Establishment and validation of a nomogram prediction model for post-stroke pneumonia in the patients with large vessel occlusion undergoing mechanical thrombectomy

Yin Xiaojiao; Wang Jianxiang; Gao Yun

doi:10.3969/j.issn.1002-1949.2025.05.009

Abstract：Objective To explore the risk factors for the occurrence of post-stroke pneumonia(PSP)in the patients with acute ischemic stroke due to large vessel occlusion(AIS-LVO)who underwent mechanical thrombectomy(MT),and to construct a nomogram model to predict the risk of PSP occurrence in order to provide a reference for clinical practice.Methods A total of 330 AIS-LVO patients undergoing MT in Kunming Yan'an Hospital and the First People's Hospital of Kunming from January 1,2017 to November 30,2023 were retrospectively included as the modeling group,and 91 patients from December 2023 to November 1,2024 were prospectively included as the external validation group.Relevant data of the patients were collected,and the patients were grouped according to the occurrence of PSP.Predictive variables were screened through univariate analysis,least absolute shrinkage and selection operator(LASSO)regression and multivariate analysis to construct a nomogram model,and the model was evaluated.Results The study showed that nasogastric tube placement,dysphagia,and neutrophil-to-lymphocyte ratio(NLR)were independent predictive factors for PSP.The nomogram model had good discriminative ability[the area under curve(AUC)in the modeling group was 0.88,and the AUC in the external validation group was 0.856],strong calibration,and good clinical applicability.Compared with non-PSP patients,patients with PSP had a higher in-hospital mortality rate(2％vs.20％,P=0.015)and a lower rate of 90-day functional independence(63％vs.27％,P=0.001).Conclusions The constructed nomogram model has a high predictive value for the risk of PSP occurrence in AIS-LVO patients who underwent MT.However,due to the relatively small sample size,further prospective studies are needed for verification and updating of the model.