The incidence of acute renal injury (AKI) in ICU is ˜ 5%. The mortality rate increases up to 50% if AKI is a part of multiple organ dysfunction syndrome [1,2]. Prerenal azotemia is responsible for 30–40% of cases of oliguria in ICU; a situation which can be usually reverted by correcting the underlying disorder, but prolonged or severe prerenal azotemia may proceed to oliguric renal azotemia .
The outcome of patients with acute renal azotemia differs significantly from those with prerenal azotemia; the latter reflects the physiological hemodynamic regulatory mechanisms of conservation of water and electrolytes in response to hypovolemia and decreased glomerular filtration rate (GFR) [1–3].
Early detection of renal and excluding prerenal azotemia improve outcome . Relying solely on the patient's response to fluid challenge in order to differentiate prerenal azotemia from renal azotemia, may lead to massive fluid overload in the latter [4,5].
Many laboratory tools have been used for differentiating renal from prerenal azotemia including blood urea/serum creatinine ratio, urine analysis (granular casts in patients with renal azotemia), urine osmolarity and specific gravity, urinary sodium level, urinary to plasma creatinine ratio (U/P Cr). However; all of them lack reasonable sensitivity and specificity [4–6].
One of these laboratory tools is the fractional excretion of sodium (FENa) which showed a reasonable sensitivity and specificity. FENa is the percentage of Na+ filtered by the kidney which is not reabsorbed and excreted in the urine. It depends upon the fact that in the state of decreased renal blood flow, as in hypovolemia, there is enhanced reabsorption of Na+ from renal tubules, making the fraction of the filtered Na+ that is excreted in urine less than normal which means lower FENa in prerenal azotemia [4,5,7].
But unfortunately FENa (despite a simple lab tool) is affected by many drugs and medical conditions. Drugs that increase Na+ excretion in urine as diuretics and dopamine  lead to false increase in FENa, and drugs that decrease Na+ excretion as norepinephrine lead to false decrease in FENa [8,9]. Because 70% of ICU pts with AKI receive diuretic therapy, and many patients with circulatory shock receive dopamine and norepinephrine, the results of FENa are not justified in such conditions . Moreover, lactic acidosis that occurs with circulatory shock may result in natriuresis and may falsely increase FENa as well .
Solutes other than Na+, including urea, have been suggested to improve the diagnostic ability in the above mentioned clinical situations instead of FENa . The fractional excretion of urea (FEurea) is mainly dependent on passive forces in the kidney and so is less influenced by the diuretic therapy .
The purpose of this study was to compare FENa and FEurea in differentiating renal from prerenal azotemia, and to detect the effect of diuretics on their handling. Few studies have specifically addressed this issue, but never in patients with AKI complicating circulatory shock.
This prospective study was conducted on 40 Egyptian patients with AKI complicating circulatory shock admitted to the general ICU of Al Fayoum University Hospital, from July 2011 to July 2012. Consents were taken from all of them according to the local ethics committee approval.
The patients were divided into two groups:
- Group-1: Included 26 patients; mean age 60.0 ± 15.15 (21–88) years (10 males) that had acute prerenal azotemia, and was subdivided into:
Group 2: Included 14 patients; mean age 56.29 ± 19.5 (23–92) years (18 males) that had acute renal azotemia.
- Group-1a: Included 12 patients who did not receive diuretics in the last 24 h before taking the urine and blood samples.
- Group-1b: Included 14 patients who received diuretics, in the last 24 h before taking the urine and blood samples.
Regarding age and sex, groups-1 and 2 were comparable.
All patients presented with AKI complicating circulatory shock during their stay in the ICU:
- Circulatory shock was diagnosed according to the Empiric Criteria , which requires 4 out of the following:
Acute kidney injury was diagnosed according to the RIFLE-criteria (Risk, Injury, Failure, Loss and End-stage Kidney) in oliguric patients : (Table 1)Prerenal azotemia diagnosis was settled when ≥6 of the following criteria was found :
- Ill appearing or altered mental status.
- Heart rate > 100 beat/min.
- Respiratory rate > 22/min, or paCO2 < 32 mmHg.
- Arterial hypotension (systolic blood pressure ≤ 90 mmHg, decrease in systolic blood pressure by 40 mmHg from baseline or mean arterial blood pressure ≤ 70 mmHg), for >20 min duration.
