Reliability of cytological follow-up after conization of the cervix; a comparison of three surgical techniques.

A retrospective analysis was made of 154 consecutive conizations between 1974 and 1982. Three surgical methods were compared: (A) Sturmdorf sutures; (B) interrupted vertical sutures; and (C) an 'open' method using only cauterization and no additional stitches. The mean blood loss during conization was not reduced by ligation of descending branches of the uterine arteries. Post-conization haemorrhage occurred in only 3% and did not differ between the three groups. Cervical smears with only ectocervical squamous epithelium or too little material to allow diagnosis were considered inadequate. Unreliable follow-up was defined as inadequacy of more than 25% of a patient's smears. Thirty per cent of all follow-up smears were inadequate (group A 33%; B 41%; C 16%). In 46% of the patients, cytological follow-up was unreliable (group A 48%; B 61%; C 18%) and additional measures such as dilatation of the cervix, endocervical curettage, or hysterectomy were required. The differences between group C and the other two groups were statistically significant (P less than 0.01; chi 2 test).