Rationale: Septic patients admitted to the intensive care unit (ICU) suffer from immune dysregulation, potentially leading to a secondary sepsis episode. This study aims to (i) assess the secondary sepsis rate, (ii) compare the second with the first episodes in terms of demographics, clinical and laboratory characteristics, and outcomes, and iii) evaluate the outcome of secondary sepsis.
Methods: A single-center, retrospective study (2014-2017) was conducted in a Greek ICU, including consecutive cases of adult patients admitted to the ICU for at least 48 h with a principal admission diagnosis of sepsis and stayed for at least 48 h. We searched for a secondary episode of sepsis following the primary-one. We performed survival analyses with Cox proportional hazard, Fine-Gray, and multistate models.
Results: In this study, 121 patients that fulfilled the eligibility criteria were included. The secondary sepsis group included 28 (23.1 %) patients, with episode onset, median (interquartile range), 9.5 (7.7-16.2) days after ICU admission, who had less frequently had a medical admission diagnosis, a microbiologically confirmed first episode, and the C-reactive protein was lower. The overall ICU mortality of the cohort was 44.6 %. The group that developed secondary sepsis had higher mortality, but significance was lost in Cox regression [Hazard ratio (95 % CI) 0.59(0.31-1.16)]. However, after multistate modeling adjustment, the attributable mortality was estimated at 43.9 % (95 %CI ± 14.8 %).
Conclusion: Secondary sepsis was evident in a quarter of the study participants and may be associated with an increased risk of death.
Keywords: Competing-risk; ICU mortality; Length of stay; Multistate; Secondary sepsis; Sepsis; Time-varying Cox.
Copyright © 2023 Elsevier Ltd. All rights reserved.