- Urine output < 0.5 mL/kg/h.
- Arterial base deficit < −5 mEq/L, or lactic acidosis (>2 mmol/L)
Acute renal failure diagnosis was settled when <6 of the previously mentioned criteria were present .
- History: Volume depletion, congestive heart failure, severe liver disease or other edematous states and anaphylaxis favor prerenal azotemia, while exposure to exogenous toxins such as medications and recent radiographic procedures with contrast, or endogenous toxins as in the case of myoglobinuria would favor acute tubular necrosis
- Physical examination: Loss of skin elasticity and skin turger in dehydrated patients, orthostatic changes in heart rate, presence of ascites or pedal edema suggests prerenal state.
- Response to fluid challenge: Favors prerenal state .
- Urine analysis: No abnormalities or hyaline and fine granular casts in prerenal states, while presence of muddy brown granular casts and renal tubular epithelial cells or casts indicates acute renal failure .
- Urine specific gravity ≤ 1010 favors acute renal failure, while >1018 favors prerenal state.
- Urinary to plasma creatinine ratio (U/P Cr): If >20 is consistent with prerenal while levels <15 suggest acute tubular necrosis.
- Blood urea/serum creatinine ratio: Would favor prerenal if >40 and acute renal failure if <40.
- Urinary sodium (UNa): If <15 mEq/L favors prerenal azotemia, while >20 mEq/L is consistent with acute tubular necrosis.
- Chronic kidney disease.
- End stage kidney disease receiving renal replacement therapy.
- Obstructive nephropathy.
- Renal transplantation.
- Osmotic diuresis as in usage of mannitol or acetazolamide .
All patients were subjected to the following:
- Full history and clinical evaluation including the routine ICU monitoring; fluid balance sheet and hourly urine output measurement.
- Routine renal function tests:
All patients had subjected to abdominal ultrasound to exclude any chronic kidney disease or obstructive nephropathy.
Fractional excretion of sodium and fractional excretion of urea
- Blood urea, serum creatinine, and Na+ and K+ levels.
- Blood urea/serum creatinine ratio.
- Urinary/serum creatinine ratio.
- Urinary Na+ level.
- Urine analysis.
- Urine specific gravity.
- Albumin/creatinine ratio or 24 h proteins in urine were used to exclude diabetic nephropathy .
- FENa was calculated as :
- FEurea nitrogen was calculated as :
Laboratory preparation of urea estimation to calculate FEurea nitrogen
- Serum or heparinized plasma (except ammonium heparin).
- Urine was diluted at 1/20 to 1/50 with distilled water before analysis.
- Serum and plasma are stable up to 24 h at room temperature, for one week at 4 °C. Frozen between −15 and −20 °C, these samples are stable for at least 2–3 months.
- Urine samples are stable up to 4 days if stored at 4–8 °C. Urines could be preserved with thymol to avoid bacterial action or by maintaining the pH below 4.
- Blood and urine chemistries and electrolytes were performed on a Beckman automated analyzer (Beckman Instruments, Fullerton, CA, USA). (SEPPIM ELITech Group Company, European manufacturer of clinical chemistry diagnostic kits.)
(5)According to the medical situation and the cause of the circulatory shock, all patients continued their medications including the vasopressor drugs during their enrollment in the study.
The collected data were organized, tabulated and statistically analyzed using Statistical Package for Social Sciences (SPSS Inc., USA) version 15.0. For quantitative data, the means ± standard deviations (SD) were calculated. The independent t-test and one-way ANOVA were used to test the statistical differences between groups. For qualitative data, the number and percent distribution were calculated, chi square (χ2) was used as a test of significance. Pearson's and Spearman's correlations were used to determine the relation between FENa and FEurea with variables of the study when applicable. ROC-curve was used to determine the cut-off point at which highest sensitivity and specificity of FENa, FEurea and other variables were differentiated between prerenal and renal failures. P-values ≤ 0.05 were considered statistically significant while ≤0.001 were considered highly statistically significant.
The patients who suffered from renal azotemia (group-2) significantly needed the mechanical ventilation and the renal replacement therapy than the patients with prerenal azotemia (group-1). Furthermore, the mortality rate was significantly higher in group-2 (Table 2). But, the number of patients who needed vasopressor drugs in their medication in both groups was comparable (Table 2).
