Cervical intraepithelial neoplasia III. An estimate of screening error rates and optimal screening interval

The detection rates of grade III cervical intraepithelial neoplasia (CIN III) in women at the initial and subsequent screenings were studied in Florence for the period 1979 to 1982 to assess the optimal interval between tests and the sensitivity of the screening (cytology plus advised colposcopy in cases with slightly abnormal Papanicolaou tests). The protective interval was calculated to be about three years. No differences in the risk of developing CIN III were observed between women with only an initial screening and women rescreened within the three-year interval. Screening sensitivity was calculated according to two different models: (1) when assumed to be a constant, the sensitivity of our cytologic-colposcopic screening was 0.71, with 0.64 for cytology alone; (2) when assumed to decrease in subsequent screening rounds, the sensitivity was calculated at 0.60 (0.50 in subsequent rounds), with 0.52 for cytology alone. The benefit of an early cytologic rescreening is discussed and compared to a wider use of colposcopy, which, with lower costs, permits a good sensitivity utilizing only one round of screening.