Comparison of Spot Urine Protein/Creatinine Ratio, Spot Urine Protein/Osmolality Ratio with Measured 24-Hour Urine Protein in HIV Subjects in Nigeria

Ernest Ndukaife Anyabolu1,2, Innocent Ijezie Chukwuonye2*, Monica Mabayoje3, ArodiweEjike4, Chinwuba K Ijoma4, Solomon Kadiri5 and EfosaOviasu6 1Division of Nephrology, Department of Internal Medicine, Imo State University Teaching Hospital, Orlu, Nigeria. 2Division of Nephrology, Department of Internal Medicine, Federal Medical Centre Owerri, Nigeria 3Division of Nephrology, Department of Internal Medicine, Lagos University Teaching Hospital, Lagos, Nigeria 4Division of Nephrology, Department of Internal Medicine, University of Nigeria Teaching Hospital, Enugu, Nigeria 5University College Hospital, Ibadan, Nigeria 6Division of Nephrology, Department of Internal Medicine, University of Benin Teaching Hospital, Benin City, Nigeria


Introduction
Chronic kidney disease (CKD) is defined by the presence of kidney damage or glomerular filtration rate (GFR) less than 60 ml/min/1.73 m 2 (1.0 ml/s/1.73 m 2 ) for three or more months, irrespective of cause [1]. HIV infection is commonly associated with chronic kidney disease [2]. Accurate and fast diagnostic tools are necessary to diagnose and monitor HIV patients who may have CKD. In majority of subjects with CKD, the presence of CKD can be detected with 2 tests. These tests include a urine test for the detection of proteinuria and a blood test to estimate the GFR. These two tests are commonly used by Nephrologists for detection of CKD in Sub Saharan Africa and other parts of the world.
Proteinuria is the commonest marker of CKD in adults and contributes to its progression by several mechanisms; and appropriate interventions that reduce proteinuria also improve patient outcome [3]. Urine protein estimation is one of the key parameters in the diagnosis and monitoring of renal functions in HIV-related renal disease states [4]. Evaluation of protein from 24-hour urine collection collection is the traditional and gold standard diagnostic test for quantification of proteinuria in the general population including in HIV subjects. The 24 HUP sample collection is often difficult, tedious, time-consuming, and is also riddled with errors [5,6]. Due to the problems associated with 24 HUP collection, some alternatives for the quantification of proteinuria in HIV seropositive subjects have been considered. The alternatives include urinary dipsticks protein estimation, SUPCR, and SUPOR.
Dipstick urinalysis protein estimation is known to have a poor sensitivity and specificity profile for establishing significant proteinuria or quantitative change in proteinuria [7]. Measurement of albumin or total protein concentration in a spot sample by tests like SUPCR and SUPOR avoids the need for collection of a timed urine specimen but is affected by some factors including the state of hydration, sex, and age [4]. However, some studies have demonstrated that SUPCR in a random, single voided urine sample reflects 24-hour urine protein [8][9][10]. Random and morning urinary protein/osmolality ratio has also been found to reflect 24 HUP excretion [11].
There is paucity of data on HIV subjects comparing SUPGR against 24 HUP and SUPOR against 24 HUP, emanating from Sub Saharan Africa. In addition, with the increasing incidence of HIV infection in the region, it is pertinent that faster, cheaper, simpler, and evidencebased reliable methods of quantifying proteinuria in HIV patients be adopted. This will help in no small measure to improve the management of this group of patients in Sub Saharan African countries. Bearing this in mind we decided to carry out this study to validate the use of SUPCR, and SUPOR in quantifying proteinuria in HIV subjects, by comparing each of them respectively, with 24 HUP.