The results of the routine lab parameters were comparable in both groups (Table 3).
Compared to group-2, the values of blood urea/serum creatinine ratio, the urinary creatinine/plasma creatinine ratio and the total urine output in the 1st 24 h were significantly higher, and the serum creatinine was significantly lower in group-1. On the other hand, the values of blood urea, serum Na+ and urine Na+ in both groups were comparable (Fig. 1).
Compared to group-2 (renal patients), both FENa and FEurea values of group-1 were significantly lower (Fig. 2).
Using one-way ANOVA test, both FENa and EFurea showed significantly incremental values with higher RIFLE-criteria from mild AKI (RIFLE-R), moderate AKI (RIFLE-I) to severe AKI (RIFLE-F) class (Table 4).
Weak correlations could be detected between the serum creatinine and both FENa and FEurea values (r = 0.36, P < 0.05 and r = 0.37, P < 0.05) respectively, but good correlation could be detected between FENa and FEurea values (r = 57, P < 0.0010 (Fig. 3).
Compared to FENa, on using FEurea as a predictor in differentiating renal from prerenal azotemia in all patients, higher sensitivity (78.6%), specificity (88.5%) and overall accuracy (85%) could be detected (Fig. 4A), than FENa; sensitivity (71.4%), specificity (69.4%) and overall accuracy (67.5%) (Fig. 4B).
For patients who received diuretics, FEurea as a predictor also could differentiate renal from prerenal azotemia (group-1b) with higher sensitivity (78.4%), specificity (92.1%) and overall accuracy (85.7%) (Fig. 4C), than FENa; sensitivity (64.3%), specificity (58.4%) and overall accuracy (59.4%) (Fig. 4D).
As a predictor of differentiation, the cutoff value of FEurea was 34.12% (P < 0.001) and the cutoff value of FENa was 1.1% (P < 0.05).
Fractional excretion of sodium (FENa) has been routinely used for differentiating renal from prerenal azotemia . However, the values of FENa are usually affected by the commonly used medications as diuretics, dopamine and norepinephrine, and also by some clinical conditions as myoglobinuria, radiocontrast induced renal failure and different causes of metabolic acidosis . On the contrary, the fractional excretion of urea (FEurea) is suggested to be used instead of FENa with better reliability to differentiate between prerenal from renal azotemia, as its values are not affected by the same drugs or medical conditions .
In our study we performed a prospective observation in a general ICU on patients with acute kidney injury (AKI) complicating circulatory shock, to explore the difference between using both FENa and FEurea in differentiating renal from prerenal azotemia, and up to our knowledge this is the first time to perform such a comparison in patients with AKI complicating circulatory shock in which 68% of them needed dopamine and norepinephrine together in their ICU medication, making the identification of the effect of each drug alone on FENa and FEurea statistically difficult.
In this study, we used the RIFLE-criteria to classify our patients with AKI into mild renal affection (RIFLE-R), moderate renal affection (RIFLE-I) and severe renal affection (RIFLE-F). We had detected significantly incremental values of both FENa and FEurea with the higher class of the RIFLE-criteria, and to our knowledge, it is the first time to compare FENa and FEurea versus the RIFLE-criteria.
Our study showed a fair correlation between the routine serum creatinine marker and both FENa (r = 0.36, P < 0.05), and FEurea (r = 0.37, P < 0.05), respectively. The poor correlations can be explained by the serum creatinine increase in both renal and prerenal azotemia, while FENa and FEurea increase in renal azotemia and decrease in prerenal azotemia . These findings agree with the Carvounis study, in which there were fair correlations between serum creatinine and both FENa and FEurea .
Also, we found significantly higher values of both FENa and FEurea in patients with renal azotemia than those with prerenal azotemia (P < 0.05 and P< 0.001, respectively). These also agree with Carvounis et al., who studied 102 patients with AKI; 76% with prerenal azotemia showed significantly lower FENa and lower FEurea values (P < 0.001), while 24% of patients with renal azotemia showed significantly higher FENa and higher FEurea values (P < 0.001) .
