The positive predictive value of postmenopausal bleeding for uterine malignancy

Abstract

Postmenopausal bleeding (PMB) is vaginal bleeding occurring after the menopause. The differing definitions of PMB used and need for a standard definition, including what this might be, are highlighted in the thesis. Benign and malignant causes of PMB have been identified, yet an accurate estimate of the risk of uterine malignancy in those with PMB remains undetermined. This thesis aims to establish the positive predictive value (PPV) of PMB for uterine malignancy by combining available evidence. Determining this will potentially improve management of PMB, ensuring detection of malignancies while not over investigating women.

A systematic review conducted identified existing studies providing a PPV in community, primary and secondary care populations. One study was identified in the community providing a PPV for uterine and endometrial cancer of 0.51% (95% CI 0.27-0.75) and 0.47% (95% CI 0.24-0.70) respectively for women aged 45-54. One primary care study was identified providing a PPV for endometrial cancer of 1.68% (95% CI 1.43-1.93) in women aged =35. Pooling findings from 26 secondary care studies via random effects meta-analysis produced an estimated PPV of 8.4% (95% CI 6.9-9.9) for endometrial cancer and 19.6% (95% CI 13.8-25.5) for uterine cancer.

The PPV derived for a community population was tested by analysing of a cohort of women with PMB from the community. No women identified had a subsequent cancer diagnosis, suggesting consistency with the low PPV established during the review.

The PPV applicable to a community population suggests women with PMB are unlikely to have cancer, indicating a public health initiative promoting awareness of PMB to be unwarranted. The primary care PPV suggests it may not be appropriate to refer all women with PMB for investigation. However, the increased PPVs in secondary care indicate it is appropriate to investigate all those presenting to secondary care with PMB for malignancy.