Estimating the Causal Effect of Sepsis on Mortality in ICU Patients with and Without a Solid Organ Transplant Using Survival Targeted Minimum Loss-Based Estimation

• Activity Number: 81 - Contributed Poster Presentations: Section on Statistics in Epidemiology
• Type: Contributed
• Date/Time: Monday, August 3, 2020 : 10:00 AM to 2:00 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Katherine Hoffman* and Iván Díaz
• Companies: Weill Cornell Medicine and Weill Cornell Medicine
• Keywords: non-parametric; TMLE; semi-parametric; targeted maximum likelihood estimation; sepsis; ICU

Abstract

Sepsis is a life-threatening illness caused by an overreaction of the immune response to an infection. The majority of sepsis cases are treated in the Intensive Care Unit (ICU), as sepsis is a well-known risk factor for mortality. Acquiring sepsis has been shown to be protective against mortality in patients with a solid organ transplant (SOT). These findings are surprising and clinically interesting, as SOT patients are pharmaceutically immunosuppressed to prevent donor rejection. We used targeted minimum loss-based estimation to compute the effects of sepsis on 28-day mortality in 28,431 NewYork-Presbyterian ICU patients with and without an SOT. After adjusting for confounders such as age, sex, comorbidities, type of infection, and white blood cell count, we found the 28-day mortality difference caused by sepsis to be 21% (95% confidence interval 19-23%) in patients without an SOT. The same difference in patients with an SOT was -3.0% (-7.0-1.0%) . The difference between these two groups (24% (20-29%)) was statistically significant and suggests patients with an SOT are over-diagnosed with sepsis, or are acquiring a less deadly form of sepsis than patients without an SOT.