It is important to mention here that good correlation could be detected between FENa and FEurea (r = 0.67, P < 0.05) in our study. This finding supports the trend of combining both FENa and FEurea in one panel; “the panel of fractional excretion of solutes” that includes FENa and FEurea which can provide the most important insight into the differentiation between renal and prerenal azotemia in various clinical states as suggested by Diskin . However, our study showed that FEurea had better sensitivity (78.1% vs. 71.4%), specificity (88.5% vs. 69.4%) and overall accuracy (85% vs. 67.5%) than FENa in differentiating renal from prerenal azotemia.
Those findings agree with Dewitte et al., in which FEurea had better sensitivity (83% vs. 49%) and specificity (75% vs. 71%) than FENa in differentiating renal from prerenal azotemia . But disagree with Pépin et al., in which 99 patients with AKI showed that sensitivity and specificity of FEurea were 79% and 33%, and concluded that FEurea could not be used as an alternative tool in differentiating renal from prerenal azotemia as it lacks reasonable specificity . The point of difference from our study is that they mainly included patients with preexisting chronic kidney disease (40% of their population) without correction of the FEurea values as described by Nguyen et al., to compensate for the effect of the chronic kidney disease on the FEurea value , while in our study all patients with chronic kidney disease were excluded.
In our study FENa had a cutoff point of 1.1% for differentiating renal from prerenal azotemia with reasonable sensitivity and specificity (71.4% and 69.4%, respectively). These agree with Carvounis et al., who had a cutoff point of 1% with high sensitivity and specificity (91% and 82%, respectively) . However, the relatively lower sensitivity and specificity of our FENa values can be attributed to; all our patients who had circulatory shock with lactic acidosis and both dopamine and noradrenaline infusion, which were additional factors that led to increase in the false FENa values.
In our study, FEurea had a cutoff value at 34.1% for differentiating renal from prerenal azotemia with a sensitivity and specificity of 78.1% and 88.5%, respectively. This agrees with Carvounis et al., which had a cutoff point of 35% with sensitivity and specificity of 85%, 92%, respectively . But disagrees with Dewitte et al., who studied 47 ICU patients with AKI, in which the best cutoff point of FEurea was 40% ; a level which is higher than ours. This can be explained as the critically ill patients, as in Dewitte et al. study, commonly have hypercatabolic disease, which may increase urea production and excretion thereby increasing FEurea . The point of difference from our study is that, all our patients had circulatory shock with decreased tissue perfusion which affects the catabolic state of the tissues making FEurea cutoff point in our study slightly lower.
In our study, FEurea was almost not affected by the use of frusemide diuretic therapy; sensitivity, specificity and overall accuracy were 78%, 92% and 85% and 78%, 88% and 85%, respectively without diuretics. The use of frusemide obviously affected FENa values and caused a decrease in the sensitivity (64% vs. 71%), specificity (58% vs. 70%) and overall accuracy (59% vs. 69%), respectively. Those findings agree with Carvounis et al., in which the sensitivity of FEurea was almost not affected by the use of diuretics (92% vs. 91%) while that of FENa was markedly decreased with the use of diuretics (92% vs. 52%). But disagree with Darmon et al., who concluded that FEurea may be of little help in distinguishing renal from prerenal azotemia in critically ill patients receiving diuretic therapy, however, they stated that few of their patients received diuretics and poor performance of the urinary indices was therefore related to low statistical power .
The mortality rate in our study was significantly higher in renal azotemia (50%) than in prerenal azotemia (19.2%), (P< 0.05). This is supported by Vincent et al. and by Angus et al., who found that the mortality rate was about 20% in single organ failure, 40% in two organ failure, 70% in three organ failure and reaching 90% in four organ failure [20,21].
The cutoff points of both FENa and FEurea as a predictor of mortality were not statistically justified. This is explained as all our patients had circulatory shock, and so patients with mild renal impairment may die from their severe shock state, and patients with severe renal affection may survive if their shock could be rapidly corrected. This cutoff point calculation is recommended only in patients with AKI without other organ affections or shock that may affect the mortality.
The limited number of our patients made it difficult to perform reliable statistical tests to explore the relations between both FENa and FEurea vs. different types of shocks, which would have added a lot of information to our study.
Although both fractional excretion of urea (FEurea) and fractional excretion of sodium (FENa) are feasible, reproducible, and inexpensive markers used in differentiating renal from prerenal azotemia, in our study FEurea showed higher sensitivity and specificity than FENa, not only in differentiating renal from prerenal azotemia in critically ill patients complicating circulatory shock, but also its values were not affected by the use of diuretics like FENa in the same group of patients.
