patients during the first week of ICU admission [7]. By applying the KDIGO classification, the authors analyzed data from more than 1,800 patients and found that 57% of them had some level of AKI severity (18, 9, and 30% for classes 1, 2 and 3 respectively). The study confirmed a stepwise increase in mortality with increasing AKI severity (stage 1: OR 2; stage 2: OR 4; stage 3: OR 7). Stages 2 and 3 remained associated with mortality even after adjustment for several confounders. Interestingly, surviving AKI patients of this large cohort showed, at ICU discharge, significantly worse renal function (4.5% were dialysis dependent and about half of them displayed an eGFR <60 mL/min/1.73 m2) compared to non-AKI patients.
Serum Creatinine Criteria for AKI Diagnosis
SCr, a metabolite of creatine (synthesized from glycine and arginine in liver, pancreas and kidneys) is excreted unchanged into urine. SCr is one of the most commonly measured analytes in clinical practice worldwide, since the cost-benefit ratio is undoubtedly favorable. Nonetheless, although SCr remains a landmark biomarker in the diagnosis and classification of AKI, the relationship between SCr and GFR should be taken into consideration in order to keep in mind some limitations for AKI diagnosis:
• SCr levels tend to remain within normal values until approximately 50% of nephrons are lost or when the GFR fall to 60 mL/min/1.73 m2 [8]. For example, kidney donors maintain, after nephrectomy, normal SCr levels (and eGFR) despite having lost an entire kidney. On the contrary, an increase in SCr occurs only after the residual nephrons (commonly <50%) encounter a further loss of kidney function. Therefore, SCr is not an accurate marker of renal function when the GFR is >60 mL/min/1.73 m2 [8].
• Up to 15% of creatinine is actively secreted by the tubules leading to a creatinine clearance (CrCl) and GFR overestimation [9]. This overestimation is quite unpredictable and might vary in the same patient with declining GFR.
• Some drugs inhibit the tubular secretion of creatinine (e.g., trimethoprim and cimetidine) leading to an increase in SCr values when GFR remains constant [10–12]. On the other hand, the production of creatinine, from muscular creatine, could be influenced negatively in severe hepatic disease and positively in rhabdomyolysis [12, 13].
• Another important limitation, especially in critically ill patients and elderly patients with cancer, is that SCr may vary with decrease in muscle mass (i.e., sarcopenia), because creatinine is a product of muscle catabolism [12]. Moreover, decreased production of creatinine has been shown in sepsis [14] and may also result in false negatives. This association with muscle mass explains why similar SCr concentrations may correspond to different values of GFR in subjects with different physical characteristics (e.g., weight, pediatric or geriatric age, gender, or ethnicity).
• SCr can be influenced by diet: food rich in proteins (e.g., red meat) can increase the SCr, and GFR also increases with such food intakes in patients with preserved renal functional reserve (RFR) [15].
• Fluid loading or overload may dilute SCr resulting in a delayed or even missed diagnosis of AKI [16].
• Finally, the KDIGO creatinine criteria require a baseline creatinine value, since the diagnosis is based on changes in concentration compared with a baseline referred to the premorbid level. Since a baseline value may not be available, several surrogates have been proposed [6]. Calculation with an eGFR equation that considers a normal eGFR (75 mL/min/m2) does not take into account the possibility of undiagnosed chronic kidney disease resulting in false positives. Moreover, using hospital or ICU admission SCr does not take into account the possibility of community-acquired AKI [17].
UO Criteria for AKI Diagnosis
UO is an inexpensive and easy test to evaluate renal function and one of the most important clinical aspects commonly observed by physicians in daily clinical practice in surgical/medical wards, ICUs, and operation theaters. It is a marker of AKI expressed in mL/kg/h, but no consensus on whether to use actual or ideal body weight has been reached so far [18]. Instances of 3 to 5 h of consecutive oliguria in patients with septic shock have shown to provide a valuable measure of AKI risk and to outperform serum creatinine criteria in a “renal angina” window of <12 h [19].
Unfortunately, in many clinical conditions, very common in ICU and perioperative medicine, UO criteria may lack specificity and may be too liberal:
• Most of the critically ill and perioperative patients may have transient periods of oliguria in the absence of decreased GFR. To distinguish such episodes from actual AKI, it may be difficult using the 0.5 mL/kg/h threshold.
• Diuretics are among the most frequently administered drugs in critically ill patients for management of fluid balance. Diuretic administration may hypothetically reverse a diagnosis of AKI based on UO criteria only.
GFR, eGFR, and RFR
A persistently reduced GFR implies a diagnosis of chronic kidney disease, whereas an abrupt reduction of the GFR may be used to describe AKI. Therefore, GFR is considered the best overall index of kidney function. Creatinine is the closest to an ideal endogenous substance for measuring GFR, and in case of stable renal function, its levels are usually constant, freely filtered at the glomerulus and not reabsorbed.
GFR can be determined by measuring CrCl using the following simple formula: (urinary creatinine × volume)/(time × plasma creatinine) [20]. Urine collection for 2, 6, 12, or 24 h is required together with the measurement of creatinine concentration in both blood and urine. CrCl tends to overestimate GFR at low levels of renal function [20]. This methodology based on CrCl is clearly time consuming and costly. Nevertheless, as renal function declines, the serum concentration of any solute produced by the body and removed by glomerular filtration will increase proportionally to the rate of GFR reduction: the assessment of serum concentration of such a solute for routine GFR estimation is attractive. Moreover, the development and application of formulae for estimating GFR (and AKI staging) based on creatinine levels is an evident confirmation of the importance of SCr as a renal biomarker.
Therefore, eGFR by creatinine-based equations is commonly used in clinical practice and several options have been proposed [21]. Although GFR and SCr are linked by an exponential relationship and by converting SCr into eGFR allows an easier interpretation of the decline in kidney function, a substantial number of nephrons may already be lost before Scr increases as previously underlined. This crucial issue keeps the eGFR estimation an insensitive marker of kidney function in many clinical conditions.
