Performance of an early quantitative swallowing assessment-based decision tree for predicting stroke-associated pneumonia

doi:10.3969/j.issn.1004-583X.2026.02.002

Abstract

Abstract:

Objective To identify risk factors for stroke-associated pneumonia (SAP) among stroke (cerebrovascular accident, CVA) patients with swallowing dysfunction using early quantitative swallowing assessment indices, and to build a decision-tree model for predicting SAP. Methods From January 2022 to December 2023, 80 CVA patients with swallowing dysfunction treated at our hospital were enrolled and divided into an SAP group (n=29) and a non-SAP group (n=51). Clinical and laboratory data were collected. Univariate analysis and multivariable logistic regression were used to identify independent predictors of SAP. A decision-tree model was constructed using the CHAID (chi-square automatic interaction detection) algorithm. Model performance was evaluated by receiver operating characteristic (ROC) curves (AUC), sensitivity, and specificity; the DeLong test compared AUCs between models. Results Significant between-group differences were observed in Standardized Swallowing Assessment (SSA) scores, presence of an abnormal swallowing-evoked cough reflex, WST (Watanabe Water Swallow Test) grading, presence of an indwelling gastric tube, and neutrophil-to-lymphocyte ratio (NLR) (all P<0.05). Multivariable logistic regression identified the following independent risk factors for SAP: higher SSA score (OR=1.555; 95%CI:1.190-2.033; P=0.001), abnormal swallowing-evoked cough reflex (OR=10.036; 95%CI: 1.889-53.306; P=0.007), WST grade V (OR=7.499; 95%CI: 1.124-50.030; P=0.037), presence of an indwelling gastric tube (OR=5.814; 95%CI: 1.117-30.263; P=0.036), and elevated NLR (OR=3.031; 95%CI: 1.378-6.669; P=0.006). Using the CHAID algorithm, the decision-tree models elected five explanatory variables: SSA score, NLR, presence of an indwelling gastric tube, abnormal swallowing-evoked cough reflex, and WST grade. The tree comprised four levels with six terminal nodes, and SSA score emerged as the strongest predictor. The decision-tree model achieved an AUC of 0.925(95%CI: 0.844-0.972), with sensitivity 79.30% and specificity 92.20% for predicting SAP in CVA patients with dysphagia. By comparison, the multivariable logistic regression modelyielded an AUC of 0.847(95%CI:0.749-0.918), sensitivity 69.00%, and specificity 86.30%. The difference in AUCs between the two models was statistically significant (DeLong test: Z=2.022, P=0.043). Conclusion SSA score, NLR, presence of an indwelling gastric tube, abnormal swallowing-evoked cough reflex, and WST grade are independent predictors of SAP in CVA patients with swallowing dysfunction. The CHAID decision-tree model based on early quantitative swallowing assessment demonstrates strong predictive performance and may be useful for early identification and prevention of SAP.

Key words: stroke, swallowing dysfunction, stroke-associated pneumonia, decision tree method, predictive modeling

CLC Number:

R743
R563.1

Xia Qi, Luo Ping, Liu Mingfeng, Lin Jiamin, Zhang Weichao. Performance of an early quantitative swallowing assessment-based decision tree for predicting stroke-associated pneumonia[J]. Clinical Focus, 2026, 41(2): 108-115.

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.lchc.cn/EN/10.3969/j.issn.1004-583X.2026.02.002

