Stroke-associated pneumonia risk estimation in patients admitted to a comprehensive stroke unit at different phases after stroke

Abstract

stroke is one of the leading causes of death and acquired disability among adults worldwide. Various complications frequently occur in patients with stroke and can significantly increase the risk for poor outcomes. Stroke-associated pneumonia, with the rate of up to 15%, is considered the most common serious medical complication in stroke patients. The aim of the study was to determine the independent predictors of the stroke-associated pneumonia in patients admitted to a comprehensive stroke unit at different time from stroke onset, as well as to develop a prognostic model for stroke-associated pneumonia risk estimation. Medical complications, including stroke-associated pneumonia, were documented in patients with clinical diagnosis of intraparenchymal cerebral stroke at admission and during their stay on the comprehensive stroke unit in 2010 to 2018. The workup at admission included assessment using the National Institutes of Health Stroke Scale and the modified Rankin scale. The diagnosis of stroke-associated pneumonia was made by specialists, in accordance with Centers for Disease Control recommendations. Qualitative variables are displayed as numbers and percentages, quantitative variables are displayed using the median (interquartile range). The method of logistic regression models was used to quantitatively assess the impact of characteristics on the PSP risk. The analysis was carried out using the package MedCalc v. 19.1. The study enrolled 539 patients, 309 (57.3%) men and 230 (42.7%) women. Median age of the patients was 66.0 years (57.9-75.1). Among the participants 101 (18.7%) were diagnosed with intracerebral hemorrhage, and 438 (81.3%) were diagnosed with ischemic stroke. Median initial National Institutes of Health Stroke Scale total score was 11 (6–18), median baseline modified Rankin scale score was 4 (3–5). Majority (64.1%) individuals were admitted within the first 30 days from onset, while 194 (35.9%) patients presented later. History of medical complications before the admission was present in 192 (35.6%) of the study participants, while during our in-hospital stay they occurred in 103 (19.1%) of the patients. The rate of new medical complications at the unit was significantly lower (p<0.001) than prior to the admission. The diagnosis of stroke-associated pneumonia before the admission was established in 20 (3.7%), and during our in-patient treatment was made in 7 (1.3%) subjects (p<0.001). Independent predictors of the stroke-associated pneumonia development were older age (odds ratio 1.11; 95% confidence interval 1.02-1.21, on average, for each additional year, p=0.02) and history of stroke-associated pneumonia at admission (odds ratio 12.2; 95 % confidence interval 2.0-72.4, p=0.006). Excellent sensitivity (85.7%), good specificity (75.0%) and area under the operating characteristic curve of 0.86 (95% confidence interval 0.83 – 0.89) for the developed prognostic model based on these of two variables, indicate their strong association with the risk of stroke-associated pneumonia during a comprehensive stroke unit stay. Therefore, stroke patients frequently develop medical complications at various stroke phases. The risk of stroke-associated pneumonia is higher in older patients and, especially, in those with stroke-associated pneumonia history. The lower rate of medical complications, in particular stroke-associated pneumonia, observed during in-hospital stay compared to the period prior to the admission may indicate the benefits of comprehensive stroke unit.