To characterize the causes and potential preventability of hospital-onset sepsis events at acute care hospitals.
Approach:
Study Design: Retrospective cohort study of adults admitted to 9 Mass General Brigham acute care hospitals from January 2021 through December 2023.
Data Collection: Medical record reviews of a sample of 250 hospital-onset sepsis events to assess the likelihood and source of infection.
Preventability Assessment: Reviewers assessed 100 infection-related cases using a 6-point Likert scale to rate potential preventability.
Statistical Analysis: Fisher exact and Wilcoxon rank-sum tests were used for comparisons; two-sided P < .05 was considered statistically significant.
Key Findings:
Hospital-onset sepsis was identified in 5,212 of 558,037 admissions (0.9%).
Infections accounted for 79% of reviewed events; respiratory infections were the most frequent source.
Among reviewed events, 13% met current US criteria for reportable healthcare-associated infections (HAIs).
Reviewers classified 31% of the infection-related cases assessed for preventability as potentially preventable.
Crude in-hospital mortality among the 250 reviewed events was 37% (93 of 250).
Potential prevention opportunities identified in the review included improving management of recognized infections and preventing nonventilator hospital-acquired pneumonia (NV-HAP).
Limitations:
The number of cases assessed for preventability was limited by the resources required for adjudication.
Retrospective preventability assessments are subjective and may be affected by reviewer judgment.
Structured consensus adjudication may not eliminate cognitive bias.
Conclusion:
In this study, hospital-onset sepsis was associated with substantial in-hospital mortality. The findings suggest that existing HAI surveillance criteria may not capture many hospital-onset sepsis events and identify potential areas for prevention. Further evaluation is needed to determine how surveillance and prevention approaches could be improved.
