Use of the protein/creatinine ratio of a single voided urine specimen in the evaluation of suspected pregnancy-induced hypertension.
Methods. Single voided urine specimens and 24-hour total urine protein collections were ordered for 66 consecutive women admitted to an antepartum unit for suspected PIH. The correlation of the protein/creatinine ratio of the single voided specimen with the 24-hour urine protein excretion was calculated.
Results. Forty-one sets of data with a 24-hour urine protein less than 1000 mg per 24 hours were obtained. The correlation of the single voided protein/creatinine ratio to the 24-hour total protein was 0.80 (P[less than].001), with a regression equation of (protein/creatinine)=0.81X(24-hour protein)-0.011. No single protein/creatinine ratio cutoff was ideal to distinguish between significant and insignificant proteinuria; however, a ratio less than .15 efficiently ruled out significant PIH.
Conclusions. We conclude that the protein/creatinine ratio of a single voided urine specimen may have a role in the management of ambulatory women with suspected PIH, although further study is needed. The main potential benefit of this method is that in institutions where women with suspected PIH are hospitalized, women with insignificant proteinuria may be identified within a matter of hours and their follow-up care handled on an outpatient basis.
Key words. Preeclampsia; proteinuria; pregnancy; hypertension; eclampsia; pregnancy-induced hypertension. (J Fam Pract 1996; 42:385-389)
The evaluation of a woman suspected of having pregnancy-induced hypertension (PIH) must include an assessment of proteinuria. The 24-hour urine collection for total protein, which is the current gold standard for this assessment, is a cumbersome process and the urine is often improperly collected.
The American College of Obstetrics and Gynecology (ACOG) defines the amount of proteinuria consistent with preeclampsia as 300 mg per 24 hours.(1) The protein/creatinine ratio of a single voided urine specimen has been shown to correlate well with the total 24-hour urine protein loss in adults with renal disease(2),(3) and in pregnant women with and without preeclampsia.(4),(5) In these studies, however, the amount of proteinuria was as much as 10 g per 24 hours, with many of the data points greater than 1 g per 24 hours. It is difficult to tell from these studies if the correlation is as good when the range of values is close to 300 mg protein per 24 hours.
In this study, we investigated the correlation between the protein/creatinine ratio of a single voided specimen and the 24-hour urine total protein in the range of zero to 1000 mg protein per 24 hours.
Permission to perform this research was obtained from the institutional review board of John Peter Smith Hospital. Patients were enrolled in this study after informed consent was obtained. In addition, the hospital agreed to fund the cost of the protein/creatinine ratio test on the single voided samples.
We enrolled 66 consecutive ambulatory women suspected of having PIH who were admitted to the antepartum service of John Peter Smith Hospital between June 1992 and June 1993 and between December 1993 and August 1994. Enrollment criteria were blood pressure greater than 140/90 mm Hg, systolic pressure greater than 30 mm Hg above baseline, or diastolic 15 mm Hg above baseline. There was only one exclusionary criterion. Women were excluded if they had been previously diagnosed as having PIH and had been placed on long-term bed rest at home or strict bed rest in the hospital for more than 36 hours. This criterion was based on the decreased correlation between the protein/creatinine ratio and the 24-hour protein resulting from prolonged recumbency.
Each participant had a 24-hour urine collection for total protein ordered as part of the study institution's standard evaluation of suspected PIH. The study design was that the single voided specimen collected when the participant emptied her bladder began the timing of the 24-hour urine collection, which occurred in 66% of the participants. For cases in which the initial bladder voiding was not collected, a single voided sample was collected immediately after the completion of the 24-hour urine collection. It was felt that this sample would still be acceptable because all previous studies that examined the correlation between the protein/creatinine ratio and the 24-hour urine protein excretion collected the single voided sample at this time. In our study, 34% of the single voided samples were collected at this time. The urine protein and creatinine concentrations were measured by a standard spectrophotometric technique using Beckman analyzers.
No specimen was collected as the first void of the morning, because earlier work has shown a diminished correlation using this specimen, presumably because of prolonged recumbency.(2)
No patient's management was changed by enrollment in this study. Twenty-one patients had incomplete data pairs for various reasons, including incomplete physician orders, computer entry errors, and inductions of labor. Invariably, if a woman was transferred from the antepartum floor to the labor and delivery area for induction of labor, the 24-hour urine collection was lost or abandoned, which occurred in six of the 21 patients.
