Mitchell H
Victorian Cervical Cytology Registry, Carlton South, Australia.
Acta Cytol. 1994 May-Jun;38(3):310-4.
A study of the intralaboratory and interlaboratory consistency of reporting endocervical cells on cytologic smears is presented. Two sets of 40 slides were compiled from previously reported material. Twenty-two laboratories each reported one set of the slides according to a single coding schedule. The intralaboratory agreement on the endocervical cell code was 80%. Considerable variation was evident between laboratories. While only 14% of the 80 slides received a unanimous endocervical code across the 11 reviewing laboratories, 73 of the 80 slides had agreement by at least 7 of the 11 reviewing laboratories. If the coding option "minor reactive and inflammatory changes in endocervical cells" was amalgamated with the coding option "normal endocervical cells present," a substantial improvement in the level of agreement resulted. The probability of agreement between laboratories did not vary according to whether the subsequent histology/cytology was reported as negative or as a low grade abnormality.