Abstract
BACKGROUND: Sepsis is a series of organ failures caused by dysregulated responses to infection. Risk factors for sepsis are multiple comorbidities, a poor nutrition status, and limited mobility. The primary purpose of the study was to determine whether ambulation ability with albumin and C-reactive protein are predictive of 28-day mortality of elderly patients with sepsis. METHODS: This was a retrospective observational study using a multicentre-based registry of elderly patients between November 2016 and February 2017. The inclusion criteria were a patient ≥65 years and a diagnosis of sepsis and exclusion criteria were a patient with covariates of ambulation ability such as central nervous system diseases, or malignancy. The area under the receiver operating characteristic curve of prediction models were calculated and compared. The survival rates according to the ambulation ability were estimated and compared by the log-rank test. RESULTS: 2291 patients ≥65 years visited with infectious diseases. 496 subjects with central nervous system diseases, 710 subjects with malignancy and 817 subjects with a Sequential Organ Failure Assessment score ≤ 1 were excluded. Ultimately, 278 subjects were included in the primary analysis. 133 (47.8%) subjects were male and the median age was 78 years. 228 (82%) subjects could ambulate independently before morbidity and 28 (10.1%) subjects expired in 28 days. In the inability to ambulate and C-reactive protein to albumin ratio model, the area under the curve predicting 28-day mortality was 0.761 with no significant difference from the Sequential Organ Failure Assessment score (0.859, p = 0.097) and the estimated survival rate on 28th day according to the ability to ambulate showed a significant difference (hazard ratio = 1.212, p < 0.001). CONCLUSION: The premorbid ambulation ability with albumin and C-reactive protein can be combined to predict 28-day mortality in elderly patients with sepsis.