Page 15 - The diagnostic work-up of women with postmenopausal bleeding

but also on their patient characteristics.Age appears to be an important risk factor, 1 but also other individual patient characteristics are associated with a higher risk of
endometrial cancer, including obesity, time since menopause, hypertension, diabetes
mellitus and nulliparity.2,17-22 On the basis of existing research into the prevalence
of these risk factors, prediction models to estimate the individual chance of having endometrial cancer have been developed. In this thesis, we systematically review the literature to map the different prediction models available on this subject.Additionally, we study their performance in internal validation to identify the model with the best performance to pinpoint women with a high risk of having endometrial cancer.
3. Is a prediction model based on patient characteristics useful in daily practice to differentiate between women with a high or a low risk of endometrial cancer?
To answer this question, we will externally validate a mathematical model based on patient characteristics with or without the combination of the measurement of the endometrial thickness. Such validation is necessary before a prediction model can be implemented into clinical practice.23 The development of a prediction model can be divided into three phases: model development, internal and external model validation, and impact analysis. In internally validated models, the performance of the model is tested in the same data set in which the model was developed, or in a group of subsequent patients within the same centre. In external validation, the goal is to demonstrate generalizability and reproducibility in patients different from the patients used for derivation of the original model. Therefore, the prediction model is evaluated on new data collected from an appropriate patient population in a different centre.24 To answer the above question, we externally validate a model based on patient characteristics showing good performance in internal validation. We will validate this model in two separate databases with women with PMB: one Dutch database and one Swedish database. External validation of this model is the first step towards implementing the model in clinical practice.
4. Is the diagnostic work-up for and the removal of benign endometrial polyps effective in women with PMB to reduce recurrent bleeding?
Although hysteroscopic polypectomy is one of the most frequently performed interventions in daily gynaecologic practice, only sparse evidence is available on
General introduction
13