Temporal Trends in Sepsis Incidence and Mortality in Patients With Cancer in the US Population

Background Few population-based studies assess the impact of cancer on sepsis incidence and mortality. Objectives To evaluate epidemiological trends of sepsis in patients with cancer. Methods This retrospective cohort study included adults (≥20 years old) identified using sepsis-indicator International Classification of Diseases codes from the Nationwide Inpatient Sample database (2006-2014). A generalized linear model was used to trend incidence and mortality. Outcomes in patients with cancer and patients without cancer were compared using propensity score matching. Cox regression modeling was used to calculate hazard ratios for mortality rates. Results The study included 13 996 374 patients, 13.6% of whom had cancer. Gram-positive infections were most common, but the incidence of gram-negative infections increased at a greater rate. Compared with patients without cancer, those with cancer had significantly higher rates of lower respiratory tract (35.0% vs 31.6%), intra-abdominal (5.5% vs 4.6%), fungal (4.8% vs 2.9%), and anaerobic (1.2% vs 0.9%) infections. Sepsis incidence increased at a higher rate in patients with cancer than in those without cancer, but hospital mortality rates improved equally in both groups. After propensity score matching, hospital mortality was higher in patients with cancer than in those without cancer (hazard ratio, 1.25; 95% CI, 1.24-1.26). Of patients with sepsis and cancer, those with lung cancer had the lowest survival (hazard ratio, 1.65) compared with those with breast cancer, who had the highest survival. Conclusions Cancer patients are at high risk for sepsis and associated mortality. Research is needed to guide sepsis monitoring and prevention in patients with cancer.