Abstract Background Influential studies conclude that each hour until antibiotics increases mortality in sepsis. However, these analyses often (1) adjusted for limited covariates, (2) included patients with long delays until antibiotics, (3) combined sepsis and septic shock, and (4) used linear models presuming each hour delay has equal impact. We evaluated the effect of these analytic choices on associations between time-to-antibiotics and mortality. Methods We retrospectively identified 104 248 adults admitted to 5 hospitals from 2015–2022 with suspected infection (blood culture collection and intravenous antibiotics ≤24 h of arrival), including 25 990 with suspected septic shock and 23 619 with sepsis without shock. We used multivariable regression to calculate associations between time-to-antibiotics and in-hospital mortality under successively broader confounding-adjustment, shorter maximum time-to-antibiotic intervals, stratification by illness severity, and removing assumptions of linear hourly associations. Results Changing covariates, maximum time-to-antibiotics, and severity stratification altered the magnitude, direction, and significance of observed associations between time-to-antibiotics and mortality. In a fully adjusted model of patients treated ≤6 hours, each hour was associated with higher mortality for septic shock (adjusted odds ratio [aOR]: 1.07; 95% CI: 1.04–1.11) but not sepsis without shock (aOR: 1.03; .98–1.09) or suspected infection alone (aOR: .99; .94–1.05). Modeling each hour separately confirmed that every hour of delay was associated with increased mortality for septic shock, but only delays >6 hours were associated with higher mortality for sepsis without shock. Conclusions Associations between time-to-antibiotics and mortality in sepsis are highly sensitive to analytic choices. Failure to adequately address these issues can generate misleading conclusions.
Journal Article The Accuracy of Infection Diagnoses Among Patients Meeting Sepsis-3 Criteria in the Emergency Department Get access Max W Adelman, Max W Adelman Division of Infectious Diseases, Department of Medicine, Houston Methodist Hospital, Houston, Texas, USACenter for Infectious Diseases, Houston Methodist Research Institute, Houston, Texas, USADepartment of Medicine, Weill Cornell Medical College, New York, New York, USADivision of Pulmonary, Critical Care, and Sleep Medicine, Houston Methodist Hospital, Houston, Texas, USA Correspondence: M. W. Adelman, Division of Infectious Diseases, Department of Medicine, Houston Methodist Hospital, 6560 Fannin Street, Scurlock Tower, Suite 1512, Houston, TX 77030, USA (mwadelman@houstonmethodist.org). Search for other works by this author on: Oxford Academic PubMed Google Scholar Edward J Septimus, Edward J Septimus Division of Infectious Diseases, Department of Medicine, Houston Methodist Hospital, Houston, Texas, USADepartment of Population Medicine, Harvard Medical School, Boston, Massachusetts, USATexas A&M College of Medicine, Houston, Texas, USA Search for other works by this author on: Oxford Academic PubMed Google Scholar Cesar A Arias Cesar A Arias Division of Infectious Diseases, Department of Medicine, Houston Methodist Hospital, Houston, Texas, USACenter for Infectious Diseases, Houston Methodist Research Institute, Houston, Texas, USADepartment of Medicine, Weill Cornell Medical College, New York, New York, USA Search for other works by this author on: Oxford Academic PubMed Google Scholar Clinical Infectious Diseases, Volume 77, Issue 2, 15 July 2023, Page 327, https://doi.org/10.1093/cid/ciad240 Published: 24 April 2023 Article history Published: 24 April 2023 Corrected and typeset: 12 May 2023
Abstract Background Administrative claims data are commonly used for sepsis surveillance, research, and quality improvement. However, variations in diagnosis, documentation, and coding practices may confound efforts to benchmark hospital sepsis outcomes using claims data. Methods We evaluated the sensitivity of claims data for sepsis and organ dysfunction relative to clinical data from the electronic health records of 193 US hospitals. Sepsis was defined clinically using markers of presumed infection (blood cultures and antibiotic administrations) and concurrent organ dysfunction. Organ dysfunction was measured using laboratory data (acute kidney injury, thrombocytopenia, hepatic injury), vasopressor administrations (shock), or mechanical ventilation (respiratory failure). Correlations between hospitals’ sepsis incidence and mortality rates by claims (using “explicit” ICD-9-CM codes for severe sepsis or septic shock) versus clinical data were measured by the Pearson correlation coefficient (r) and relative hospital rankings using either data source were compared. All estimates were reliability-adjusted to account for random variation using hierarchical logistic regression modeling. Results The study cohort included 4.3 million adult hospitalizations in 2013 or 2014. The sensitivity of hospitals’ claims data for sepsis and organ dysfunction was low and variable: median sensitivity 30% (range 5–54%) for sepsis, 66% (range 26–84%) for acute kidney injury, 39% (range 16–60%) for thrombocytopenia, 36% (range 29–44%) for hepatic injury, and 66% (range 29–84%) for shock (Figure 1). There was only moderate correlation between claims and clinical data for hospitals’ sepsis incidence (r = 0.64) and mortality rates (r = 0.61), and relative hospital rankings for sepsis mortality differed substantially using either method (Figure 2). Of 48 (46%) hospitals, 22 ranked in the lowest sepsis mortality quartile by claims shifted to higher mortality quartiles using clinical data. Conclusion Variation in the completeness and accuracy of claims data for identifying sepsis and organ dysfunction limits their use for comparing hospital sepsis rates and outcomes. Sepsis surveillance using objective clinical data may facilitate more meaningful hospital comparisons. Disclosures All authors: No reported disclosures.
Objective. To estimate and compare the impact on healthcare costs of 3 alternative strategies for reducing bloodstream infections in the intensive care unit (ICU): methicillin-resistant Staphylococcus aureus (MRSA) nares screening and isolation, targeted decolonization (ie, screening, isolation, and decolonization of MRSA carriers or infections), and universal decolonization (ie, no screening and decolonization of all ICU patients). Design. Cost analysis using decision modeling. Methods. We developed a decision-analysis model to estimate the health care costs of targeted decolonization and universal decolonization strategies compared with a strategy of MRSA nares screening and isolation. Effectiveness estimates were derived from a recent randomized trial of the 3 strategies, and cost estimates were derived from the literature. Results. In the base case, universal decolonization was the dominant strategy and was estimated to have both lower intervention costs and lower total ICU costs than either screening and isolation or targeted decolonization. Compared with screening and isolation, universal decolonization was estimated to save $171,000 and prevent 9 additional bloodstream infections for every 1,000 ICU admissions. The dominance of universal decolonization persisted under a wide range of cost and effectiveness assumptions. Conclusions. A strategy of universal decolonization for patients admitted to the ICU would both reduce bloodstream infections and likely reduce healthcare costs compared with strategies of MRSA nares screening and isolation or screening and isolation coupled with targeted decolonization.
