Development and validation of the nomogram of aspiration pneumonia in stroke patients with dysphagia: A retrospective observational study

Although poststroke dysphagia has been shown to be associated with aspiration pneumonia (AP), the prevalence and influencing factors of AP in stroke patients with dysphagia are still not entirely clear. This study aims to explore the prevalence and influencing factors of AP in stroke patients with dysphagia and to establish a nomogram model. This retrospective study included stroke patients with dysphagia from 2020 to 2024. Information was collected on basic data, nutritional status, swallowing function, blood routine tests, Barthel index and AP. Univariate and the multivariate analyses were used to identify the influencing factors of AP. A nomogram model was developed, and internal validation was conducted using k-fold cross-validation while external validation was conducted using the data from an additional database. Totally, 761 participants were included, with an AP prevalence of 31.41%. After adjusting for potential confounders, age ([56-65]: adjusted odds ratio [aOR] 95% confidence interval [CI]: 1.756 [1.132, 2.724]; [66-75]: aOR [95% CI]: 2.313 [1.300, 4.116]; >75: aOR [95% CI]: 2.928 [1.552, 5.525]), denture status (wearing: aOR [95% CI]: 1.429 [1.115, 2.096]), paralysis (hemiplegia: aOR [95% CI]: 1.676 [1.104, 2.544]; quadriplegia: aOR [95% CI]: 1.829 [1.184, 2.826]), nutrition support mode (intermittent oro-esophageal tube feeding: aOR [95% CI]: 0.552 [0.377, 0.807]), stroke location (brainstem: aOR [95% CI]: 1.634 [1.173, 2.276]), history of reflux esophagitis (yes: aOR [95% CI]:1.979 [1.372, 2.855]), National Institutes of Health Stroke Scale (≤5: aOR [95% CI]: 1.921 [1.226, 3.009]; [6-20]: aOR [95% CI]: 3.154 [1.902, 5.230]; >20: aOR [95% CI]: 0.856 [0.781, 0.939]), mini-mental state examination (aOR [95% CI]: 0.856 [0.781, 0.939]), malnutrition (yes: aOR [95% CI]: 1.424 [1.099, 2.399]), penetration-aspiration scale (aOR [95% CI]: 1.479 [1.344, 7.987]), neutrophil percentage (>70%: aOR [95% CI]: 1.929 [1.186, 3.3136]), white blood cell count (abnormal: aOR [95% CI]: 2.019 [1.247, 3.270]), and Barthel index (aOR [95% CI]: 0.948 [0.910, 0.988]) were the independent factors associated with AP in stroke patients with dysphagia. K-fold cross-validation revealed a mean area under curve of 0.729. The C-index for the training set and validation set were 0.778 and 0.748, respectively, and the model exhibited good stability and accuracy. The decision curve analysis revealed a net benefit when the risk threshold was between 0 and 0.85. The nomogram can help identify the risk of AP in stroke patients with dysphagia.