Page 74 - The diagnostic work-up of women with postmenopausal bleeding

Chapter 4
Introduction
Postmenopausal bleeding (PMB) is a common complaint in postmenopausal women and in about 10% of women endometrial cancer is the underlying cause of PMB.1
In the 1990s endometrial thickness measurement with transvaginal ultrasonography (TVS) was introduced as a test to distinguish between women with a low and a high risk of endometrial cancer.2The cut-off point for a thin endometrium, and thus a low risk of endometrial cancer, varies in different guidelines between three and five millimetres.3 Patients with a thin endometrium can be reassured as their post-test probability of endometrial cancer is lower than one per cent, which is a worldwide-accepted threshold for patient reassurance.1,4,5 The post-test probability depends not only on the endometrial thickness, but also on the pre-test probability, which depends on patient’s characteristics. In women with PMB, characteristics that define the pre-test probability of endometrial cancer are: age, time since menopause, body mass index (BMI), hypertension, diabetes mellitus, anticoagulants use and parity.6-12
Several studies have described the prevalence of these characteristics and developed different prediction models to estimate the individual chance of having endometrial cancer13. However, none of the existing prediction models have yet been externally validated, which is necessary for successful implementation14. All models were internally validated in their development database and two models showed the best performance.15,16 Opolskiene et al concluded that their model excludes endometrial cancer reasonably well when power Doppler is added, but because Doppler is not commonly used in daily practice, we decided to validate twomultivariablemodelswithoutDoppler,describedbyOpmeeretal15,16Thesetwo models had been internally validated in their development database.13
The aim of the present study was to externally validate the diagnostic performance of these two models and estimate the clinical consequences of the three management strategies suggested in this article by retrospectively applying the models on two independently prospectively collected databases of women with PMB.
Methods
The multivariable models and management strategies
Opmeer et al developed two multivariable logistic regression models for the
72