Abstract
We describe a modification of the second-derivative spectrophotometric assay for carboxyhaemoglobin in blood. Using the original procedure we have often observed a time-dependent change in the signal leading to unreliable results. By using phosphate-buffered-saline (pH 7·4) instead of ammonia to dissolve the deoxygenating agent (sodium dithionite), we obtained stable and reproducible second-derivative signals from which the percentage carboxyhaemoglobin in the patients' samples can be calculated.