https://www.lchc.cn/EN/Y2026/V41/I2/108

Figures/Tables 7

References 27

[1]	Dogariu OA, Dogariu I, Vasile CM, et al. Diagnosis and treatment of watershed strokes:AA narrative review[J]. J Med Life, 2023, 16(6):842-850.doi: 10.25122/jml-2023-0127.
[2]	Anderson CS. Progress-defining risk factors for stroke prevention[J]. Cerebrovasc Dis, 2021, 50(6):615-616. doi: 10.1159/000516996.
[3]	D'Netto P, Rumbach A, Dunn K, et al. Clinical predictors of dysphagia recovery after stroke: A systematic review[J]. Dysphagia, 2023, 38(1):1-22.doi: 10.1007/s00455-022-10443-3.
[4]	Eltringham SA, Bray BD, Smith CJ, et al. Are differences in dysphagia assessment, oral care provision, or nasogastric tube insertion associated with stroke-associated pneumonia? A nationwide survey linked to national stroke registry data[J]. Cerebrovasc Dis, 2022, 51(3):365-372.doi: 10.1159/000519903.
[5]	Schumann-Werner B, Becker J, Nikoubashman O, et al. The relationship between neurogenic dysphagia, stroke-associated pneumonia and functional outcome in a cohort of ischemic stroke patients treated with mechanical thrombectomy[J]. J Neurol, 2023, 270(12):5958-5965.doi: 10.1007/s00415-023-11940-7. pmid: 37632565
[6]	Jannini TB, Ruggiero M, Viganò A, et al. The role of the Sapienza GLObal Bedside Evaluation of Swallowing after Stroke (GLOBE-3S) in the prevention of stroke-associated pneumonia (SAP)[J]. Neurol Sci, 2022, 43(2):1167-1176.doi: 10.1007/s10072-021-05449-y.
[7]	Wang J, Chang E, Jiang Y. Effects of vitamin C stimulation on rehabilitation of dysphagia after stroke: Arandomized trial[J]. Eur J Phys Rehabil Med, 2022, 58(4):558-564.doi: 10.23736/S1973-9087.22.07337-3.
[8]	Flayer CH, Perner C, Sokol CL. A decision tree model for neuroimmune guidance of allergic immunity[J]. Immunol Cell Biol, 2021, 99(9):936-948.doi: 10.1111/imcb.12486. pmid: 34115905
[9]	全国第四届脑血管病学术会议. 各类脑血管疾病诊断要点[J]. 中华神经科杂志, 1996, 29(6):379-381.
[10]	中国卒中学会急救医学分会, 中华医学会急诊医学分会卒中学组, 中国老年医学学会急诊医学分会, 等. 卒中相关性肺炎诊治中国专家共识(2019更新版)[J]. 中国急救医学, 2019, 39(12):1135-1143.
[11]	Alemseged F, Rocco A, Arba F, et al. Basilar Artery Treatment and Management (BATMAN) collaboration investigators. Posterior national institutes of health stroke scale improves prognostic accuracy in posterior circulation stroke[J]. Stroke, 2022, 53(4):1247-1255.doi: 10.1161/STROKEAHA.120.034019.
[12]	Tai J, Hu R, Fan S, et al. Theta-burst transcranial magnetic stimulation for dysphagia patients during recovery stage of stroke:AA randomized controlled trial[J]. Eur J Phys Rehabil Med, 2023, 59(5):543-553.doi: 10.23736/S1973-9087.23.08023-1.
[13]	Labeit B, Michou E, Trapl-Grundschober M, et al. Dysphagia after stroke: Research advances in treatment interventions[J]. Lancet Neurol, 2024, 23(4):418-428.doi: 10.1016/S1474-4422(24)00053-X. pmid: 38508837
[14]	赵林娟, 顾永梅, 葛春霞, 等. 脑卒中吞咽障碍患者相关性肺炎危险因素分析及其预防策略[J]. 中华保健医学杂志, 2023, 25(1):97-99.
[15]	王甜梦, 曾泓辑, 李彩霞, 等. 脑卒中后吞咽障碍患者吸入性肺炎的影响因素分析及预测模型构建[J]. 中华物理医学与康复杂志, 2024, 46(7):618-623.
[16]	孔玉明, 李治璋, 岳蕴华. 急性缺血性脑卒中静脉溶栓患者并发卒中相关性肺炎的危险因素分析[J]. 中风与神经疾病杂志, 2021, 38(8):693-695. doi:10.19845/j.cnki.zfysjjbzz.2021.0184.
[17]	Patel UK, Kodumuri N, Dave M, et al. Stroke-associated pneumonia: A retrospective study of risk factors and outcomes[J]. Neurologist, 2020, 25(3):39-48.doi: 10.1097/NRL.0000000000000269. pmid: 32358460
[18]	么美康, 尹丽丽. 老年卒中相关性肺炎患者多重耐药菌感染危险因素研究进展[J]. 临床荟萃, 2023, 38(1):84-87. doi: 10.3969/j.issn.1004-583X.2023.01.014
[19]	高新梅, 邢雷, 王清涛, 等. 脑卒中相关性肺炎患者多药耐药菌感染影响因素分析及预测模型构建[J]. 检验医学与临床, 2025, 22(4):475-484.
[20]	于凯娜, 高耐芬, 范素芳, 等. 早期肺康复干预治疗老年2型糖尿病合并卒中相关性肺炎患者的疗效观察[J]. 中华保健医学杂志, 2024, 26(6):765-768.
[21]	段霞, 侯景明, 洪健, 等. 脑卒中后吞咽障碍患者发生相关性肺炎的危险因素及预测模型分析[J]. 中国康复医学杂志, 2022, 37(5):616-622.
[22]	李晨红, 姜晨黎, 王金慧, 等. 脑出血患者微创颅内血肿清除术后肺部感染的影响因素分析及预测模型构建[J]. 中国卒中杂志, 2024, 19(5):532-538.
[23]	任向利, 任向杰, 白玉, 等. 卒中相关性肺炎临床特点及危险因素分析[J]. 解放军医药杂志, 2021, 33(1):44-48.
[24]	涂秀, 张克昌, 吴战. 胃肠疾病手术患者术后呼吸道感染的病原菌分布药敏性及影响因素分析[J]. 河北医学, 2023, 29(12):2077-2083.
[25]	李园, 马婷, 杨媛. 慢性支气管炎患者合并肺部感染后肺功能与免疫反应的变化及对预后的影响[J]. 中国医刊, 2025, 60(2):144-148.
[26]	陈左然, 张晔, 王铁军, 等. 探讨NPAR、NLR对老年缺血性脑卒中相关性肺炎患者的预测价值[J]. 脑与神经疾病杂志, 2023, 31(11):703-708.
[27]	Wang K, Liu Q, Mo S, et al. A decision tree model to help treatment decision-making for severe spontaneous intracerebral hemorrhage[J]. Int J Surg, 2024, 110(2):788-798.doi: 10.1097/JS9.0000000000000852.

