KSKelly Yvonne Roger Stevens

Kelly Yvonne Roger Stevens

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Kelly Yvonne Roger Stevens has a total of 17 co-authors. Here are the top co-authors by number of shared publications:

  • Benedictus C. Schoot

    Catharina Ziekenhuis

    3 shared publications
  • Saskia Houterman

    Catharina Ziekenhuis

    3 shared publications
  • Liesbet Lagaert

    Department of Obstetrics and Gynaecology, Catharina Hospital, Eindhoven, the Netherlands

    2 shared publications
  • Malou Evi Gelderblom

    Department of Obstetrics and Gynaecology, Catharina Hospital, Eindhoven, the Netherlands

    2 shared publications
  • Tanja Gijsen

    Department of Obstetrics and Gynaecology, Elkerliek Hospital, Helmond, the Netherlands

    2 shared publications
  • Tom Bakkes

    Department of Department of Electrical Engineering, Biomedical Diagnostics Lab, TU Eindhoven, Eindhoven, The Netherlands

    2 shared publications

Latest Publications

  • Open Access

    External Validation of Models to Predict Unsuccessful Endometrial Ablation: a Retrospective Study

    by Kelly Stevens, Saskia Houterman, Steven Weyers, Iris Muller, Benedictus C. Schoot

    Study Objective: External validation of our previously presented and locally established prediction models to help counsel patients for failure of endometrial ablation (EA) or surgical re-intervention within 2 years after EA, called ‘Failure model’ and ‘Re-intervention model’ respectively. Design: Retrospective external validation study, minimal follow-up time of 2 years.Setting: Two non-academic teaching hospitals in the Netherlands.Patients: Pre-menopausal women (18+) who had undergone EA for abnormal uterine bleeding problems between January 2010 and November 2012. A total of 329 patients were eligible for analysis.Interventions: Interventions used for EA were Novasure (Hologic, Marlborough, Massachusetts, US) and ThermaChoice III (Ethicon, Sommerville, US). Measurements and Main Results: The Area Under the Receiver Operating characteristics Curve (AUROC) for the outcome parameter of failure within 2 years after EA was 0.59 (95% CI 0.53 – 0.65). Variables in this model were dysmenorrhea, age, parity ≥5 and preoperative menorrhagia. The Hosmer-Lemeshow test showed no significant difference between the observed and predicted outcome. (Chi-square: 4.62, P-value: .80) The AUROC for the outcome parameter surgical re-intervention within 2 years was 0.62 (95% CI 0.53 – 0.70) Variables in this model were dysmenorrhea, age, menstrual duration> 7 days, parity ≥5 and a previous caesarean section. The Hosmer-Lemeshow test showed no significant difference between the observed and predicted outcome (Chi-square 11.34, P-value .18).Conclusion: Both the failure model and the re-intervention model can be used to predict unsuccessful endometrial ablation in the general population within two years after the procedure. It can be used prior to surgery to facilitate tailor-made shared decision-making, and help counsel patients with regards to the potential outcome of their treatment with the use of a personally calculated percentage.

  • Prediction models in gynaecology: Transparent reporting needed for clinical application

    by Malou E. Gelderblom, Kelly Y.R. Stevens, Saskia Houterman, Steven Weyers, Benedictus C. Schoot


  • Open Access

    Prediction of unsuccessful endometrial ablation: random forest vs logistic regression

    by Kelly Yvonne Roger Stevens, Liesbet Lagaert, Tom Bakkes, Malou Evi Gelderblom, Saskia Houterman, Tanja Gijsen, Benedictus C. Schoot

    Background: Five percent of pre-menopausal women experience abnormal uterine bleeding. Endometrial ablation (EA) is one of the treatment options for this common problem. However, this technique shows a decrease in patient satisfaction and treatment efficacy on the long term. Study objective: To develop a prediction model to predict surgical re-intervention (for example re-ablation or hysterectomy) within 2 years after endometrial ablation (EA) by using machine learning (ML). The performance of the developed prediction model was compared with a previously published multivariate logistic regression model (LR). Design: This retrospective cohort study, with a minimal follow-up time of 2 years, included 446 pre-menopausal women (18+) that underwent an EA for complaints of heavy menstrual bleeding. The performance of the ML and the LR model was compared using the area under the receiving operating characteristic (ROC) curve. Results: We found out that the ML model (AUC of 0.65 (95% CI 0.56–0.74)) is not superior compared to the LR model (AUC of 0.71 (95% CI 0.64–0.78)) in predicting the outcome of surgical re-intervention within 2 years after EA. Based on the ML model, dysmenorrhea and duration of menstruation have the highest impact on the surgical re-intervention rate. Conclusion: Although machine learning techniques are gaining popularity in development of clinical prediction tools, this study shows that ML is not necessarily superior to the traditional statistical LR techniques. Both techniques should be considered when developing a clinical prediction model. Both models can identify the clinical predictors to surgical re-intervention and contribute to the shared decision-making process in the clinical practice.


