No association between breast pain and breast cancer: a prospective cohort study of 10 830 symptomatic women presenting to a breast cancer diagnostic clinic

Abstract

Background Women with breast pain constitute >20% of breast clinic attendees. Aim To investigate breast cancer incidence in women presenting with breast pain and establish the health economics of referring women with breast pain to secondary care. Design and setting A prospective cohort study of all consecutive women referred to a breast diagnostic clinic over 12 months. Method Women were categorised by presentation into four distinct clinical groups and cancer incidence investigated. Results Of 10 830 women, 1972 (18%) were referred with breast pain, 6708 (62%) with lumps, 480 (4%) with nipple symptoms, 1670 (15%) with ‘other’ symptoms. Mammography, performed in 1112 women with breast pain, identified cancer in eight (0.7%). Of the 1972 women with breast pain, breast cancer incidence was 0.4% compared with ~5% in each of the three other clinical groups. Using ‘breast lump’ as reference, the odds ratio (OR) of women referred with breast pain having breast cancer was 0.05 (95% confidence interval = 0.02 to 0.09, P<0.001). Compared with reassurance in primary care, referral was more costly (net cost £262) without additional health benefits (net quality-adjusted life-year [QALY] loss –0.012). The greatest impact on the incremental cost-effectiveness ratio (ICER) was when QALY loss because of referral-associated anxiety was excluded. Primary care reassurance no longer dominated, but the ICER remained greater (£45 528/QALY) than typical UK National Health Service cost-effectiveness thresholds. Conclusion This study shows that referring women with breast pain to a breast diagnostic clinic is an inefficient use of limited resources. Alternative management pathways could improve capacity and reduce financial burden.