Abstract
We performed comparative investigation in visible-spectrophotometric methods for determining carboxyhemoglobin (CO-Hb) in blood samples. About 58% carboxyhemoglobin-saturation (%CO-Hb) of blood samples (sample H) was prepared from control blood by carbon monoxide bubbling and this sample was diluted to be 4/11 and 3/25 with control human blood to prepare moderate and weak CO-Hb saturated samples (sample M and L, respectively). We measured %CO-Hb of four samples, samples H, M, L and control human blood (relative %CO-Hb were 1.00, 0.36, 0.12 and 0, respectively), by four different methods in five different forensic science laboratories. By summing up the measurement results, the method (1), which is described in “Standard method of chemical analysis in poisoning (edited by the Pharmaceutical Society of Japan)”, gave %CO-Hb values that reflected the relative %CO-Hb of the four samples. The method (2), which is an isosbestic point method (developed by Department of Forensic Medicine, Kagawa University) gave higher values compared to the expected ones. The method (3), which is performed with a strong alkaline condition, gave higher values for low %CO-Hb samples by Katsumata's formula (method (3)-1). But the values calculated using the formula improved by Forensic Science Laboratory, Hokkaido Prefectural Police H.Q. (method (3)-2), gave the values reflecting the relative %CO-Hb. The method (4), which is also performed with a strong alkaline method, gave values which reflected the relative %CO-Hb when the Fukui's formula was used for calculation. But the formula modified by Aoki (method (4)-2) gave higher values for the blood samples of low %CO-Hb. In comparison of the three methods that gave the values reflecting the relative %CO-Hb, the method (1) and (3)-2 gave similar values but the measured values obtained by method (4)-1 was lower than the values obtained by method (1) and (3)-2. On the other hand, the method (3)-2 and method (1) showed the large dispersion in the measured values among the laboratories, but the dispersion by the method (4)-1 was small.
