Reducing Excessive Continuous Pulse Oximetry and Promoting Patient Safety: An Electronic Health Record-Based Quality Improvement Effort

BACKGROUND: While continuous pulse oximetry monitoring (CPO) is ubiquitous in hospital settings, overutilization may contribute to alarm fatigue, disproportionately burdening bedside nurses and monitoring technicians. Achieving appropriate utilization remains a national patient safety goal. OBJECTIVE: We investigated whether an electronic health record (EHR)-based quality improvement (QI) effort could safely reduce inpatient CPO monitoring. INTERVENTION: This QI intervention took place from June 2022 to May 2024 across all non-ICU inpatient units of a 796-bed tertiary medical center. Following semi-structured interviews to understand CPO ordering practices, we implemented a revised CPO order, admission and discharge order sets, and new best practice advisory (BPA) to promote timely discontinuation. MAIN MEASURES: The primary outcome was change in monthly CPO hours using interrupted time series analysis comparing pre-intervention (June 2022-May 2023) and post-intervention (June 2023-May 2024) periods. Secondary outcomes included average monthly CPO alarms and CPO duration per admission, along with rate of escalation-of-care and serious patient harm events. KEY RESULTS: There were 18,351 CPO-associated hospitalizations during the pre-intervention period and 18,713 in the post-intervention period. Compared to the pre-intervention monthly average of 120,771 CPO hours, the post-intervention period was associated with an immediate reduction of 22,680 h (95% CI -33,591, -11,769; p < 0.01). Monthly CPO alarms (-332,819; 95% CI -510,870, -154,769; p < 0.01) and CPO duration per admission (-34.2 h; 95% CI -45.5, -22.9; p < 0.01) decreased during the post-implementation period. ICU transfer rates (+9.6 per 1000 admissions; 95% CI 0.87, 18.3; p = 0.03) increased without changes in rates of rapid responses, code blue events, nor mortality. CONCLUSION: An EHR-based QI initiative achieved significant, sustained reductions in CPO utilization, exemplifying how clinical decision support interventions can effectively reduce CPO utilization without increasing serious patient harm events in the form of code blue events or in-hospital mortality.