Back

A Novel Easy-To-Access Model That Predicts Stroke-Associated Pneumonia

Chung, C.-Y.; Wang, W.-H.; Yin, C.-H.; Chen, J.-S.; Chen, Y.-S.; Yen, C.-C.; Lin, C.-H.

2023-07-16 infectious diseases
10.1101/2023.07.13.23292637 medRxiv
Show abstract

BACKGROUNDStroke-associated pneumonia (SAP) is a common poststroke complication but the influence of early neurological deterioration (END) on SAP risk remains unclear. We aimed to develop an easy-to-access model to predict SAP and evaluate the SAP-END relationship using the Glasgow Coma Scale (GCS). METHODSThis retrospective study enrolled consecutive patients aged [&ge;]20 years with first-ever acute ischemic stroke at Kaohsiung Veterans General Hospital between January 1, 2010, and November 30, 2020. SAP was defined according to modified Centers for Disease Control and Prevention criteria. Patients baseline characteristics, laboratory data within 24 h, neurological findings, and serial GCS scores within 48 h were collected. Regression analysis was used to identify independent risk factors for SAP. RESULTSAmong 1009 enrolled patients, SAP occurred in 85 (8.4%) patients. Using multivariate analysis, END after admission (adjusted odds ratio [aOR] 2.94, 95% confidence interval [CI] 1.74-4.95, P<0.001) remained significant after adjusting for confounders. Initial GCS score <10 (aOR 2.30, 95% CI 1.30-4.06, P=0.004), National Institutes of Health Stroke Scale (NIHSS) score 5-15 (aOR 2.02, 95% CI 1.09-3.73, P=0.026) or [&ge;]16 (aOR 3.45, 95% CI 1.72-6.89, P<0.001), cardioembolism (aOR 4.76, 95% CI 1.90-11.91, P=0.014), undetermined etiology (aOR 3.41, 95% CI 1.29-9.03, P=0.001), and neutrophil-to- lymphocyte ratio (NLR) >2.5 (aOR 2.10, 95% CI 1.28-3.46, P=0.004) were also significant. The area under the curve (AUC) of combined GCS score, END, NIHSS score, cardioembolism, stroke with undetermined type, and NLR was 0.83 (95% CI 0.78-0.87, P<0.001), which was superior to that of PANTHERIS scores (0.62, 95% CI 0.55-0.68, P<0.001). CONCLUSIONSThis study developed a simple predictive model for SAP using easily accessible and generally available parameters. GCS-based END was an independent risk factor for SAP. Early identification of SAP risk factors and reversible END causes may lower SAP incidence.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.