Randall D. Jenkins

Early experience with continuous arteriovenous hemofiltration in critically ill pediatric patients

The applicability of continuous arteriovenous hemofiltration (CAVH) for renal replacement therapy was evaluated in three infants and two young children with catastrophic medical and surgical illnesses. In the first four patients, CAVH was used in conjunction with either peritoneal or hemodialysis. In the fifth patient, CAVH was the sole renal replacement therapy employed; in this critically ill anuric infant, we were best able to evaluate the ability of CAVH to continuously control fluid, electrolyte, and acid-base balance, and allow the administration of adequate parenteral nutrition. The difficulties encountered were related to anticoagulation, establishment of adequate vascular access, and selection of an appropriate hemofilter for the performance of the technique. Despite the application of suction-assistance, we were unable to effectively employ a prototype pediatric hemofilter to attain a level of plasma ultrafiltration consistent with the objectives of therapy. However, we were able to effectively and safely employ an adult hemofilter for these purposes; modifications were made in the adult hemofilter system before its application in the smallest pediatric patients. Our experience suggests that, even in critically ill infants, CAVH can be successfully applied as an effective renal replacement therapy. However, further experience is required before its potential impact on patient survival can be assessed.

Accuracy of intravenous infusion pumps in continuous renal replacement therapies.

Most extracorporeal continuous renal replacement therapies (CRRT) require inflow pumping of either dialysate, filtrate replacement solution, or both. Outflow of spent dialysate and ultrafiltrate can be accomplished by gravity drainage or pump. Intravenous infusion pumps have been commonly used for these purposes, although little is known about the accuracy of these pumps. To evaluate accuracy of two different types of intravenous infusion pumps used in CRRT, we studied flow rates at nine different pressure variations in three piston type and three linear peristaltic pumps. The results showed that error of either pump was not different for flow rates of 4 and 16 ml/min. Both types of pumps were affected by fluid circuit pressures, although pressure conditions under which error was low were different for each pump type. The linear peristaltic pumps were most accurate under conditions of low pump inlet pressure, whereas piston pumps were most accurate under conditions of low pump pressure gradient (outlet minus inlet) of 0 or −100 mmHg. The magnitude of error outside these conditions was substantial, reaching 12.5% for the linear peristaltic pump when inlet pressure was −100 mmHg and outlet pressure was 100 mmHg. Error may be minimized in the clinical setting by choosing the pump type best suited for the pressure conditions expected for the renal replacement modality in use.

Maximum ultrafiltration rate in continuous arteriovenous hemofiltration does not occur at the lowest level of the ultrafiltrate collection chamber.

Chronic renal failure is known to raise serum magnesium concentration. However, its effect on intracellular free cytosolic magnesium concentration ([Mg2+]i) has not been clearly delineated. Likewise, whereas hemodialysis (HD) lowers serum magnesium concentration, its effect on [Mg2+]i is unknown. It should be noted that, based on other studies, [Mg2+]i is inversely related to arterial blood pressure in the general population. Given the high prevalence of hypertension in end-stage renal disease (ESRD), the study of [Mg2+]i in this population is of interest. We measured platelet [Mg2+]i in nine ESRD patients, six of whom were hypertensive. Samples were obtained from arterial and venous lines at the start of an HD session, and from the arterial line at the end of an HD session. Five normal volunteers served as controls. [Mg2+]i was measured by spectrofluorometry using Mag-fura-2-AM. Serum magnesium and calcium concentrations were determined by atomic absorption. Compared with the normal control group, the ESRD patients exhibited significantly lower [Mg2+]i but higher serum magnesium concentrations. As expected, HD lowered the serum magnesium concentration. However, platelet [Mg2+]i was unaffected by either a single passage through the dialyzer or the entire dialysis procedure. [Mg2+]i did not correlate with serum magnesium or calcium concentrations, nor with the presence or absence of hypertension.

Operational instability in extracorporeal filtration of blood

After encountering recurrent flow stasis in continuous arteriovenous hemofiltration (CAVH) clinical systems, a combined experimental and theoretical investigation was undertaken of hydraulic operational behavior of extracorporeal systems operated with blood of high hematocrit. Theoretical analyses were developed for selected modalities of extracorporeal blood purification. These modalities include hemofiltration, volume-controlled hemofiltration, CAVH, and volume-controlled CAVH. These analyses have revealed that steady state hydraulic operation may not always be possible when the venous return blood hematocrit becomes elevated. For all modalities, except volume-controlled hemofiltration, one would expect to encounter a departure from steady state operation at a critical hematocrit specific to that system. One would typically not expect to encounter this operational instability phenomenon, unless venous hematocrit was greater than 50. When operational instability is encountered, progressive hemoconcentration of blood and elevation of hemofilter blood pressure occur despite no further perturbation of the system. For modalities in which blood is pumped, the pressure rises until such time as mechanical integrity of the circuit is breached. For the arteriovenous modalities, the pressure rises, and blood flow and uhrafiltration ultimately cease. Laboratory experiments were performed to validate this theory. Blood was perfused through hemofilters of high hydraulic permeability. The ultrafiltrate pressure was varied stepwise from a positive pressure sufficient to prevent net uhrafiltration (or backfiltration) to approximately – 50 mm Hg. Examination of hemofilter pressures and uhrafiltration rates in twenty-four experiments showed a return to steady state operation after most perturbations in ultrafiltrate pressure. In eleven of these experiments (only those in which the venous hematocrit was greater than 47), a subsequent small perturbation in ultrafiltrate pressure was followed by a large upward pressure excursion in the pre-and posthemofilter pressures such that the experiment had to be terminated to avoid disruption of tubing or membrane. No clotting occurred, and the experiments were repeatable with the same or different hemofilters. The experimental results are in reasonably good agreement with the theoretical analysis for pumped uhrafiltration. Hematocrit values at which this operational instability occurs can be predicted based on the shape of the apparent viscosity versus hematocrit curve and system design parameters, in particular hemofilter hydraulic permeability and venous catheter diameter. These hematocrit values may define an upper limit of postfilter hematocrit at which some blood purification systems can operate.

A Mathematical Model for Flow, Pressure, and Ultrafiltration Rate in Extracorporeal Filtration of Blood

A mathematical model of extracorporeal hydraulics (pressures and flows) has been developed for common extracorporeal blood purification modalities. The model includes a description of the blood pathwa