Reformulating real-time random safety analysis during the SARS-CoV-2 pandemic

Introduction: From a safety perspective, the pandemic imposed atypical work dynamics that led to noticeable gaps in clinical safety across all levels of health care.

Objectives: To verify that Real-Time Random Safety Analyses (AASTRE) are feasible and useful in a high-pressure care setting.

Design: Prospective study (January-September 2022).

Setting: University Hospital with 350 beds. Two mixed ICUs (12 and 14 beds).

Interventions: Two safety audits per week were planned to determine the feasibility and usefulness of the 32 safety measures (grouped into 8 blocks).

Main variables of interest: 1) Feasibility: Proportion of completed audits compared to scheduled audits and time spent. 2) Utility: Changes in the care process made as a result of implementing AASTRE.

Results: A total of 390 patient-days were analyzed (179 were Non-COVID patients and 49 were COVID patients). In the COVID patient subgroup, age, ICU stay, SAPS 3, and ICU mortality were significantly higher compared to the Non-COVID patient subgroup. Regarding feasibility, 93.8% of planned rounds were carried out with an average audit time of 25 ± 8 min. Overall, changes in the care process were made in 11.8% of the measures analyzed.

Conclusions: In a high-complexity care environment, AASTRE proved to be a feasible and useful tool with only two interventions per week lasting less than 30 min. Overall, AASTRE allowed unsafe situations to be turned safe in more than 10% of the evaluations.