A total of 86422 mammograms were read and 4156 patients were recalled for assessment.
Overall recall rate was 4.8%.
3253 of the recalled patients (78,4%) were coded as BI-RADS 4A or below,
while the number of patients coded as BI-RADS 4B,
4C and 5 were as follows,
321 (7,7%),
368 (8,8%) and 214 (5,1%) respectively (see Table 1).
1279 patients underwent core tissue biopsy,
and 736 (58%) of these appeared to be malignant.
Overall cancer detection rate was 0.85%.
However,
in 543 patients,
the biopsy result was reported as non-malignant,
and overall non-malignant biopsy rate was 0.62%.
608 patients,
recorded as BI-RADS 4A,
underwent core tissue biopsy,
and of those biopsies 410 resulted as non-malignant,
and non-malignant biopsy rate for BI-RADS 4A was 14,1%.
Whereas,
206 patients recorded as BI-RADS 5 underwent biopsy and only 5 of those biopsies came up as non-malignant,
and non-malignant biopsy rate for BI-RADS 5 was just 2,3%.
The non-malignant biopsy rate for BI-RADS 4B category was slightly higher than BI-RADS 4A (16,5% versus 14,1%),
however,
malignancy rate was substantially higher,
31,1% vs 6,8%.
The malignancy rate of each BI-RADS categories was consistent with BI-RADS system recommendation [1] (see Table 2 for further detail).
Fibrocystic changes (42%) and fibroadenoma (25%) were the most common benign finding.
In 32 patients,
the biopsy was reported as normal breast tissue.
There were 88 borderline lesions,
of which 74 were surgically or image guided removed,
and two lesions (one ADH,
one papilloma) were upgraded to malignancy (see Fig. 1).
Apart from core tissue biopsies,
180 of the recalled patients underwent cyst aspiration.
168 of these patients had been recorded as BI-RADS 4A and below.