Huiyu Xu1,2,3,4, Guoshuang Feng5, Haiyan Wang1,2,3,4, Yong Han6, Rui Yang1,2,3,4, Ying Song1,2,3,4, Lixue Chen1,2,3,4, Li Shi1,2,3,4, Meng Qian Zhang1,2,3,4, Rong Li7,8,9,10, Jie Qiao1,2,3,4. 1. Center for Reproductive Medicine, Department of Obstetrics and Gynecology, Peking University Third Hospital, 49 Huayuan North Road, Haidian District, Beijing, 100191, People's Republic of China. 2. National Clinical Research Center for Obstetrics and Gynecology, Beijing, 100191, People's Republic of China. 3. Key Laboratory of Assisted Reproduction (Peking University), Ministry of Education, Beijing, 100191, People's Republic of China. 4. Beijing Key Laboratory of Reproductive Endocrinology and Assisted Reproductive Technology, Beijing, 100191, People's Republic of China. 5. Center for Clinical Epidemiology & Evidence-Based Medicine, Beijing Children's Hospital, National Center for Children's Health, Capital Medical University, Beijing, China. 6. Clinical Research Institute, Zhejiang Provincial People's Hospital, Hangzhou, 310014, Zhejiang Province, China. 7. Center for Reproductive Medicine, Department of Obstetrics and Gynecology, Peking University Third Hospital, 49 Huayuan North Road, Haidian District, Beijing, 100191, People's Republic of China. roseli001@sina.com. 8. National Clinical Research Center for Obstetrics and Gynecology, Beijing, 100191, People's Republic of China. roseli001@sina.com. 9. Key Laboratory of Assisted Reproduction (Peking University), Ministry of Education, Beijing, 100191, People's Republic of China. roseli001@sina.com. 10. Beijing Key Laboratory of Reproductive Endocrinology and Assisted Reproductive Technology, Beijing, 100191, People's Republic of China. roseli001@sina.com.
Abstract
PURPOSE: To establish a mathematical model for assessing the true ovarian reserve based on the predicted probability of poor ovarian response (POR). METHODS: In this retrospective cohort study, a total of 1523 GnRH-antagonist cycles in 2017 were firstly analyzed. The ovarian responses were calculated based on the number of retrieved oocytes. The continuous variables were converted into categorical variables according to cutoff values generated by the decision tree method. The optimal model was identified using forward stepwise multiple logistic regression with 5-fold cross-validation and further verified its performances using outer validation data. RESULTS: The predictors in our model were anti-Müllerian hormone (AMH), antral follicle counts (AFC), basal follicle-stimulating hormone (FSH), and age, in order of their significance, named AAFA model. The AUC, sensitivity, specificity, positive predictive value, and negative predictive value of AAFA model in inner validation and outer validation data were 0.861 and 0.850, 0.603 and 0.519, 0.917 and 0.930, 0.655 and 0.570, and 0.899 and 0.915. Ovarian reserve of 16 subgroups was further ranked according to the predicted probability of POR and further divided into 4 groups of A-D using clustering analysis. The incidence of POR in the four groups was 0.038 (0.030-0.046), 0.139 (0.101-0.177), 0.362 (0.308-0.415), and 0.571 (0.525-0.616), respectively. The order of ovarian reserve from adequate to poor followed the order of A to D. CONCLUSION: We have established an easy applicable AAFA model for assessing true ovarian reserve and may have important implications in both infertile women and general reproductive women in Chinese or Asian population.
PURPOSE: To establish a mathematical model for assessing the true ovarian reserve based on the predicted probability of poor ovarian response (POR). METHODS: In this retrospective cohort study, a total of 1523 GnRH-antagonist cycles in 2017 were firstly analyzed. The ovarian responses were calculated based on the number of retrieved oocytes. The continuous variables were converted into categorical variables according to cutoff values generated by the decision tree method. The optimal model was identified using forward stepwise multiple logistic regression with 5-fold cross-validation and further verified its performances using outer validation data. RESULTS: The predictors in our model were anti-Müllerian hormone (AMH), antral follicle counts (AFC), basal follicle-stimulating hormone (FSH), and age, in order of their significance, named AAFA model. The AUC, sensitivity, specificity, positive predictive value, and negative predictive value of AAFA model in inner validation and outer validation data were 0.861 and 0.850, 0.603 and 0.519, 0.917 and 0.930, 0.655 and 0.570, and 0.899 and 0.915. Ovarian reserve of 16 subgroups was further ranked according to the predicted probability of POR and further divided into 4 groups of A-D using clustering analysis. The incidence of POR in the four groups was 0.038 (0.030-0.046), 0.139 (0.101-0.177), 0.362 (0.308-0.415), and 0.571 (0.525-0.616), respectively. The order of ovarian reserve from adequate to poor followed the order of A to D. CONCLUSION: We have established an easy applicable AAFA model for assessing true ovarian reserve and may have important implications in both infertile women and general reproductive women in Chinese or Asian population.
Authors: Carlo Alviggi; Claus Y Andersen; Klaus Buehler; Alessandro Conforti; Giuseppe De Placido; Sandro C Esteves; Robert Fischer; Daniela Galliano; Nikolaos P Polyzos; Sesh K Sunkara; Filippo M Ubaldi; Peter Humaidan Journal: Fertil Steril Date: 2016-02-26 Impact factor: 7.329
Authors: Ilse A J van Rooij; Frank J M Broekmans; Gabrielle J Scheffer; Caspar W N Looman; J Dik F Habbema; Frank H de Jong; Bart J C M Fauser; Axel P N Themmen; Egbert R te Velde Journal: Fertil Steril Date: 2005-04 Impact factor: 7.329
Authors: Z Heidar; M Bakhtiyari; M Mirzamoradi; S Zadehmodarres; F S Sarfjoo; M A Mansournia Journal: J Endocrinol Invest Date: 2015-05-17 Impact factor: 4.256
Authors: L Kotanidis; K Nikolettos; S Petousis; B Asimakopoulos; E Chatzimitrou; G Kolios; N Nikolettos Journal: J Endocrinol Invest Date: 2016-07-27 Impact factor: 4.256
Authors: Kannamannadiar Jayaprakasan; Bruce Campbell; James Hopkisson; Ian Johnson; Nick Raine-Fenning Journal: Fertil Steril Date: 2008-11-30 Impact factor: 7.329