Identifying high-risk phenotypes and associated harms of delayed time-to-antibiotics in patients with ICU onset sepsis: A retrospective cohort study

PURPOSE

To identify phenotypes of Intensive Care Unit (ICU) onset sepsis and its associated harms of delayed time-to-antibiotics.

MATERIALS AND METHODS

The Medical Information Mart for Intensive Care IV (MIMIC-IV) database was employed to identify patients with ICU onset sepsis. The primary exposure was time-to-antibiotics, as measured from sepsis recognition to first antibiotic administered. Latent profile analysis (LPA) was used to identify phenotypes of sepsis based on individual organ failure score derived from Sequential Organ Failure Assessment (SOFA). Interactions between phenotypes and time-to-antibiotics on 28-day mortality were explored.

RESULTS

6246 patients were enrolled in final analysis. The overall 28-day mortality was 12.7%. Delayed time-to-antibiotics was associated with increased 28-day mortality in patients with ICU onset sepsis (HR 1.12, 95% CI 1.08-1.18). Four phenotypes of sepsis were identified: phenotype 1 was characterized by respiratory dysfunction, phenotype 2 was characterized by cardiovascular dysfunction, phenotype 3 was characterized by multiple organ dysfunction, and phenotype 4 was characterized by neurological dysfunction. The adjusted HR of 28-day mortality was 1.16 (95% CI 1.08-1.25) in phenotype 1, and 1.06 (95% CI 1.00-1.13) in phenotype 2, while no significant interaction was observed.

CONCLUSIONS

Septic patients with respiratory or cardiovascular dysfunction were associated with harms of delayed time-to-antibiotics.

Septic patients without obvious signs of infection at baseline are more likely to die in the ICU

Objective

Early identification of sepsis is mandatory. However, clinical presentation is sometimes misleading given the lack of infection signs. The objective of the study was to evaluate the impact on the 28-day mortality of the so-called “vague” presentation of sepsis.

Design

Single centre retrospective observational study.

Setting

One teaching hospital Intensive Care Unit.

Subjects

All the patients who presented at the Emergency Department (ED) and were thereafter admitted to the Intensive Care Unit (ICU) with a final diagnosis of sepsis were included in this retrospective observational three-year study. They were classified as having exhibited either “vague” or explicit presentation at the ED according to previously suggested criteria. Baseline characteristics, infection main features and sepsis management were compared. The impact of a vague presentation on 28-day mortality was then evaluated.

Interventions

None.

Measurements and main results

Among the 348 included patients, 103 (29.6%) had a vague sepsis presentation. Underlying chronic diseases were more likely in those patients [e.g., peripheral arterial occlusive disease: adjusted odd ratio (aOR) = 2.01, (1.08–3.77) 95% confidence interval (CI); p = 0.028], but organ failure was less likely at the ED [SOFA score value: 4.7 (3.2) vs. 5.2 (3.1), p = 0.09]. In contrast, 28-day mortality was higher in the vague presentation group (40.8% vs. 26.9%, p = 0.011), along with longer time-to-diagnosis [18 (31) vs. 4 (11) h, p < 0.001], time-to-antibiotics [20 (32) vs. 7 (12) h, p < 0.001] and time to ICU admission [71 (159) vs. 24 (69) h, p < 0.001]. Whatever, such a vague presentation independently predicted 28-day mortality [aOR = 2.14 (1.24–3.68) 95% CI; p = 0.006].

Conclusions

Almost one third of septic patient requiring ICU had a vague presentation at the ED. Despite an apparent lower level of severity when initially assessed, those patients had an increased risk of mortality that could not be fully explained by delayed diagnosis and management of sepsis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-022-07210-y.

Time-to-antibiotics and clinical outcomes in patients with sepsis and septic shock: a prospective nationwide multicenter cohort study

Background

Timely administration of antibiotics is one of the most important interventions in reducing mortality in sepsis. However, administering antibiotics within a strict time threshold in all patients suspected with sepsis will require huge amount of effort and resources and may increase the risk of unintentional exposure to broad-spectrum antibiotics in patients without infection with its consequences. Thus, controversy still exists on whether clinicians should target different time-to-antibiotics thresholds for patients with sepsis versus septic shock.

Methods

This study analyzed prospectively collected data from an ongoing multicenter cohort of patients with sepsis identified in the emergency department. Adjusted odds ratios (ORs) were compared for in-hospital mortality of patients who had received antibiotics within 1 h to that of those who did not. Spline regression models were used to assess the association of time-to-antibiotics as continuous variables and increasing risk of in-hospital mortality. The differences in the association between time-to-antibiotics and in-hospital mortality were assessed according to the presence of septic shock.

Results

Overall, 3035 patients were included in the analysis. Among them, 601 (19.8%) presented with septic shock, and 774 (25.5%) died. The adjusted OR for in-hospital mortality of patients whose time-to-antibiotics was within 1 h was 0.78 (95% confidence interval [CI] 0.61–0.99; p = 0.046). The adjusted OR for in-hospital mortality was 0.66 (95% CI 0.44–0.99; p = 0.049) and statistically significant in patients with septic shock, whereas it was 0.85 (95% CI 0.64–1.15; p = 0.300) in patients with sepsis but without shock. Among patients who received antibiotics within 3 h, those with septic shock showed 35% (p = 0.042) increased risk of mortality for every 1-h delay in antibiotics, but no such trend was observed in patients without shock.

Conclusion

Timely administration of antibiotics improved outcomes in patients with septic shock; however, the association between early antibiotic administration and outcome was not as clear in patients with sepsis without shock.