Abstract
Background
Patient-reported outcome measures (PROMs) are an important tool for evaluating outcomes following breast device procedures, and are used by breast device registries. PROMs can assist with device monitoring through benchmarked outcomes, but need to account for demographic and clinical factors which may affect PROMs responses.
Objectives
This study aimed to develop appropriate risk-adjustment models for the benchmarking of PROMs data to accurately track device outcomes and identify outliers in an equitable manner.
Methods
Data for this study were obtained from the Australian Breast Device Registry, which consists of a large prospective cohort of patients with primary breast implants. The five-question BREAST-Q implant surveillance module was used to assess PROMs at one-year following implant insertion. Logistic regression models were used to evaluate associations between demographic and clinical characteristics and PROMs separately by implant indication. Final multivariate risk-adjustment models were built sequentially assessing the independent significant association of these variables.
Results
2,221 reconstructive and 12,045 aesthetic primary breast implants with complete one-year follow-up PROMs were included in the study. Indication for operation (post-cancer, risk-reduction, developmental deformity) was included in the final model for all reconstructive implant PROMs. Site type (private or public hospital) was included in the final breast reconstruction model for look, rippling and tightness. Age at operation was included in the reconstruction models for rippling and tightness and in the aesthetic models for look, rippling, pain and tightness.
Conclusions
These multivariate models will be useful for equitable benchmarking of breast devices by PROMs to help track device performance.