Operating room (OR) capacity utilization plays a critical role in improving patient care and enhancing financial performance in healthcare. However, traditional OR planning approaches and limited flexibility in resource allocation often struggle to balance scheduled elective surgeries with unpredictable emergency cases, leading to inefficiencies such as overtime and undertime. We propose a design science approach comprising four key stages to balance OR utilization. First, the problem framing and conceptualization phase involved close collaboration with hospital managers and surgeons to identify OR capacity challenges and explore demand-side flexibility strategies. Through in-depth discussions, we critically assessed the limitations of existing planning methods and identified opportunities for a more adaptive and efficient approach. Second, inspired by the volatility portfolio approach and the layered spackling strategy, we introduce a volatility-stratified surgical portfolio, categorizing surgeries according to medical advice and patient preferences. Surgeries are classified into emergency (high volatility), general elective (moderate volatility) and flexible elective (low volatility). By leveraging the low volatility as a buffer to absorb fluctuations, we accommodate emergency demand uncertainty while offering theoretical and practical insights into flexibility management in operations. Third, a simulation incorporating uncertainty and sensitivity analysis was conducted using real data from two hospitals before and during COVID-19. The stage, combined with machine learning, mixed integer programming, and empirical-based algorithm, demonstrates the effectiveness of the framework in balancing OR capacity utilization. Finally, the evaluation phase combined simulation-based assessments with practitioner validation workshops to examine the feasibility and applicability of implementing the framework in healthcare and other industries.