Objective To investigate the value of Nomogram prediction model on pregnancy outcomes with ovarian insufficiency(POI) treated by acupuncture based on the least absolute shrinkage and selection operator (LASSO)-Cox model.Methods A total of 137 POI patients who received acupuncture and moxibustion in Hubei Third People′s Hospital from January 2019 to January 2021were gathered. Relevant data were collected, and the combined LASSO regression and Cox regression methods were carried out toscreen the main influencing factors of acupuncture on the pregnancy outcome of POI patients, R software (version R3.6.3) was carriedout to construct a Nomogram prediction model for predicting the pregnancy outcome of POI patients with acupuncture and moxibustion,and receiver operating characteristic (ROC) curve and calibration curve were carried out to evaluate the predictive performance of themodel.Results There were significant differences between the pregnant group and the non-pregnant group in terms of menstrual conditions, basal FSH level, ratio of FSH to LH, basal AFC, acupuncture times, needle retention time, and needle tools (P<0.05). The multivariate Cox regression model was constructed, and the results showed that menstrual conditions, basal FSH level, ratio of FSH to LH,basal AFC, and the number of acupuncture were the independent influencing factors of acupuncture on the pregnancy outcome of POIpatients. The ROC curve showed that the areas under the curve of the 6-month pregnancy probability and the 12-month pregnancy probability were 0.78 [95%CI: (0.69, 0.88) and 0.74 [95%CI: (0.63, 0.85)], respectively. The calibration curve showed that the calibrationcurve of the prediction model was very close to the standard curve.Conclusion The Nomogram prediction model based on 5 factors screened by LASSO-Cox model, including menstrual condition, basal FSH level, ratio of FSH to LH, basal AFC, and acupuncturetimes, can individually predict the pregnancy outcome of POI patients after acupuncture. |