The biochemistry and clinical application of urinary oestriol measurement during late pregnancy in the 1990's.

Part 1. The present disillusionment with oestriol measurement as a test of fetoplacental function could be explained by the use of poor methodology and inappropriate normal ranges rather than that the test has lost its usefulness. We have updated the Lever method for measuring oestriol in urine, and examined the automatic TDX system supplied by Abbott Laboratories. The precision of the methods and consistency of results between methods have been determined and normal ranges have been established for both methods. The overall accuracy of collection of 24-hour urine specimens in a routine laboratory setting has also been calculated. The normal ranges suggested as a guide for the TDX method by Abbott were based on those derived from the original method of Brown and were found to be too low and therefore unsuitable for clinical use. This study reports appropriate lower limits of normal for both the updated Lever and the TDX methods. Part 2. The results obtained using the updated Lever method since its introduction in 1991 have been compared with those obtained by the original Brown method during the years 1971-1989. The new, user-friendly Lever method of oestriol assay measurement used in 1991-1993 gave results of equivalent clinical value to the Brown method used in 1971-1989, although the perinatal mortality rates in the tested populations fell from 0.95% to 0.37%. During 1971-1989, low oestriol excretion on 1 or more occasions was associated with a 5.6-fold increase in the perinatal mortality rate (0.66% to 3.67%), whereas in 1991-1993, the factor was 4.4 (0.27% to 1.2%).(ABSTRACT TRUNCATED AT 250 WORDS)