within hours, unlike the usual time course of renal decline in the general nonpregnant medical population. Our study found a strong correlation (r ⫽ 0.88) between the spot P:C ratio and 24 hour urine protein results. Based on
Obstetrics
www.AJOG.org
FIGURE
Receiver operator characteristic curves for 24 hour urine protein results are shown
A, For 300 mg, the P:C ratio cut-off is greater than 0.21. B, For 1000 mg, the P:C ratio cut-off is greater than 0.46. C, For 2000 mg, the P:C ratio cut-off is greater than 0.82. D, For 5000 mg, the P:C ratio cut-off is 3.0 or greater.
our ROC analysis, the optimal P:C ratio of 0.21 was identified for 300 mg per 24 hours and 3.0 for 5000 mg per 24 hours. The areas under the ROC curves appeared to show that these cut-off spot P:C ratios were accurate in predicting the 24 hour urine collection results. However, 13.2% (9 of 68) of women with significant proteinuria (more than 300 mg per 24 hours) were missed by this P:C ratio cut off (ie, false-negative rate). We concluded the following: (1) spot P:C ratio is strongly correlated (r ⫽ 0.88) with 24 hour urine collections for protein, (2) the NPV for P:C ratios 0.21 of 83.3% is too low to be a substitute for 24 hour collections, and (3) despite a 100% NPV and PPV, the sample size for P:C ratio greater than 3.0 (5000 mg per 24 hours) of n ⫽ 5 was too small to justify eliminating the 24 hour collection for this value. Weaknesses of our study include the small sampling of women with greater than 5000 mg protein per 24 hours. This study also included patients with renal disease, chronic hypertension, and dia-
betes, in whom preexisting proteinuria could exist, but did not subgroup them for analysis. Hence, the impact of these conditions and preexisting proteinuria on the strength of correlation between the spot P:C and 24 hour urine collections for protein is unknown in this study subpopulation. We suggest that a clinical study with a larger sample size to determine the best P:C cut point for 5000 mg is needed. It is also important to clarify the impact of comorbid disease and/or preexisting proteinuria on the correlation between spot P:C and 24 hour urine collection for protein.22 f REFERENCES 1. Ginsberg JM, Chang BS, Matarese RA, Garella S. Use of single voided urine sample to estimate quantitative proteinuria. N Engl J Med 1982; 309:1543-6. 2. Phelan LK, Brown MA, Davis GK, Mangos G. A prospective study of the impact of automated dipstick urinalysis on the diagnosis of preeclampsia. Hypertens Pregnancy 2004;23: 135-42.
Research
3. Meyer NL, Mercer BM, Friedman SA, Sibai BM. Urinary dipstick protein: a poor predictor of absent or severe proteinuria. Am J Obstet Gynecol 1994;170:137-41. 4. Christopher-Stine L, Petri M, Astor, BC, Fine D. Urine protein-to-creatinine ratio is a reliable measure of proteinuria in lupus nephritis. J Rheumatol 2004;31:1557-9. 5. Chitalia VC, Kothari J, Wells EJ, et al. Costbenefit analysis and prediction of 24-hour proteinuria from the spot urine protein-creatinine ratio. Clin Nephrol 2001;55:436-47. 6. Xin G, Wang M, Jiao L, Xu G, Wang H. Protein-to-creatinine ratio in spot urine samples as a predictor of quantification of proteinuria. Clinica Chimica Acta 2004;350:35-9. 7. Torng S, Rigatto C, Rush DN, Nickerson P, Jeffery JR. The urine protein to creatinine ration (P/C) as a predictor of 24-hour urine protein excretion in renal transplant patients. Transplantation 2001;72:1453-6. 8. Tamura S, Shimizu T, Kawakatsu H, Tateishi S. Correlation between 24-hour urinary protein excretion and protein/creatinine ratio in the first voided morning urine samples. Jpn J Nephrol 2004;46:26-34. 