Serum Creatinine, an important marker for renal function, is routinely measured with the automated Jaffe's method using picric acid. This method suffers from interference by plasma hemoglobin, serum bilirubin, ketones and cephalosporins which are often present in sick babies. We, therefore, developed a specific, sensitive, and rapid method for serum creatinine in neonates without interference from endogenous compounds (bilirubin, glucose, acetoacetic acid, acetone, uric acid, pyruvic acid, hemoglobin, albumin) using capillary electrophoresis (CE) and tested it in six neonates [mean birth weight=1377.5 g (range: 700-3295 g), gestational age=28.3 wks (24-40 wks), postnatal age=4.8 days (1-9 days)]. Blood samples (0.4 ml) were drawn for creatinine assay using Jaffe's technique in the clinical laboratory and an aliquote (0.2 ml) from the same sample for CE. A CE instrument (Waters Quanta 4000E) with a variable wave length, a U.V. detection at 214 nm, an untreated fused silica capillary tube 50 mm i.d. by 40 cm long, sodium phosphate 50μM in 0.1% Brij solution (pH-11.0) were used. Samples were injected electrokinetically for 20 sec. Data were collected with a millenium 2010 software (Millipore Inc., Bedford, MA). The detection limit was 0.88 μmol/L and was linear to 194.8 μmol/L, with a signal to noise ratio of 3.0. The peak area and the creatinine concentration (0.88 μmol /L - 194.8 μmol/L) was linear and passed through the origin. The correlation coefficient was r=0.999 with inter and intra assay coefficient of variations of <6.0%. Mean(±SE) serum creatinines were 86.2±13.3 and 64.3±8.1μmol/L by Jaffe's and CE respectively. Jaffe's was 21.9% greater than CE. We conclude that the routine automated picric acid method overestimates serum creatinine by at least 20%.
