Predictive factors for ovarian response in a corifollitropin alfa/GnRH antagonist protocol for controlled ovarian stimulation in IVF/ICSI cycles

Table 3 Test characteristics and equations for models for high and low ovarian response

				Predictive value
Model	Cutoff	Sensitivity	Specificity	Positive	Negative
High ovarian response^a	0.15	0.84	0.80	0.41	0.97
Low ovarian response^b	0.37	0.77	0.87	0.64	0.93
Combined high ovarian response^c	0.13	0.87	0.78	0.39	0.97
Combined low ovarian response^d	0.36	0.78	0.86	0.62	0.93

Model-based probability \( P=\frac{e^{LP}}{1+{e}^{LP}} \) where LP is the linear predictor
^aLP = 0.6953 – 0.1232 × age [years] + 0.6596 × AMH [ng/mL] + 0.1829 × AFC [count] – 0.2517 × FSH [IU/mL]
^bLP = 5.1380 + 0.0961 × age [years] – 1.6821 × AMH [ng/mL] – 0.1690 × AFC [count] – 0.2304 × CLn [days]
^cLP = −1.1213 – 0.1258 × age [years] + 0.7010 × AMH [ng/mL] + 0.1942 × AFC [count]
^dLP = −0.7701 + 0.0828 × age [years] – 1.7373 × AMH [ng/mL] – 0.1635 × AFC [count]
If the model-based predicted probability is above the cutoff, a patient would be classified as a potential high (respectively, low) responder
AMH anti-Müllerian hormone, AFC antral follicle count, CLn cycle length, FSH follicle-stimulating hormone

ISSN: 1477-7827