[Part 1 of 2]
Table 1. Patients with Complete and Incomplete Data Pairs
Patient Complete Set Incomplete Set Characteristics (n=45) (n=21)
Gravida 1.78 1.62 Parity 0.51 0.43 EGA (wks) 33.4 32.8 PIH diagnosis, % 57.8 61.9 Chronic medical 24.4 33.3 condition, %
[Part 2 of 2]
Table 1. Patients with Complete and Incomplete Data Pairs
Patient Characteristics Statistics(1)
Gravida t=0.57, df=64 Parity t=0.36, df=64 EGA (wks) t=0.49, df=64 PIH diagnosis, % [[chi].sup.2]=0.10, df=1 Chronic medical [[chi].sup.2]=0.77, df=1 condition, %
EGA denotes estimated gestational age; PIH, pregnancy-induced hypertension.
The value of the protein/creatinine ratio (mg/g) of the single voided specimen was plotted against the total amount of protein excreted in the 24-hour urine collection (g) for all 24-hour protein excretions less than 1000 mg. The Pearson product moment correlation was used to analyze the relationship between the protein/creatinine ratio in the single voided specimen and the 24-hour protein excretion. The regression equation of the line that best fit the data was calculated using the method of least squares.
A receiver operating curve (ROC) was generated to examine the performance of various single cutoff points for defining a 24-hour protein excretion of greater than 300 mg. A series of four protein/creatinine ratio cutoff pairs was then created to examine how well values falling below a lower limit cutoff point or greater than an upper limit cutoff point predicted proteinuria less than or greater than 300 mg per 24 hours. Two-by-three tables were constructed for each of the four cutoff pairs, and the results analyzed using the method of chi-square.
The group of women for whom we did not have complete data was compared with the group of women with complete data (Table 1). The two groups did not differ significantly by gravida, parity, estimated gestational age, final diagnosis of PIH, or significant coexisting medical illness, eg, chronic hypertension, diabetes, and asthma.
The plot of the protein/creatinine ratio of the single voided sample vs the 24-hour total protein excretion for the 41 women whose 24-hour protein was less than 1000 mg is shown in Figure 1. The correlation coefficient (r) is .80 (P[less than].001) and the regression equation of the line of best fit is (protein/creatinine ratio)=0.81X(24-hour protein)-0.011. The 95% confidence interval of the regression equation is shown in Figure 1.
The correlation of the protein/creatinine ratio and the 24-hour urine protein, including data from the four women with 24-hour urine protein values greater than 1000 mg, is 0.89. This value is slightly lower than the range found in other studies (0.92 to 0.99).(2),(3),(4),(5)
The correlation was also calculated separately for patients whose single voided specimen was collected before and after the 24-hour collection. The correlation coefficient for the specimens collected before the 24-hour collection is 0.78 (n=27). The correlation coefficient for those collected after the 24-hour collection is very similar at 0.83 (n=14).
The ROC for various protein/creatinine ratio cutoff points, using all 45 complete data sets, is shown in Figure 2 (area under the curve = 0.82). The sensitivity and specificity of each of the cutoff points is shown in Table 2. No single cutoff point appears to be satisfactory in determining whether a patient has significant proteinuria.
The performance of four protein/creatinine ratio cutoff pairs is shown in Table 3. The purpose of this analysis was to determine if a lower limit cutoff point could be used to rule out significant proteinuria and an upper limit cutoff point could be used to rule in significant proteinuria, with an indeterminate range between the two points. For example, if the cutoff points of .100 and .250 are examined, 17 patients are correctly classified, one is incorrectly classified as a false negative, but 27 patients have indeterminate protein/creatinine ratios that fall between .100 and .250. Using the cutoff points of .150 and .250, 24 patients are correctly classified and 15 patients fall between .150 and .250 at a cost of four false positives and two false negatives.
Seven additional patients had specimens collected for protein/creatinine ratios after they had been placed on strict bed rest in the hospital for more than 24 hours. The protein/creatinine ratio in this group of patients did not correctly predict the 24-hour total protein value.
