Indications for Continuous Renal Replacement Therapy: Renal Replacement Versus Renal Support

Fluid Overload

Fluid overload may occur as a complication of acute kidney injury or as a complication of other critically clinical syndromes such as congestive heart failure, iatrogenic fluid overload in shock, and acute respiratory distress syndrome. There is increasing evidence that fluid overload in critically ill patients with AKI is associated with adverse outcomes as shown in Table 163.2 . With most of the data coming from observational studies, the question if a positive fluid balance has a direct causative effect on patient outcome or if it is a marker of clinical severity remains to be answered. Nevertheless, one randomized controlled trial in patients with acute distress respiratory syndrome provided evidence of a causal relationship.

In terms of type of RRT to be used in patients with fluid overload, CRRT offers some advantages over intermittent hemodialysis (IHD). Bouchard et al. showed that patients on CRRT were more likely to progress with a lower percentage of fluid accumulation compared with patients treated with IHD. In that observational study, mean fluid accumulation in patients treated with CRRT on the tenth day of treatment was lower than 10%, as compared with patients treated with IHD, in whom mean fluid accumulation on the tenth day was higher than 15% (with an increment from baseline mean fluid accumulation). The study also showed that the adjusted odds ratio for death associated with fluid overload at RRT cessation was 2.52 (95% CI 1.55–4.08). Another study that compared IHD with CRRT found similar results: median cumulative total fluid balances during the first 3 days of therapy were −4005 mL for patients treated with CRRT and +1539 mL for patients treated with IHD, and this negative did not correlate with a decrease in urine output. Furthermore, for IHD patients during the treatment there was a significant decrease in the mean arterial blood pressure from baseline, whereas for CRRT patients the mean arterial blood pressure remained unchanged from baseline.

CRRT allows continuous and slow fluid removal to at least match fluid inputs; on the other hand IHD fluid removal goals must be met in 3 to 4 hours of therapy that leads to transient intravascular underfilling, intradialytic hypotension, and recurrent injury to kidneys, which increases the risk of nonrecovery of renal function. Likewise, a systematic review of 16 studies has found a higher rate of dialysis dependence among survivors who initially received IHD as compared with CRRT (RR 1.99, 95% CI 1.53–2.59).

Fluid removal is an important goal and often is the major goal of RRT for AKI. Although the interpretation of the studies assessing the relationship between fluid overload and mortality is difficult, because sicker patients may receive more fluids, and more severe AKI is often oliguric, CRRT should be considered for patients not achieving adequate fluid balance in IHD techniques.

To optimize fluid overload managing patients' goals for fluid management must be defined initially and adjusted according to the patient's clinical condition as recommended by a recent Acute Dialysis Quality Initiative (ADQI) consensus. A proposed approach to achieve the target fluid balance using CRRT (replacement fluid technique) is to set a fix net ultrafiltration rate and use a variable replacement fluid rate. With this method, fluid balance is achieved by adjusting the amount of replacement fluids, the output is fixed to achieve a solute clearance goal, and replacement fluid rates are changed to allow flexibility in reaching net fluid balance goals. This method allows for constant solute clearance and dissociates clearance parameters from fluid balance. However, one of the disadvantages of this method is that it requires hourly calculations of the amount of replacement fluid to be given with risk for fluid imbalance if rate is not calculated correctly.

Cardiovascular Instability

When applied to hemodynamically unstable patients, CRRT generally is viewed as superior to IHD techniques based on several studies comparing changes in mean arterial pressure (MAP), systemic vascular resistance (SVR), and other hemodynamic parameters. However, CRRT-associated hypotension remains a frequent problem in intensive care unit (ICU) settings. Several physiologic mechanisms have been described. In the early phase of CRRT (within 12 minutes after connection to the extracorporeal circuit), bradykinin release has been shown to be associated with the induction of temporary hypotension. Another determinant of hemodynamic stability during CRRT is the ultrafiltration (UF) rate. Although the UF rate in CRRT is 5 to 6 times slower when compared with IHD, it may nonetheless lead to hypotension if the UF rate exceeds the rate of interstitial fluid movement into the plasma and depletes the intravascular compartment volume. Similarly, rapid removal of urea with high UF rates may decrease plasma osmotic pressure, further decreasing the rate of interstitial fluid movement into the plasma. In the neonate there is an increased risk of hemodynamic instability if more than 10% of the neonate's blood volume is in the extracorporeal circuit. As a result, most neonates weighing less than 8 to 10 kg require blood priming to mitigate this hypotension.

