Jones S, Thomas G D, Williamson P
Pathology Laboratory, Royal Halifax Infirmary, England.
Acta Cytol. 1996 Mar-Apr;40(2):226-34. doi: 10.1159/000333743.
To measure variation in reporting within and between cytotechnologists in a routine cytology laboratory and to assess if attending training courses and discussing many cases alter agreement levels.
The study involved nine cytotechnologists each screening 100 cervical smears and recording the results. The same 100 smears were given a different identity number and the process repeated about six months later. Between the two occasions, two participants attended a training course, and two others discussed many other cases with the aim of reaching agreement on diagnostic criteria.
The number of smears judged by the participants on either occasion as adequate varied between 87 and 100, and as adequate and negative, between 36 and 65. Percentage intraobserver agreement on the first round diagnosis of a negative smear varied between 88.5 and 100, and of an abnormal smear, between 58.5 and 91.2.
Attending the same training course and in-depth discussion of other cervical smears appear to have had the effect of increasing some aspects of interobserver agreement between the two pairs of individuals involved.