Table 2. Sensitivity and Specificity of Various Protein/Creatinine Ratio Cutoff Points
Cutoff Point Sensitivity, % Specificity, %
0.05 100 0 0.10 96 18 0.15 91 41 0.20 78 59 0.25 65 82 0.30 57 100 0.35 48 100
In pregnant women with suspected PIH, it is imperative to quantify the amount of proteinuria to assess the severity of disease, if present. The 24-hour urine collection is often the slowest part of the workup of a pregnant woman with hypertension, which has ramifications for the health of both the mother and the fetus. It would be desirable to develop a method to assess proteinuria more rapidly and with fewer collection problems than the traditional 24-hour urine collection, especially in institutions where patients with suspected PIH are initially hospitalized. Even in our inpatient setting, 12% of women at risk did not have a successfully collected 24-hour urine sample.
[Part 1 of 2]
Table 3. Performance of the Protein/Creatinine Ratio as a Function of Various Cutoff Pairs
Correctly Classified Protein/Creatinine Ratio Cutoffs True + True - Total
0.100 & 0.250(1) 15 4 19 0.100 & 0.300(2) 13 4 17 0.150 & 0.250(3) 15 9 24 0.150 & 0.300(4) 13 9 22
[Part 2 of 2]
Table 3. Performance of the Protein/Creatinine Ratio as a Function of Various Cutoff Pairs
Incorrectly Classified Protein/Creatinine Ratio Cutoffs False + False - Total Indeterminate
0.100 & 0.250(1) 4 1 4 22 0.100 & 0.300(2) 0 1 1 27 0.150 & 0.250(3) 4 2 6 15 0.150 & 0.300(4) 0 2 2 21
(1) df = 2, [[chi].sup.2] = 10.6, P[less than] .01.
(2) df = 2, [[chi].sup.2] = 18.0, P[less than] .001.
(3) df = 2, [[chi].sup.2] = 11.5, P[less than] .01.
(4) df = 2, [[chi].sup.2] = 18.7, P[less than] .001.
NOTE: The lower ratio rules out pregnancy-induced hypertension, the upper pair rules it in, and those between are nondiagnostic.
Many guidelines on the workup of PIH include evaluation of proteinuria made on the basis of a urine dipstick.(1),(6),(7) Two recent studies show that the dipstick testing result is a poor predictor of the 24-hour protein excretion.(8),(9) The authors of both of these studies recommend against using the urine dipstick result, stating that the 24-hour urine collection for total protein should be the basis for classifying the degree of proteinuria in these women. The protein/creatinine ratio of a single voided specimen may, however, be an acceptable alternative to the 24-hour urine collection, especially if the limitations of this method are understood.
For the protein/creatinine ratio to be valid, the patient cannot have been confined to strict bed rest prior to the collection of the single voided urine specimen. Ginsberg et al(2) observed a worsening of the correlation between the protein/creatinine ratio and the 24-hour collection for samples collected at night and at the first void of the morning. They speculated that postural changes in the renin-angiotensin system might explain this change. The correlation coefficient of our group of patients whose single voided specimen was collected before the 24-hour collection period (r= .78) was similar to the group of patients who had the single voided specimen collected after the 24-hour collection (r= .83). Therefore, it seems that the protein/creatinine ratio remains valid as long as urine is collected within 24 hours of a patient having been prescribed strict bed rest, and the single voided specimen is not the first void of the morning. Beyond this 24-hour period, the ratio becomes unreliable. It is not known if this same restriction is necessary for women with a diagnosis of PIH who have been prescribed modified bed rest at home and are being followed as outpatients.
What should the lower limit cutoff point of the protein/creatinine ratio be? Two women had a false-negative ratio of less than 0.150, but a 24-hour total protein excretion of greater than 300 mg. One woman had 4400 mL of urine collected with a total protein of 308 mg per 24 hours. Her total creatinine excreted per 24 hours was 1.85 g, which is more than two standard deviations greater than that expected in a young woman.(10) Her 24-hour urine collection took place while she was in the labor and delivery suite and the antepartum floor. It is likely that her urine was collected for more than 24 hours. She gave birth at term 5 weeks after this admission with no further hypertension or proteinuria. The other patient had a 24-hour total protein excretion of 330 mg. Her condition was diagnosed as mild preeclampsia, which was responsive to bed rest, and she was later discharged home. She reported decreased fetal movements 2 weeks later and had labor induced after oligohydramnios was diagnosed by ultrasound. Therefore, even in the two patients who had a false-negative result using a lower-limit cutoff of 0.150, there was no apparent adverse effect on the ultimate outcome of the mothers or their babies.
Our study has several weaknesses, including a relatively small sample size, inconsistent collection times for single voided specimens, and a significant number of eligible study participants (21/66) with incomplete data. This work should be duplicated in other centers. Easily measured values that may improve correlation between the protein/creatinine ratio and the 24-hour protein, such as urine specific gravity and single voided specimen volume, should be measured as well.
