132 Chapter 9 reconstructions, registered in the Dutch Breast Implant Registry (DBIR) between 2017 and 2021. Implant loss occurred in 8.1%. Even though the observed implant loss rate increased with the number of risk factors, the observed probability and the predicted probability of implant loss did not match. Only obesity and smoking were significantly related to implant loss among the four risk factors in the risk model (obesity, smoking, nipple-preserving procedure, and DTI reconstruction). In conclusion, the multicenter risk model could not be validated with nationwide data from the DBIR. Knowing the predicted risk of implant loss during the preoperative workup remains valuable. This can guide the treating physician in planning the mastectomy and type of reconstruction, which will improve counseling women who are considering implant-based breast reconstruction. Therefore, the study in chapter 7 aimed to create a validated risk prediction model for implant loss after breast reconstruction using perioperative risk factors. Patients who had undergone either a twostage or DTI breast reconstruction were identified from the DBIR. The cohort was divided into a training (80%) and a validation cohort (20%). A risk prediction model for implant loss was created in the training cohort with multivariate logistic regression, which was subsequently validated in the validation cohort. Risk factors included smoking, BMI, pre-pectoral placement and previous radiotherapy. The model predicted an increasing probability of implant loss from 4.5% without any risk factors to 38% with four risk factors present. Due to the model’s successful validation, it may be used in general practice and is a useful tool for preoperative counseling in women who are considering implant-based breast reconstruction.
RkJQdWJsaXNoZXIy MTk4NDMw