9. Yamasmit W, Wongkitisophon K, Charoenvidhya D, Uerpairojkit B, Chaithongwongwatthana S. Correlation between random urinary protein-to creatinine ratio and quantitation of 24-hour proteinuria in preeclampsia. J Med Assoc Thai 2003;86:69-73. 10. Neithardt AB, Dooley SL, Borensztajn J. Predicition of 24-hour protein excretion in pregnancy with a single voided urine protein-to-creatinine ratio. Am J Obstet Gynecol 2002;186: 883-6. 11. Rodriguez-Thompson D, Lieberman E. Use of random urinary protein-to-creatinine ratio for the diagnosis of significant proteinuria during pregnancy. Am J Obstet Gynecol 2001; 185:808-11. 12. Ramos JG, Martins-Costa SH, Mathias MM, Guerin YL, Barros EG. Urinary protein/creatinine ratio in hypertensive pregnant women. Hypertens Pregnancy 1999;18:209-18. 13. Robert M, Sepandj F, Liston R, Dooley K. Random protein-creatinine ratio for the quantization of proteinuria in pregnancy. Obstet Gynecol 1997;90:893-5. 14. Saudan PJ, Brown MA, Farrrell T, Shaw L. Improved methods of assessing proteinuria in hypertensive pregnancy. Br J Obstet Gynaecol 1997;104:1159-64. 15. Young RA, Buchanan RJ, Kinch RA. Use of the protein/creatinine ratio of a single voided urine specimen in the evaluation of suspected pregnancy-induced hypertension. J Fam Pract 1996;42:385-9. 16. Quadri KH, Bernardini J, Greenberg A, Laifer S, Syed A, Holley JL. Assessment of renal function during pregnancy using a random urine protein to creatinine ratio and Cockcroft-Gault formula. Am J Kidney Dis 1994;24:416-20. 17. Jaschevatzky OE, Rosenberg RP, Shalit A, Zonder HB, Grunstein S. Protein/creatinine ra-
MAY 2007 American Journal of Obstetrics & Gynecology
465.e3
Research
Obstetrics
tio in random urine specimens for quantization of proteinuria in preeclampsia. Obstet Gynecol 1990;75:604-6. 18. Al RA, Baykal C, Karacay O, Geyik PO, Altun S, Dolen I. Random urine protein-creatinine ratio to predict proteinuria in new-onset mild hypertension in late pregnancy. Obstet Gynecol 2004;104:367-71. 19. Durnwald C, Mercer B. A prospective comparison of total protein/creatinine ratio versus 24-hour urine protein in women with suspected preeclampsia. Am J Obstet Gynecol 2003;189:848-52. 20. Combs CA, Wheeler BC, Kitzmiller JL. Urinary protein/creatinine ratio before and during
465.e4
www.AJOG.org pregnancy in women with diabetes mellitus. Am J Obstet Gynecol 1991;165:920-3. 21. Haas DM, Sabi F, McNamara M, RiveraAlsina M. Comparing ambulatory spot urine protein/creatinine ratios and 24-h urine protein measurements in normal pregnancies. J Matern Fetal Neonatal Med 2003;14:233-6. 22. Price C, Newall R, Boyd J. Use of protein: creatinine ratio measurements on random urine samples for prediction of significant proteinuria: a systematic review. Clin Chem 2005;9: 1577-86. 23. ACOG Committee on Obstetric Practice. ACOG practice bulletin. Diagnosis and man-
American Journal of Obstetrics & Gynecology MAY 2007
agement of preeclampsia and eclampsia. Number 33, January 2002. Int J Gynecol Obstet 2002;77:67-75. 24. Ortho-Clinical Diagnostics Inc. TP slide: microslide instructions for use manual and Vitros chemistry products. Version 5. Ortho-Clinical Diagnostics Inc; 2004:1-10. 25. Ortho-Clinical Diagnostics Inc. CREA slide: microslide instructions for use manual and Vitros chemistry products. Version 5. OrthoClinical Diagnostics Inc.; 2004:1-12. 26. Chan P, Brown M, Simpsom J, Davis G. Proteinuria in pre-eclampsia: how much matters? Br J Obstet Gynaecol 2005;112:280-5.