Response:

We appreciate Dr Oud’s interest in our study¹  and his insightful comments. The mixed linear model used in our study for sepsis trends was adjusted for age and sex. The age-sex adjusted model demonstrated increased rates of sepsis with decreased sepsis-related mortality in patients with cancer compared with patients without cancer.

The propensity score–adjusted models present similar trends of 30-day in-hospital all-cause mortality in patients with sepsis either with or without cancer. Age, sex, and comorbidity scores were the most important contributors to sepsis outcomes reported in prior studies. Therefore, these covariates were included for the mortality analyses between patients with cancer and patients without cancer (see Supplemental Table 4¹ ). We did not include organ dysfunction in the propensity score because organ dysfunction is an intermediate variable rather than a confounder. The intermediate variable is part of the causal pathway in the relationship between cancer and the...