 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 2005;294(7):813-818.
 Abernethy VE, Lieberthal W. Acute renal failure in the critically ill patient. Crit Care Clin 2002;18(2):203-223 [Review].
 Hirsch S. Prerenal success in chronic kidney disease. Am J Med 2007;120(9):754-759 [Review].
 Westhuyzen J, Endre ZH, Reece G, Reith DM, Saltissi D, Morgan TJ. Measurement of tubular enzymuria facilitates early detection of acute renal impairment in the intensive care unit. Nephrol Dial Transplant 2003;18(3):543-551.
 Fahimi D, Mohajeri S, Hajizadeh N, Madani A, Esfahani ST, Ataei N, et al. Comparison between fractional excretions of urea
in children with acute kidney injury
. Pediatr Nephrol 2009;24(12):2409-2412.
 Krutzik M, Foster G. Endocrinologic and nephrologic disorders. In: Brenner M, editors. Critical care and cardiac medicine. Current Clinical Strategies Publishing, Inc.; 2006. p. 95-107.
 Carvounis CP, Nisar S, Guro-Razuman S. Significance of the fractional excretion
in the differential diagnosis of acute renal failure. Kidney Int 2002;62(6):2223-2229.
 Bacic D, Capuano P, Baum M, Zhang J, Stange G, Biber J, et al. Activation of dopamine D1-like receptors induces acute internalization of the renal Na+
/phosphate cotransporter NaPi-IIa in mouse kidney and OK cells. Am J Physiol Renal Physiol 2005;288(4):F740-F747.
 Van Biesen W, Yegenaga I, Vanholder R, Verbeke F, Hoste E, Colardyn F, et al. Relationship between fluid status and its management on acute renal failure in intensive care unit patients with sepsis: a prospective analysis. J Nephrol 2005;18(1):54-60.
 Lattanzio MR, Kopyt NP. Acute kidney injury
: new concepts in definition, diagnosis, pathophysiology, and treatment. J Am Osteopath Assoc 2009;109(1):13-19.
 Kline JA. Shock in Rosen's emergency medicine: concepts and clinical practice [Edited by Marx JA, Hockberger RS], 5th ed., vol. 3, No. 4. Mosby Inc.; 2002. p. 39-41.
 Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, et al The Acute Kidney Injury
Network. Acute Kidney Injury
Network: report of an initiative to improve outcomes in acute kidney injury
. Crit Care 2007;11(2):R31.
 Klahr S, Miller SB. Acute oliguria. N Engl J Med 1998;338(10):671-675.
 Dwinnell BG, Anderson RJ. Diagnostic evaluation of the patient with acute renal failure. In: Schrier RW, editor. Atlas of diseases of kidney. Philadelphia: Current Medicine Inc.; 1999. p. 12.1-12.12.
 Diskin CJ, Stokes TJ, Dansby LM, Radcliff L, Carter TB. Toward the optimal clinical use of the fraction excretion of solutes in oliguric azotemia. Renal Failure 2010;32(10):1245-1254.
 Dewitte A, Biais M, Petit L, Cochard JF, Hilbert G, Combe C, et al. Fractional excretion
as a diagnostic index in acute kidney injury
in intensive care patients. J Crit Care 2012;27(5):505-510.
 Pépin MN, Bouchard J, Legault L, Ethier J. Diagnostic performance of fractional excretion
and fractional excretion
in the evaluations of patients with acute kidney injury
with or without diuretic treatment. Am J Kidney Dis 2007;50(4):566-573.
 Nguyen MT, Maynard SE, Kimmel PL. Misapplications of commonly used kidney equations: renal physiology in practice. Clin J Am Soc Nephrol 2009;4(3):528-534.
 Darmon M, Vincent F, Dellamonica J, Schortgen F, Gonzalez F, Das V, et al. Diagnostic performance of fractional excretion
in the evaluation of critically ill patients with acute kidney injury
: a multicenter cohort study. Crit Care 2011;15(4):R178.
 Vincent JL, de Mendonça A, Cantraine F, Moreno R, Takala J, Suter PM, et al. Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study. Working group on “sepsis-related problems” of the European Society of Intensive Care Medicine. Crit Care Med 1998;26(11):1793-1800.
 Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001;29(7):1303-1310.