Effects of Postmortem Hemolysis and Ultrafiltration on Creatinine Detection Results

To investigate the interference of postmortem hemolysis on the detection of creatinine and whether ultrafiltration can reduce the interference.

A total of 33 non-hemolyzed whole blood samples from the left heart were collected. Hemolyzed samples with 4 hemoglobin mass concentration gradients H1-H4 were artificially prepared. Ultrafiltration was performed on each hemolyzed sample. Creatinine concentrations in non-hemolyzed serum (baseline serum), hemolyzed samples and ultrafiltrate were detected. Bias (B), Pearson correlation and receiver operator characteristic (ROC) of baseline creatinine concentration between before and after ultrafiltration were analyzed.

As the hemoglobin mass concentration increased, B of the hemolyzed samples in the H1-H4 groups gradually increased, the |B| was 2.41(0.82, 8.25)-51.31(41.79, 188.25), reaching a maximum of 589.06%, and there was no statistically significant between the creatinine concentration and the baseline creatinine concentration (P=0.472 7, r=0.129 5). After ultrafiltration of hemolyzed samples, the interference of creatinine concentration in ultrafiltrate was significantly reduced, the |B| was 5.32(2.26, 9.22)-21.74(20.06, 25.58), reaching a maximum of 32.14%, and there was a positive correlation with baseline creatinine concentration (P<0.05, r=0.918 2). In the hemolyzed samples of H3 and H4 groups, there were 7 false-positive samples and 1 false-negative sample; in the ultrafiltrate samples, there were no false-positive sample and 1 false-negative sample. ROC analysis results showed the hemolyzed samples were lack of diagnostic value (P=0.117 5).

The postmortem hemolysis significantly interferes creatinine detection results of blood samples, ultrafiltration can reduce hemolysis-induced interference in postmortem creatinine detection.