项目	SAP组(n=29)	非SAP组(n=51)	t/χ²值	P值
年龄(岁)	65.33±4.51	64.82±4.30	0.501	0.618
性别[例(%)]
男	19(65.5)	37(72.5)	0.435	0.509
女	10(34.5)	14(27.5)	0.435	0.509
病程(d)	42.62±12.87	38.77±11.65	1.368	0.175
BMI(kg/m²)	23.58±2.12	23.10±2.06	0.991	0.325
饮酒史[例(%)]
是	11(37.9)	20(39.2)	0.013	0.91
否	18(62.1)	31(60.8)	0.013	0.91
吸烟史[例(%)]
是	9(31.0)	15(29.4)	0.023	0.879
否	20(69.0)	36(70.6)	0.023	0.879
糖尿病[例(%)]
是	7(24.1)	13(25.5)	0.018	0.893
否	22(75.9)	38(74.5)	0.018	0.893
高血脂[例(%)]
是	9(31.0)	18(35.3)	0.15	0.699
否	20(69.0)	33(64.7)	0.15	0.699
高血压[例(%)]
是	24(82.8)	30(58.8)	0.216	0.642
否	5(17.2)	11(21.6)	0.216	0.642
NIHSS评分(分)	16.23±2.89	15.02±2.74	1.862	0.066
SSA评分(分)	28.34±3.25	24.65±2.98	5.18	0
渗漏[例(%)]
异常	14(48.3)	20(39.2)	0.621	0.431
正常	15(51.7)	31(60.8)	0.621	0.431
食物残留[例(%)]
异常	17(58.6)	25(49.0)	0.683	0.408
正常	12(41.4)	26(51.0)	0.683	0.408
吞咽咳嗽反射[例(%)]
异常	15(51.7)	10(19.6)	8.876	0.003
正常	14(48.3)	41(80.4)	8.876	0.003
WST分级[例(%)]
Ⅲ~Ⅳ级	17(58.6)	45(88.2)	9.298	0.002
Ⅴ级	12(41.4)	6(11.8)	9.298	0.002
留置胃管[例(%)]
是	20(69.0)	17(33.3)	9.442	0.002
否	9(31.0)	34(66.7)	9.442	0.002
MLR	0.38±0.10	0.34±0.09	1.835	0.07
NLR	3.96±1.04	2.98±0.90	4.423	0
PLR	180.83±35.29	169.54±33.87	1.412	0.162