Materials and Methods
This was a cross-sectional comparative study of SUPCR and SUPOR against the "gold Standard, " 24 HUP, in newly-diagnosed HIVseropositive subjects. The study was carried out between March, 2011 and August, 2011 in Federal Medical Centre, Owerri (FMC), Imo State, Nigeria. The subjects consisted of 393 newly-diagnosed, drug-naïve, HIV-seropositive adult subjects within the age range of 18-65 years, and 136 age-and sex-matched HIV-sero-negative subjects as control. Exclusion criteria included subject below 18 years or above 65 years, febrile illnesses, evidence of heart failure, urinary tract infection, diabetes mellitus, hypertension, evidence of urological disease, and use of drugs that could interfere with urinary creatinine excretion. The same exclusion criteria applied to the control group. An interviewer structured questionnaire was administered and relevant data collected. These included patient's age, sex, diagnosis and co-morbidities (diabetic mellitus, hypertension etc.).
Approval for this study was obtained from the Ethical and Research Committee of FMC Owerri. Investigations performed by the subjects included: HIV screening and confirmatory tests, SUP, SUCr, SUOsm, serum creatinine, 24 HUP, FBS, and urinalysis.
Clear instructions were given to all the subjects on how to collect 24-hour urine sample. 24-hour urine protein was measured in the urine samples thus collected. A day-time random spot urine sample was collected at the end of 24-hour urine sample collection and was used in calculating SUPCR and SUPOR. This was to facilitate rapid analysis in our laboratory [12]. Urine protein was measured by photometric method, urine osmolality by freezing point depression method (using Precision System Osmette 5002 osmometer) and creatinine by modified Jaffe method. All the laboratory tests were carried out in the laboratory of FMC, Owerri, while urine osmolality tests were done by E,N Anyabolu. FMC, Owerri, is one of two tertiary health institutions in the State, and has a good number of laboratory scientists. SUPCR and SUPOR were determined.

Statistical analysis
SPSS version 17.0 (SPSS Inc., Chicago, Il, USA) statistical software was used to analyze the data. The 2 × 2 contingency tables were used to calculate the sensitivity, specificity, positive predictive value, negative predictive value and accuracy obtained. The cut-off value for the 2 by 2 table was 24 HUP 0.150 g, SUPCR 0.150 mg/mg, SUPOR 0.150 mg/dL/ mOsm/Kg H 2 O. Predictive indices of SUPCR and SUPOR for 24 HUP were determined. The strength of association between the SUPCR and against 24 HUP, and also SUPOR and against 24 HUP was determined by correlation statistics. P<0.05 was taken as statistically significant. Bland Altman's plot was used to compare SUPCR, SUPOR, each with 24 HUP (Medcalc statistical software) BLAND ALTMAN [13]. Cutoff values, sensitivity and specificity of SUPCR and SUPOR were assessed for predicting 24 HUP excretion "threshold" of 0.150 g/day by ROC curve. These four instruments, namely 2 × 2 contingency table analysis, correlation statistics, receiver operator characteristics (ROC) curve analysis and Bland Altman plots were used to compare SUPCR, SUPOR with 24 HUP in HIV subjects and Control. P<0.05 was taken as statistically significant
The female/male ratio was approximately 3:1 in both the HIV subjects and control subjects. However, 22 subjects were excluded from the study on account of inadequate 24-hour urine collection.

