Interobserver agreements in assessing 549 labor admission tests after a standardized training program

Abstract

The labor admission test is a short cardiotocography (CTG) performed upon admission to the maternity ward. The aim of the present study is to examine interobserver agreements when the labor admission tests were assessed by midwives and obstetricians who had received training in interpreting CTG. Five hundred forty-nine high- and low-risk women who delivered at Hammer-fest Hospital were included. The tests were assessed by three midwives and three obstetricians who had completed a standardized training program. The traces were assessed as normal, intermediary, or abnormal. Weighted kappa (kappaw), proportion of agreement (Pa), and predictive values were calculated. Between the pairs of observers, kappaw varied between 0.57 and 0.75; Pa for a normal test between 0.78 and 0.88, and Pa for an intermediary/abnormal test between 0.56 and 0.69. At a cutoff intermediary test, mean sensitivity was 0.43 (range=0.39 -- 0.48), specificity 0.75 (range=0.69 - 0.81, positive predictive value 0.13 (range=0.12 -- 0.15), negative predictive value 0.94 (range=0.94 -- 0.94), likelihood ratio (LR) for a positive test result 1.73 (range=1.53--1.99), and LR for a negative test result 0.76 (range=0.75--0.77). Midwives and obstetricians who had completed the training program achieved good levels of agreements in assessing labor admission tests. The agreements in normal tests were better than those in intermediary/abnormal tests. Obstetric staff should be aware that there are disagreements in assessing labor admission tests; especially in tests assessed as intermediary/abnormal.