SAFE-CARE: Reversible Privacy-preserving Physician Feedback Framework to Improve Patient Care Quality

Current quality improvement efforts in clinical medicine face significant challenges, such as delivering timely performance data to clinicians, lacking relevant benchmarks, and infrequent reference to national guidelines during performance reviews. There is an urgent need to help physicians identify inefficiencies in their practices compared to local, regional, or national peers. However, proposed solutions often meet resistance due to privacy concerns and potential repercussions. To address these issues, we propose SAFE-CARE, a privacy-preserving feedback framework that provides timely performance data while safeguarding physician anonymity. SAFE-CARE employs a cryptography-based threshold secret sharing scheme, allowing for reversible privacy that protects against unauthorized identification while enabling de-anonymization for legitimate purposes, such as audits. This framework facilitates the collection and correlation of data across different clinical repositories within a healthcare system, comparing it to optimal care patterns to assess physician performance and resource use. We demonstrate the feasibility of SAFE-CARE through a prototype validated with synthetic data in a blood transfusion case study.