| Abstrakt: |
A serum estrogen assay was evaluated for use in an indigent population in terms of its practicality as well as accuracy, precision, and other statistical parameters. Seventy patients were chosen retrospectively from the obstetric high-risk clinic at the University of Alabama Medical Center on the basis of delivery of a normal infant (Apgar score of 7 or greater). Twenty-four-hour total urinary estrogen (UE), the estrogen/creatinine ratio (E/C), and serum estrogen (SE) were determined on these patients throughout the third trimester of pregnancy. Correlation coefficients for 218 paired samples were highly significant (p< 0.01) between SE and UE (r= 0.31), SE and E/C (r= 0.28), and UE and E/C r= 0.72). The over-all means and standard deviations for UE, E/C, and SE, respectively, were 27.5 ± 15.3 mg. per 24 hours, 22.1 ± 11.2 mg. per gram, and 26.0 ± 11.6 ng. per milliliter while the equations for the best-fit regressions were y = −31.94 + 1.79x, −30.37 + 1.57x, and −16.46 + 1.26x, demonstrating that the rise in weekly SE levels is slower than for UE parameters. Since only one case of overscreening existed in 70 patients, it can be concluded that SE is a most useful clinical aid in screening the high-risk obstetric population for indications of possible fetal compromise. This assay system has the necessary practicability to be useful under the conditions presented by an indigent population seen predominantly as outpatients. |
