Improving Patient Selection for Surgical Treatment of Bone Metastases

Abstract

To improve patient selection for surgery of bone metastases, one must first have an accurate understanding of the incidence and consequences of surgery before weighing adverse event risks in comparison to benefits. Therefore, this thesis aims at improving patient selection for surgical treatment of bone metastases by evaluating quality of life outcomes, identifying, and predicting adverse events with the help of AI tools using patient and tumor characteristics, and discussing the challenges associated with these AI tools. Part I explores the incidence and outcomes of patients with bone metastases undergoing surgery by using a large, national database representative of the United States. Part II studies the quality-of-life benefits of surgical treatment, which are considered the most important outcomes in this vulnerable patient population. Part III identifies postoperative adverse events, including mortality, complications, blood transfusions, prolonged hospital stays, and reoperations. These adverse events may substantially undermine the benefits of surgery and have significant impact on patient reported outcomes, primarily quality-of-life benefits. Part IV presents AI tools that predict these adverse events and might aid in the decision-making process of choosing the optimal candidate for surgical intervention. Part V concludes with a portrayal of the challenges of using AI tools in orthopaedic surgical care based on three reviews.