The choice of whether a clinician should use the protein/creatinine ratio is a difficult one. The potential benefit of a more efficient management strategy for women with suspected PIH, through avoided or shortened hospitalizations, and a more reliably collected measure of the degree of proteinuria in ambulatory patients must be balanced against the risk to patients with a false-negative protein/creatinine ratio. With this caution in mind, we believe that there are two clinical scenarios for which the protein/creatinine ratio method holds the greatest promise.
Our institution encounters women who come to a prenatal clinic for the first time in their third trimester, including some who have PIH diagnosed at this first encounter. The use of the protein/creatinine ratio to identify patients with severe preeclampsia, as defined by the amount of proteinuria, may allow induction of labor to proceed sooner, if indicated by the assessment of other maternal and fetal risk factors. This will reduce the time of exposure these patients have to the risk of developing eclampsia.
The greater potential for the use of the protein/creatinine ratio is for ambulatory women who have little or insignificant proteinuria. An ambulatory woman who is seen for a prenatal care visit and has blood pressure greater than 140/90 mm Hg does not meet the criteria for PIH at that point. The ACOG guidelines state that the "...[elevated] blood pressures must be manifested on at least two occasions six or more hours apart."(1) At our institution, a woman with one elevated blood pressure reading is admitted to the antepartum unit for a "tox watch" and usually spends at least 3 days in the hospital, even in uncomplicated cases. The rate-limiting step for discharge of a woman with an uncomplicated course is always the 24-hour urine collection.
It may be possible for a patient with this presentation to be observed for only 6 hours. During this time, the protein/creatinine ratio of a single voided urine specimen could be determined and other necessary laboratory tests performed. At the time of the second blood pressure reading 6 hours later, the clinician is able to rule in or rule out a diagnosis of PIH, and if the diagnosis is made, to determine how severe the disease is. This should allow the clinician to discharge home patients who may be safely followed on an outpatient basis and further observe patients at higher risk in the hospital.
The authors wish to thank the nurses and clerks of the antepartum floor of John Peter Smith Hospital for their efforts in identifying and recruiting patients for this study. We also thank the administration of John Peter Smith Hospital for funding the cost of the protein/creatinine ratio urine test for the enrolled patients.
(1) . American College of Obstetricians and Gynecologists. Management of preeclampsia. Washington, DC: American College of Obstetricians and Gynecologists, Feb 1986. ACOG Technical Bulletin No. 91, February 1986.
(2) . Ginsberg JM, Chang BS, Matarese RA, Garella S. Use of single voided urine samples to estimate quantitative proteinuria. N Engl J Med 1983; 309:1543-6.
(3) . Schwab SJ, Christensen L, Dougherty K, Klahr S. Quantitation of proteinuria by the use of protein-to-creatinine ratios in single urine samples. Arch Intern Med 1987; 147:943-4.
(4) . Boler L, Zbella EA, Gleicher N. Quantitation of proteinuria in pregnancy by the use of single voided urine samples. Obstet Gynecol 1987;70:99-100.
(5) . Jaschevatzky OE, Rosenberg RP, Shalit A, Zonder HB, Grunstein S. Protein/creatinine ratio in random specimens for quantitation of proteinuria in preeclampsia. Obstet Gynecol 1990; 75:604-6.
(6) . Cunningham FG, Lindheimer MD. Hypertension in pregnancy. N Engl J Med 1992; 326:927-32.
(7) . Dildy GA, Cotton DB. Management of severe preeclampsia and eclampsia. Crit Care Clin 1991; 7:829-50.
(8) . Kuo VS, Koumantakis G, Gallery EDM. Proteinuria and its assessment in normal and hypertensive pregnancy. Am J Obstet Gynecol 1992; 167:723-8.
(9) . 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.
(10) . Kampmann J, Siersbaek-Nielsen K, Kristensen M, Molholm Hansen J. Rapid evaluation of creatinine clearance. Acta Med Scand 1974; 196:517-20.
|Printer friendly Cite/link Email Feedback|
|Author:||Young, Richard A.; Buchanan, Rebecca J.; Kinch, Robert A.H.|
|Publication:||Journal of Family Practice|
|Date:||Apr 1, 1996|
|Previous Article:||The completion of advance directives in primary care.|
|Next Article:||The Papanicolaou smear: its value and limitations.|