Objective. To quantify glucose balance related to continuous arteriove
nous hemofiltration (CAVH) when a glucose-rich replacement fluid is us
ed for the plasma ultrafiltrate removed. Design: Prospective, noninter
vention study. Setting. Medical/surgical and cardiac surgical intensiv
e care units of a university hospital. Patients: Critically ill patien
ts (n = 20) with acute oliguric renal failure undergoing CAVH. Interve
ntions: None. Measurements and Main Results. Timed collections of CAVH
effluent were analyzed and other observations were made. Serum creati
nine and blood urea nitrogen concentrations decreased substantially du
ring CAVH in each patient. The mean measured glucose concentration of
the replacement fluid (Dianeal 1.5%) was 1.40 +/- 0.11 (SD) g/dL (77 /- 6 mmol/L) and rate of infusion was 1.39 +/- 0.43 L/hr. Effluent vol
ume was 1.51 +/- 0.49 L/hr and glucose was 0.47 +/- 0.10 g/dL (26 +/-
5 mmol/L). The glucose content of the replacement fluid infused was co
nsistently and substantially greater than that value of the effluent r
emoved during the same period. Thus, the average net glucose uptake in
relation to the CAVH circuit was 11.9 +/- 3.1 g/hr (range 4.3 to 17.6
). Serum glucose concentrations increased in each patient with initiat
ion of CAVH (from 135 +/- 44 to 278 +/- 80 mg/dL 7.4 +/- 2.4 to 15.3
+/- 4.4 mmol/L!; p < .001). Negative fluid balance achieved during CAV
H was approximately 100 mL/hr. Conclusions: CAVH using predilution wit
h replacement fluid is effective in managing uremia and fluid overload
in critically ill patients. The use of a glucose-rich replacement flu
id is accompanied by the net uptake of large amounts of glucose, appro
aching 300 g/day on average in our patients and representing a major e
xogenous calorie source. This finding has important implications for t
he metabolic management of critically in patients during CAVH and shou
ld be taken into account in prescribing their nutritional support.