We investigated the efficiency, accuracy, and reliability of the ultra
filtration/dipstick methodology commonly used to diagnose myoglobinuri
a. Twenty-five myoglobin-containing urine specimens were filtered by c
entrifugation for 15 min at 1500g through a Centricon-30(TM) membrane
filter, Both the original specimen and filtrate were assayed for myogl
obin. The amount of myoglobin recovered subsequent to filtration varie
d from <1-38%, This poor and variable recovery was independent of samp
le matrix or precentrifugation of the specimens, This was most critica
l for urine specimens with myoglobin concentrations <60 000 mu g/L. Fo
urteen of 18 such filtrates had concentrations <350 mu g/L, a concentr
ation below which a negative result would be obtained by using convent
ional dipstick methods, Thus, the use of this procedure has the potent
ial to misdiagnose patients with myoglobin concentrations associated w
ith increased risk of subsequent renal dysfunction, in particular when
urine myoglobin concentrations are <60 000 mu g/L.