Incidence and burden of infectious disease or sepsis in hospitalized adults with specific comorbidities: a nationwide population-based surveillance

Abstract

AbstractPurpose:Although numerous epidemiological investigations have focused on sepsis, its definition has been revised by the Sepsis-3 criteria since 2016. To update the epidemiologic sepsis information on individuals who are immunocompromised due to specific comorbidities, a retrospective cohort based on the population-based databasewas studied.Methods:According to ICD-9-CM codes, the first and subsequent hospitalizations for infections or sepsis episodes were identified. Patients with targeted comorbidities were regarded as the study cohort; their corresponding control cohort was recognized through matching by sex and age.Results:In total, 231,016 adults with comorbid diabetes mellitus, 41,205 with liver cirrhosis, and 137,661 with hemato-oncology were identified; and their corresponding control patients included the control cohort I (115,461 adults), II (20,702), and III (68,847), respectively. Through the validation in our hospital, interobserver agreement in infections (k=0.81) and sepsis (k=0.88) was excellent. Compared to the corresponding control cohorts, patients with the targeted comorbidities were more likely to have experienced a first hospitalization due to infection or sepsis at a younger age; and the incidence rate ratios of infections in patients with diabetes mellitus, liver cirrhosis, or hemato-oncology were 1.92, 4.14, and 2.76; and the incidence rate ratios of sepsis were 1.83, 5.42, and 2.68, respectively.Conclusions:Compared with the general population, patients with immunocompromised status caused by specific comorbidities, particularly in cirrhotic patients, have a higher frequency of hospitalization for subsequent infections or sepsis. We believe that our findings might offer the support for improving the quality of patient.