ABSTRACT
The Surviving Sepsis Campaign recommends standard operating procedures for patients with sepsis. Real-world evidence about sepsis order set implementation is limited. OBJECTIVES: To estimate the effect of sepsis order set usage on hospital mortality. DESIGN: Retrospective cohort study. SETTING AND PARTICIPANTS: Fifty-four acute care hospitals in the United States from December 1, 2020 to November 30, 2022 involving 104,662 patients hospitalized for sepsis. MAIN OUTCOMES AND MEASURES: Hospital mortality. RESULTS: The sepsis order set was used in 58,091 (55.5%) patients with sepsis. Initial mean sequential organ failure assessment score was 0.3 lower in patients for whom the order set was used than in those for whom it was not used (2.9 sd [2.8] vs 3.2 [3.1], p < 0.01). In bivariate analysis, hospital mortality was 6.3% lower in patients for whom the sepsis order set was used (9.7% vs 16.0%, p < 0.01), median time from emergency department triage to antibiotics was 54 minutes less (125 interquartile range [IQR, 68-221] vs 179 [98-379], p < 0.01), and median total time hypotensive was 2.1 hours less (5.5 IQR [2.0-15.0] vs 7.6 [2.5-21.8], p < 0.01) and septic shock was 3.2% less common (22.0% vs 25.4%, p < 0.01). Order set use was associated with 1.1 fewer median days of hospitalization (4.9 [2.8-9.0] vs 6.0 [3.2-12.1], p < 0.01), and 6.6% more patients discharged to home (61.4% vs 54.8%, p < 0.01). In the multivariable model, sepsis order set use was independently associated with lower hospital mortality (odds ratio 0.70; 95% CI, 0.66-0.73). CONCLUSIONS AND RELEVANCE: In a cohort of patients hospitalized with sepsis, order set use was independently associated with lower hospital mortality. Order sets can impact large-scale quality improvement efforts.
ABSTRACT
Two HIV-1-infected children on antiretroviral therapy were enrolled into a clinical study of retroviral-mediated transfer of a gene that inhibits replication of HIV-1, targeting bone marrow CD34+ hematopoietic stem/progenitor cells. Two retroviral vectors were used, one encoding a "humanized" dominant-negative REV protein (huM10) that is a potent inhibitor of HIV-1 replication and one encoding a nontranslated marker gene (FX) to serve as an internal control for the level of gene marking. Peripheral blood mononuclear cells (PBMC) containing the huM10 gene or FX gene were detected by quantitative PCR at frequencies of approximately 1/10,000 in both subjects for the first 1-3 months following re-infusion of the gene-transduced bone marrow, but then were at or below the limits of detection (<1/1,000,000) at most times over 2 years. In one patient, a reappearance of PBMC containing the huM10 gene, but not the FX gene, occurred concomitant with a rise in the HIV-1 viral load during a period of nonadherence to the antiretroviral regimen. Unique clones of gene-marked PBMC were detected by LAM-PCR during the time of elevated HIV-1 levels. These findings indicate that there was a selective survival advantage for PBMC containing the huM10 gene during the time of increased HIV-1 load.
