Surgical mammography reporting in a limited resource environment.

Background: Lack of resources and dedicated breast radiology has forced professionals other than radiologists to read mammography. We report a series of surgeon-read mammography in a developing country.

Methods: Data were collected prospectively form January 2003 to June 2008 for all mammography performed at a tertiary hospital. The data acquired were demographics, previous treatment for breast pathology, and indication for and outcome of mammography, including diagnostic procedures and their results. The results were benchmarked against standard performance indicators.

Results: Of 8,743 mammograms performed, 1,468 were for palpable lumps; the cancer diagnosis rate was 640 per 1,000 investigations (for 885 of these reported as compatible with malignancy, the cancer diagnosis rate was 95%; for 183 reported as indeterminate 60%, for 400 reported as benign/no abnormality 4%). Of 4,998 cancer follow-up mammograms, the cancer diagnosis rate was 15 per 1,000. For 76 of these reported as compatible with malignancy, the cancer diagnosis rate was 67%, for 161 reported as indeterminate, the biopsy rate was 36%, and cancer diagnosis rates of biopsy was 41%. Of 75 cancers diagnosed, 26 were locally recurrent after breast conservation and 49 new contralateral primaries, 84% of these were stage 0-I. Of 2,001 mammograms performed for screening, the biopsy rate was 4.5% and the cancer diagnosis rate of biopsy was 27%; the overall cancer diagnosis rate was 11 per 1,000 examinations; 96% of these were stage 0-1.

Conclusions: Dedicated breast surgeons are equally proficient at mammography interpretation as radiologists. In resource-restricted environments, nonavailability of dedicated breast radiologists should not preclude development of breast units.