INTRAOBSERVER AND INTEROBSERVER VARIABILITY OF THE AMNIOTIC-FLUID INDEX

OBJECTIVE: We assessed intraobserver and interobserver variations in t he performance of the amniotic fluid index. STUDY DESIGN: A total of 3 4 women in the third trimester with intact membranes were examined twi ce by each of three examiners. The amniotic fluid index was measured o nce by each of the three examiners in order, then repeated by each exa miner in the same order. Numeric displays on the video terminal screen were covered so that the examiners were blinded to actual measurement values. Results were recorded on hard copy for later data analysis. A senior investigator supervised all examinations to monitor consistenc y of technique. RESULTS: Overall, the coefficient of variation for amn iotic fluid index measurements varied from 10.8% within examiners to 1 5.4% between examiners. Both between and within variation of absolute differences increased as the amniotic fluid index value increased, whi le the percent difference decreased. Simply repeating the test by the same examiner will decrease the coefficient of variation significantly . CONCLUSION: Although intraobserver and interobserver variability can not be eliminated, its impact can be minimized by use of the same exam iner for serial studies.