An analysis of referral patterns from a breast screening unit

Integrating Self-Referral for Mammography into Organised Screening: Results from an Italian Experience

OBJECTIVE

To compare self-referred screenees with respondents to invitation for main performance indicators of mammography screening.

SETTING

First round of an organised, population-based screening programme in six districts of northern Italy.

METHODS

The screening test was a two-view mammography. Eligible women aged 50-69 years were invited. Self-referred attendees were accepted if they were eligible for screening and had not yet been invited or had been invited >6 months before presentation. Age-specific performance indicators were compared with the calculation of their ratio and 95% confidence intervals (CIs). Total ratios were age standardised. In situ carcinomas were excluded.

RESULTS

The eligible population was 183 542 women. There were 112 188 respondents to invitation and 20 280 self-referred attendees. Self-referral rate was inversely related to age. Performance indicators were as follows: recall rate, 5.6% for self-referred attendees vs 5.5% for respondents (ratio 1.02, 95% CI 0.96 to 1.08); total aspiration cytology rate, 37.3% vs 28.3% (1.37, 1.24 to 1.51); biopsy rate, 17.0 vs 12.6 x 1000 (1.51, 1.35 to 1.67); total detection rate, 10.7 vs 7.5 x 1000 (1.70, 1.48 to 1.94); detection rate of pT1 carcinoma, 7.0 vs 6.1 x 1000 (1.35, 1.14 to 1.59); detection rate of pT2-4 carcinoma, 3.5 vs 1.2 x 1000 (3.51, 2.75 to 4.43); false-positive rate, 4.5% vs 4.7% (0.93, 0.87 to 0.99); positive predictive value (PPV) of mammography, 19.1% vs 13.5% (1.59, 1.39 to 1.82); PPV of biopsy, 63.7% vs 60.6% (1.13, 0.98 to 1.29); detected:expected ratio, 5.02 vs 3.37 (1.49, 1.28 to 1.74). All differences were more pronounced among or restricted to women aged 50-54 years.

CONCLUSIONS

Self-referred screenees were similar to respondents to invitation in main indicators of screening feasibility such as recall rate and PPV of biopsy, while showing important increases in detection rates and detected:expected ratios, especially among women aged 50-54 years.

Mammography screening--reasons for recall and the influence of experience on recall in the Finnish system

The reasons for recall and the influence of experience on recall were studied in 579 women (3.21%) recalled from mammography screening. The proportion of recalls for further studies was the greatest (6.30%) at the onset of screening. With increased experience, the proportion decreased by stages and in the second screening round it was only 2.04%. The proportion of cases referred for surgical biopsy in the first screening round was 0.7% and the proportion of screening-detected breast cancers was 0.52%. These change little with increasing experience. In the second screening round, however, the proportion of referrals for surgical biopsy (0.43% of those screened) and of screening-detected breast cancers, (0.30%) were both low. A tumour-like density was the commonest finding resulting in recall for further studies, and the number caused by superimposition of normal parenchymal structures decreased with experience, while the proportion of breast cancers and benign tumours increased. The proportion of cases with parenchymal distortion resulting in recall for further studies was similar in both screening rounds but the number of those referred for biopsy fell dramatically in the second screening round. The number of recalls for microcalcification also fell with experience and the proportion of breast cancers in this group increased over the two screening rounds. A high ratio of malignancies in surgical biopsies can be expected when the radiologists undertaking primary screening also perform all further studies.