Statistical Learning to Measure the Efficacy of Digital Surgical Workflows

The crisis facing many National Health Services, arising from economic resources, often limited, impacts their ability to integrate new technologies and meet citizens' healthcare demands. To tackle this challenge, optimizing existing resources, particularly in surgical activities, which require considerable effort, is necessary to improve operational efficiency. Standardizing surgical procedures represents a significant commitment to advancing healthcare equity, minimizing errors, and optimizing resources. This study investigates the standardization of surgical procedures, focusing on laparoscopic right hemicolectomy, given that colorectal cancer is the third most commonly diagnosed cancer. Comparative analyses were conducted among conventional, robotic, and Surgical Process Institute (SPI) supported approaches. The performance of the different operative approaches in terms of surgical duration times (minutes) and lengths of hospital stay (days) was analyzed through a statistical learning approach. In particular, statistical analyses and regression modeling were conducted on 29 patients. The regression analysis showed that the intervention mode is confirmed to be the most significant predictor of intervention duration and length of stay. The results indicate that SPI-supported procedures exhibit more stability and less variability in intervention duration compared to the other two groups, reduced average surgical time durations compared to robotic surgeries, and shorter lengths of stay. Due to the size of the dataset, further data collection and analysis are needed to better evaluate the effect of standardization on surgical outcomes. Despite this, it is anticipated that greater use of standardization software, such as that offered by SPI, will increase the quality standards of interventions, allow for better resource management, and enhance patient satisfaction and safety.