项目	SAP组(n=29)	非SAP组(n=51)	t/χ²值	P值
年龄(岁)	65.33±4.51	64.82±4.30	0.501	0.618
性别[例(%)]
男	19(65.5)	37(72.5)	0.435	0.509
女	10(34.5)	14(27.5)	0.435	0.509
病程(d)	42.62±12.87	38.77±11.65	1.368	0.175
BMI(kg/m²)	23.58±2.12	23.10±2.06	0.991	0.325
饮酒史[例(%)]
是	11(37.9)	20(39.2)	0.013	0.91
否	18(62.1)	31(60.8)	0.013	0.91
吸烟史[例(%)]
是	9(31.0)	15(29.4)	0.023	0.879
否	20(69.0)	36(70.6)	0.023	0.879
糖尿病[例(%)]
是	7(24.1)	13(25.5)	0.018	0.893
否	22(75.9)	38(74.5)	0.018	0.893
高血脂[例(%)]
是	9(31.0)	18(35.3)	0.15	0.699
否	20(69.0)	33(64.7)	0.15	0.699
高血压[例(%)]
是	24(82.8)	30(58.8)	0.216	0.642
否	5(17.2)	11(21.6)	0.216	0.642
NIHSS评分(分)	16.23±2.89	15.02±2.74	1.862	0.066
SSA评分(分)	28.34±3.25	24.65±2.98	5.18	0
渗漏[例(%)]
异常	14(48.3)	20(39.2)	0.621	0.431
正常	15(51.7)	31(60.8)	0.621	0.431
食物残留[例(%)]
异常	17(58.6)	25(49.0)	0.683	0.408
正常	12(41.4)	26(51.0)	0.683	0.408
吞咽咳嗽反射[例(%)]
异常	15(51.7)	10(19.6)	8.876	0.003
正常	14(48.3)	41(80.4)	8.876	0.003
WST分级[例(%)]
Ⅲ~Ⅳ级	17(58.6)	45(88.2)	9.298	0.002
Ⅴ级	12(41.4)	6(11.8)	9.298	0.002
留置胃管[例(%)]
是	20(69.0)	17(33.3)	9.442	0.002
否	9(31.0)	34(66.7)	9.442	0.002
MLR	0.38±0.10	0.34±0.09	1.835	0.07
NLR	3.96±1.04	2.98±0.90	4.423	0
PLR	180.83±35.29	169.54±33.87	1.412	0.162

变量	赋值方式
SAP的发生情况	非SAP组=“0”;SAP组=“1”
SSA评分	实测值
吞咽咳嗽反射	正常=“0”;异常=“1”
WST分级	Ⅲ-Ⅳ级=“0”;Ⅴ级=“1”
留置胃管	否=“0”;是=“1”
NLR	实测值

变量	赋值方式
SAP的发生情况	非SAP组=“0”;SAP组=“1”
SSA评分	实测值
吞咽咳嗽反射	正常=“0”;异常=“1”
WST分级	Ⅲ-Ⅳ级=“0”;Ⅴ级=“1”
留置胃管	否=“0”;是=“1”
NLR	实测值

因素	回归系数	标准误	Waldχ²值	P值	OR值	95%CI
因素	回归系数	标准误	Waldχ²值	P值	OR值	下限	上限
SSA评分	0.442	0.137	10.426	0.001	1.555	1.190	2.033
吞咽咳嗽反射	2.306	0.852	7.326	0.007	10.036	1.889	53.306
WST分级Ⅴ级	2.015	0.968	4.330	0.037	7.499	1.124	50.030
留置胃管	1.760	0.842	4.375	0.036	5.814	1.117	30.263
NLR	1.109	0.402	7.597	0.006	3.031	1.378	6.669
常量	-18.168	4.527	16.103	0.000	0.000