BACKGROUND: Procedures performed after-hours have been associated with worse outcomes in several procedural specialties. Data examining the relationship between procedure timing and adverse event severity in interventional radiology remain limited. This study retrospectively evaluated adverse events reviewed at interventional radiology morbidity and mortality conferences to assess whether procedure timing was associated with differences in adverse event severity and quality-of-care assessment. RESULTS: A total of 547 adverse events occurring across four affiliated hospital sites were included. The cohort comprised 341 males and 206 females with a mean age of 56.2 years. Vascular procedures accounted for 369 events (67.5%), and non-vascular procedures accounted for 178 events (32.5%). Most adverse events occurred during weekday daytime hours (n = 459, 83.9%), followed by weekday after-hours (n = 50, 9.1%) and weekend periods (n = 38, 6.9%). Compared with weekday daytime procedures, a greater proportion of adverse events resulting in death occurred during weekday after-hours and on weekends, as reflected by a higher proportion of Society of Interventional Radiology class F events and Harm Score 9 outcomes (p < 0.001). In contrast, no significant difference in Quality Control System scores was observed across procedure timing categories (p = 0.73). CONCLUSIONS: Among adverse events reviewed at interventional radiology morbidity and mortality conferences, procedures performed during weekday after-hours and weekend periods were associated with greater adverse event severity, while quality-of-care assessments did not differ by procedure timing. These findings describe patterns within morbidity and mortality-reviewed complications and do not estimate complication rates or population-level risk by procedure timing. The absence of variation in Quality Control System scores supports the hypothesis that observed differences in severity may reflect patient acuity and procedural urgency rather than differences in procedural quality.
Wu et al. (Fri,) studied this question.