Discussion
This study compared the SUPCR and SUPOR against "gold standard" 24 HUP in HIV subjects. In the 2 by 2 table analysis in HIV subjects at cut off of 0.150 for the three methods, the sensitivity for SUPCR, and SUPOR was, 43%, and 11% respectively, while the specificity was 92%, and 99.2% respectively in predicting 24 HUP. These shows that both spot urine tests had low sensitivity and high specificity. This tends to suggest that they are both poor screening tests, good diagnostic tests. Similar results were also obtained in the control. We did not, however, evaluate these predictive indices of SUPCR and SUPOR for 24 HUP at higher cut-off values of proteinuria.
The correlation of SUPCR and SUPOR, with 24 HUP, respectively, in the HIV subjects, showed This study showed that SUPCR had a high correlation coefficient (0.734) with 24 HUP (r=0.734, p<0.001) , while SUPOR had a moderate correlation (0.417) with 24 HUP (r=0.417, p<0.001) in HIV subjects. The p values were significant in both cases. In the Control the SUPCR had a moderate correlation (0.518), while SUPOR had a mild (0.336) correlation. This correlation values, of SUPCR with 24 HUP, in HIV subjects, where slightly lower than those reported in some studies [14][15][16][17][18]. None the less one of these studies was conducted in an HIV population, while the other was in a general population. This study also showed lower correlation values for SUPCR and 24 HUP, compared with those of two previous studies [11,12].
Bland-Altman analysis is typically used to compare measurement techniques against a reference value, usually an accepted gold standard.     We want to know by how much the new method is likely to differ from the old; if this is not enough to cause problems in clinical interpretation we can replace the old method by the new or use the two interchangeably. " [13]. Bland Altman plot was used in this study to compare SUPCR and SUPOR against 24HUP, respectively. The bias and precisions were low, and though good correlation was observed between SUPCR, SUPOR with 24 HUP in this study, it is expedient that reasonable limits of agreement be demonstrated for them to be used interchangeably. The 95% limits of agreements values were within acceptable range.
We found, in this study, for both SUPCR and SUPOR, when each was compared with 24 HUP, that the results were similar within 95% limits of agreement. These findings, and correlation, strongly suggest that both tests may be used interchangeably with the 24 HUP, especially the SUPCR that had a high correlation with 24 HUP in HIV subjects. Similar results were reported by previous studies which also showed that SUPCR and 24 HUP could be used interchangeably in HIV subjects and in the general population [14,18]. This is in keeping with findings from some previous studies [14,18].
The predictive indices of SUPCR and SUPOR against measured 24 HUP in HIV were also evaluated using the ROC curves. Based on ROC comparison, our study showed low sensitivity and specificity for SUPCR (at values ≥0.150 mg/mg) and SUPOR (at values ≥0.150 mg/dl/mOsm/ Kg H 2 O) methods in predicting 24 HUP of ≥0.150 g in HIV subjects. There was no difference between SUPCR and SUPOR in predicting 24 HUP at this level. This is similar to the findings reported by two studies with slightly higher AUROC [19,20]. This is in keeping with some earlier studies [19,20]. However SUPCR was found to be better than SUPOR when compared with 24 HUP in another study [21]. In addition, with further evaluation in our study, we found that in SUPCR, of 0.042 mg/ mg represented the best threshold to reliably predict 24 HUP of 0.150 g, with high sensitivity, and high specificity in this study. This is consistent with previously published reports [20][21][22]. In addition, SUPOR of 0.01 mg/dl/mOsm/kg H 2 O represented the best threshold to reliably predict 24 HUP of 0.150 g, with high sensitivity and high specificity. This is similar to the study by Wilson et al. [11] in which they found SUPOR of 0.120 with sensitivity of 95% and specificity of 93% to represent normal range proteinuria in 24-hour urine collection. This compared to a SUPCR of less than 0.05 (sensitivity 96%, specificity 90%).

Conclusion
SUPCR is a fast, convenient and reliable method of estimating proteinuria in both HIV and non HIV population from the correlation, Bland Altman plot analysis and ROC curves. The same applies to the SUPOR test. HIV clinics in Sub-Saharan African County may use the tests in quantifying proteinuria in this group of subjects. However, a conversion equation for SUPCR and SUPOR are needed for effective use of both methods in quantitative urine protein estimation.

Limitations of the Study
The sample size is relatively small. A larger sample size preferably involving many centers is desirable. Subjects that took part in the study were not on admission; as a result compliance with strict urine collection may have been compromised by some of the subjects. Ideally the urine should be stored in a fridge at a temperature 2-6ºC during the collection of the urine, it was not certain that all patient complied strictly with this guideline.