The weekend effect on 28-day mortality in septic patients admitted to the ICU: A retrospective study from the MIMIC-IV database

Objective: Numerous studies have shown that patients admitted on weekends or holidays have higher mortality rates and poorer prognoses than those admitted on weekdays. However, the specific impact of the weekend effect on patients with sepsis remains unclear and requires further investigation. Methods: This study included adult patients with sepsis who were first admitted to the ICU and stayed for ≥24 hours, using data from the Medical Information Mart for Intensive Care (MIMIC)-IV (version 2.2),with the data collection period from 2008 to 2019. Data on age, gender, type of ICU admission, vital signs, disease severity scores, and medications were collected, with patients categorized into weekend and weekday admission groups. The primary outcome was 28-day mortality, while secondary outcomes included 90-day mortality, hospital mortality, ICU mortality, and survival days without vasoactive drugs, ventilator, or ICU stay. COX regression analyses with propensity score matching (PSM) were employed to assess the impact of weekend admissions on the survival of septic patients in the ICU. Results: A total of 20,261 septic patients met the inclusion criteria, with 14,469 in the weekday group and 5,792 in the weekend group. The weekend admission group showed no statistically significant differences in 28-day mortality, hospital mortality, ICU mortality, survival days without vasoactive drugs, survival days without ventilator, survival days without ICU, and length of ICU stay compared to the weekday group. Subgroup analyses for 28-day mortality revealed that key baseline characteristics such as gender, age, BMI, race, ICU type, hypertension, diabetes mellitus, and SOFA score did not independently influence the prognosis of patients with sepsis based on weekend admission. Conclusion: The study found no significant weekend effect on the prognosis of septic patients admitted to the ICU, based on both univariate and multivariate analyses.