Enhancing the Understanding of Abdominal Trauma During the COVID-19 Pandemic Through Co-Occurrence Analysis and Machine Learning

Abstract

Background: This study examines the impact of the COVID-19 pandemic on abdominal trauma management by comparing pre-pandemic (17 February 2018–26 February 2020) and pandemic periods (27 February 2020–7 March 2022). Methods: Analyzing data from 118 patients at the Emergency County Clinical Hospital of Craiova, we identified significant shifts in clinical practices affecting patient outcomes. Results: During the pandemic, a moderate increase in surgical interventions for specific abdominal traumas indicated the effective adaptation of the medical system. Prioritizing critical cases and deferring non-urgent procedures optimized limited resources. Demographic and clinical factors—including age, sex, body mass index (BMI), and red cell distribution width (RDW)—significantly influenced the hospitalization duration and recovery outcomes. Gender disparities in mortality lessened during the pandemic, possibly due to standardized interventions and the physiological effects of SARS-CoV-2. The link between occupation and obesity highlighted how work environments impact trauma severity, especially as lifestyle changes affect BMI. While age remained a major predictor of mortality, its influence slightly decreased, potentially due to improved protocols for elderly patients. RDW emerged as an important prognostic marker for disease severity and mortality risk. Conclusions: Employing advanced co-occurrence analysis enhanced with machine learning, we uncovered complex relationships between clinical and demographic variables often overlooked by traditional methods. This innovative approach provided deeper insights into the collective impact of various factors on patient outcomes. Our findings demonstrate the healthcare system’s rapid adaptations during the pandemic and offer critical insights for optimizing medical strategies and developing personalized interventions in global crises.

Keywords

• abdominal trauma
• COVID-19 pandemic
• co-occurrence
• machine learning