Neighborhood socioeconomic deprivation, healthcare access, and 30-day mortality and readmission after sepsis or critical illness: findings from a nationwide study

To determine if neighborhood socioeconomic deprivation independently predicts 30-day mortality and readmission for patients with sepsis or critical illness after adjusting for individual poverty, demographics, comorbidity burden, access to healthcare, and characteristics of treating healthcare facilities. We performed a nationwide study of United States Medicare beneficiaries from 2017 to 2019. We identified hospitalized patients with severe sepsis and patients requiring prolonged mechanical ventilation, tracheostomy, or extracorporeal membrane oxygenation (ECMO) through Diagnosis Related Groups (DRGs). We estimated the association between neighborhood socioeconomic deprivation, measured by the Area Deprivation Index (ADI), and 30-day mortality and unplanned readmission using logistic regression models with restricted cubic splines. We sequentially adjusted for demographics, individual poverty, and medical comorbidities, access to healthcare services; and characteristics of treating healthcare facilities. A total of 1,526,405 admissions were included in the mortality analysis and 1,354,548 were included in the readmission analysis. After full adjustment, 30-day mortality for patients was higher for those from most-deprived neighborhoods (ADI 100) compared to least deprived neighborhoods (ADI 1) for patients with severe sepsis (OR 1.35 95% [CI 1.29-1.42]) or with prolonged mechanical ventilation with or without sepsis (OR 1.42 [95% CI 1.31, 1.54]). This association was linear and dose dependent. However, neighborhood socioeconomic deprivation was not associated with 30-day unplanned readmission for patients with severe sepsis and was inversely associated with readmission for patients requiring prolonged mechanical ventilation with or without sepsis. A strong association between neighborhood socioeconomic deprivation and 30-day mortality for critically ill patients is not explained by differences in individual poverty, demographics, measured baseline medical risk, access to healthcare resources, or characteristics of treating hospitals.