INTEROBSERVER AND INTRAOBSERVER VARIATION IN THE AMNIOTIC-FLUID INDEX

Objective: To measure inter- and intra-observer reliability in the amn iotic fluid index (AFI). Methods: To measure inter-observer reliabilit y, three observers measured the AFI at random in 32 patients without k nowing each other's results. To determine intra-observer reliability, each investigator made repeated measurements on ten other patients. Pa tients at 26-41 weeks' gestation who were referred for an antenatal ul trasound examination were asked to participate. Intra-observer reliabi lity and interobserver reliability were expressed as intra-class corre lation coefficients. Inter-observer agreement was also expressed as th e kappa statistic after dichotomizing the AFI (less than 8 cm, 8 cm or more). Results: The intra-class correlation coefficient for intraobse rver reliability exceeded 0.8, and for inter-observer reliability it w as 0.67 (95% confidence interval [CI] 0.52-0.79), with an overall kapp a of 0.44 (95% CI 0.22-0.65). The percentage disagreement between pair s of observers reached 22-28%. Conclusion: Measurements of AFI have go od intraobserver and moderate inter-observer agreement. In clinical us e, an AFI level near a cutoff value should be interpreted with caution .