  • Open Access

    Prediction of Unsuccessful Endometrial Ablation: Random Forest vs Logistic Regression

    by Kelly Yvonne Roger Stevens, Liesbet Lagaert, Tom Bakkes, Malou Evi Gelderblom, Saskia Houterman, Tanja Gijsen, Benedictus Schoot

    Background: Five percent of premenopausal women experience abnormal uterine bleeding. Endometrial ablation (EA) is one of the treatment options for this common problem. However, this technique shows a decrease in patient satisfaction and treatment efficacy on the long term Study objective: To develop a prediction model to predict surgical re-intervention (for example re-ablation or hysterectomy) within two years after EA by using Machine Learning (ML). The performance of the developed prediction model was compared with a previously published multivariate logistic regression model (LR). Design: This retrospective cohort study, with a minimal follow up time of two years, included 446 pre-menopausal women (18+) that underwent an EA for complaints of heavy menstrual bleeding. The performance of the ML- and the LR model was compared using the area under the Receiving Operating Characteristic (ROC) curve. Results: We found out that the ML model (AUC of 0.65 (95% CI 0.56-0.74)) is not superior compared to the LR model (AUC of 0.71 (95% CI 0.64-0.78)) in predicting the outcome of surgical re-intervention within two years after EA. Conclusion: Although Machine Learning techniques are gaining popularity in development of clinical prediction tools, this study shows that ML is not necessarily superior to the traditional statistical LR techniques. The performance of a prediction model is influenced by the sample size, the number of features of a dataset, hyperparameter tuning and the linearity of associations. Both techniques should be considered when developing a clinical prediction model.

  • Open Access

    Prediction of unsuccessful endometrial ablation: a retrospective study

    by K. Y. R. Stevens, D. Meulenbroeks, S. Houterman, T. Gijsen, S. Weyers, B. C. Schoot

    Background: Endometrial ablation (EA) is a frequently used treatment for abnormal uterine bleeding, mainly due to the low risks, low costs and short recovery time associated with the procedure. On the short term, it seems successful, long-term follow-up however, shows decreasing patient satisfaction as well as treament efficacy. There even is a post-ablation hysterectomy rate up to 21%. Multiple factors seem to` influence the outcome of EA. Due to dissimilarities in and variety of these factors, it has not been possible so far to predict the success rate of EA based on pre-operative factors. Therefore, the aim of this study is to develop two prediction models to help counsel patients for failure of EA or necessity of surgical re-intervention within 2 years after EA. Methods: We designed a retrospective two-centred cohort study in Catharina Hospital, Eindhoven and Elkerliek Hospital, Helmond, both non-university teaching hospitals in the Netherlands. The study population consisted of 446 pre-menopausal women who underwent EA for abnormal uterine bleeding, with a minimum follow-up time of 2 years. Multivariate logistic regression analysis was used to create the prediction models. Results: The mean age of the patients was 43.8 years (range 20–55), 97.3% had complaints of menorrhagia, 57.4% of dysmenorrhoea and 61.0% had complaints of intermittent or irregular bleeding. 18.8% of patients still needed a hysterectomy after EA. The risk of re-intervention was significantly greater in women with menstrual duration > 7 days or a previous caesarean section, while pre-operative menorrhagia was significantly associated with success of EA. Younger age, parity ≥ 5 and dysmenorrhea were significant multivariate predictors in both models. These predictors were used to develop prediction models, which had a C-index of 0.71 and 0.68 respectively. Conclusion: We propose two multivariate models to predict the chance of failure and surgical re-intervention within 2 years after EA. Due to the permanent character of EA, the increasing number of post-operative failure and re-interventions, these prediction models could be useful for both the doctor and patient and may contribute to the shared decision-making.