Solute Removal: Uremic Symptoms and Signs

In critically ill patients, uremic syndrome is characterized by multiple-organ deterioration. The most serious consequences are observed in the cardiovascular, neurologic, hematologic, and immunologic systems. Critically ill patients with AKI have an increased protein catabolic rate with negative nitrogen balance and variable urea water distribution. An early and aggressive approach to hyperazotemia is an important therapeutic goal, because the reduction of the level of plasma urea could reduce the rate of complications of acute kidney injury and improves survival because higher blood urea nitrogen also was associated with mortality in logistic regression models. For example, early initiation of CRRT based on blood urea nitrogen levels (<60 mg/dL) in patients with posttraumatic acute kidney injury had a better survival rate as compared with patients with a late CRRT initiation (blood urea nitrogen > 60 mg/dL).

After the initial prescription of CRRT, it is important that providers frequently reassess the response to prescribed CRRT dose using quality measures focused on CRRT dose, such as delivered clearance; ratio of delivered to prescribed dose; effective treatment time; and other measures of solute control. CRRT prescription may require additional modifications, which is why it is important that solute control should be adapted to the changing clinical needs of critically ill patients.

Metabolic Acidosis

In AKI, metabolic acidosis is the most common acid-base abnormality and is due to reduced regeneration of bicarbonate and failure to excrete ammonium ions. Acidosis is related to increased mortality resulting from myocardial electrical and contractility alterations. Severe refractory metabolic acidosis, usually defined as a pH value less than 7.10 or 7.15, is considered a standard indication for RRT. Under most circumstances, RRT can correct abnormal blood pH within 24 to 48 hours. Specific physiologic end points of acid-base intervention with RRT in the critically ill population have not been defined. A complete correction of blood pH is not necessary nor a rapid correction in most if not all critically ill patients.

Four different buffers have been used to correct acidemia in RRT (i.e., bicarbonate, lactate, citrate, and acetate). The most widely used buffer solution is bicarbonate. In IHD, bicarbonate concentrations usually vary between 31 and 39 mEq/L and in commercial solutions used for CRRT, between 25 and 35 mEq/L. Despite a lower concentration, bicarbonate levels more often are normalized with continuous venovenous hemodiafiltration (CVVHDF) than with IHD (71.5% vs. 59.2%, p = .007). When using CRRT for treating acidosis, clinicians must consider that different modalities also may cause different acid-base disturbances. In a retrospective study, continuous venovenous hemofiltration (CVVH) was associated with a lower incidence of metabolic acidosis than CVVHDF (13.8% vs. 34.5%; p < .0001) and a higher incidence of metabolic alkalosis (38.9% vs. 1.1%; p < .0001).

CRRT can be effective to manage patients with lactic acidosis. In one of the largest series, six patients with metformin-associated lactic acidosis were treated with either CVVH (n = 3) or CVVHDF (n = 3). Metabolic acidosis, as well as metformin plasma concentrations, were reduced dramatically in the first 24 hours and/or normalized on the second day in every case with no rebound in acidosis.

When the acidosis is severe or not controlled during dialysis, the highest concentration of bicarbonate levels should be used, and the therapy can be optimized to accommodate the patient's needs. Initiation of CRRT leads to increased bicarbonate and is the preferred modality in this setting, if tolerated, because it provides a greater rapid clearance, whereas CRRT offers more continuous and sustained correction of metabolic acidosis.