Risk Factors for Adverse Outcomes in Cancer Patients With Sepsis

Abstract

Background: Cancer and its various treatment modalities increase susceptibility to the development of sepsis. Because of the complex relationship between sepsis and cancer, we aimed to study the differences in risk factors and outcomes of sepsis in patients with cancer (SCa) compared to patients without cancer (SnoCa).

Methods: A retrospective cohort analysis of all adult patients who received care for sepsis in an urban tertiary healthcare center was conducted. Risk factors and outcomes were compared between the SCa and SnoCa groups.

Results: SCa group (n = 310) was older than SnoCa group (n = 628) (66.8 vs. 61.5 years; P < 0.01). There were higher associations of certain variables in the SCa group compared to the SnoCa group, such as male sex (55.8% vs. 48.2%; P = 0.03), White race (60.6% vs. 51.7%; P = 0.01), lower body mass index (BMI) (28.10 ± 9.3 vs. 30.02 ± 10.4 kg/m²; P = 0.01), and history of transient ischemic attack (TIA) (6.1% vs. 2.7%; P = 0.01). Conversely, there were lower associations of recreational drug use (10.0% vs. 17.0%; P = 0.01) and diabetes mellitus (DM) (35.9% vs. 45.9%; P = 0.01). Simple linear regression found that the SCa group had lower length of stay (LOS) (β = –0.08; P = 0.03). Logistic regression model showed that having cancer increased odds of all-cause mortality (odds ratio (OR) 1.82, 95% confidence interval (CI) 1.35–2.46; P < 0.01); however, the SCa group had comparable readmission rates, bloodstream infection, and in-hospital mortality.

Conclusion: Compared to patients with sepsis without cancer, patients with sepsis and cancer have higher association with older age, male sex, White race, lower BMI, and TIA, and lower association with recreational drug use and DM. Patients with sepsis and cancer have lower LOS, higher all-cause mortality and have no difference in readmissions, bloodstream infections, and in-hospital mortality.