Application of cardiopulmonary ultrasound in adult acute dyspnea: construction of differential diagnosis model between cardiogenic pulmonary edema and pneumonia
10.3760/cma.j.cn131148-20220623-00455
- VernacularTitle:心肺超声在成人急性呼吸困难中的应用：心源性肺水肿与肺炎的鉴别诊断模型构建
- Author:
Haotian ZHAO
1
;
Li LI
;
Heling ZHAO
;
Hongyuan XUE
;
Yuanlin LIU
;
Yang BAI
;
Yi LIU
;
Guangyao YAO
;
Peng ZHAO
;
Yaru YAN
Author Information
1. 河北省人民医院超声科，石家庄 050000
- Keywords:
Lung ultrasound;
Dyspnea, acute;
Cardiogenic pulmonary edema;
Pneumonia;
Diagnosis, differential
- From:
Chinese Journal of Ultrasonography
2023;32(3):242-249
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To analyze the cardio-pulmonary ultrasound features of cardiogenic pulmonary edema (CPE) and pneumonia in adults with acute dyspnea, and to construct a differential diagnosis model.Methods:Seven hundred and forty-three patients with sudden acute dyspnea admitted to Hebei General Hospital from November 2018 to May 2022 were retropectively included. Ultrasonographer A performed lung ultrasound with 12 zone method, and interpreted and recorded the ultrasonic signs (including A-lines area, B-lines area, consolidation area and pleural effusion area) together with ultrasonographer B. According to the ultrasonic characteristics of the whole lung, it was divided into A-profile and B-profile. According to the continuity and symmetry of the distribution of B-lines in bilateral lung fields, it could be divided into bilateral lung continuous and discontinuous B-profile, bilateral lung symmetric and asymmetric B-profile. Left ventricular ejection fraction (LVEF), left ventricular filling pressure (E/e′), right ventricular dilatation, tricuspid annular systolic displacement (TAPSE) and inferior vena cava diameter (IVCD) were evaluated by echocardiography, and all the indexes were transformed into binary variables. According to the final clinical diagnosis and treatment results, the disease was divided into CPE group and pneumonia group. Binary Logistic regression model was used to screen independent influencing factors, and partial regression coefficient β value was used as a weight to assign a score, and a differential diagnosis model was established based on the total score. The predictive value of the model was evaluated by the receiver operating characteristic curve (ROC) and area under curve (AUC). After the model was built, 30 patients with CPE or pneumonia were independently collected by ultrasonographer C as external validation data, which were included in the model to draw ROC curve and evaluate the differential diagnosis efficiency of the model. The consistencies between ultrasonographer A and B, A and C in observing lung ultrasound were explored.Results:A total of 743 patients from 43 clinical departments were included, including 246 cases in CPE group and 497 cases in pneumonia group. Multivariate logistic regression analysis showed that bilateral lung continuous B-profile, bilateral lung symmetric B-profile, ≥1 pleural effusion area, LVEF<50%, E/e′>14 were the risk factors for CPE (all OR>1, P<0.05), and ≥1 consolidation area and ≥1 pleural sliding disappearance area were the protective factors for CPE (all OR>1, P<0.05). The sensitivity, specificity and AUC of combined cardio-pulmonary ultrasound index β value weight score in the differential diagnosis of CPE and pneumonia were 0.939, 0.956 and 0.986, respectively. The AUC of external validation data was 0.904. Ultrasonographer A and B, A and C had good consistency in the interpretation of lung ultrasound signs ( P<0.05). Conclusions:The differential diagnosis model based on combined cardio-pulmonary ultrasound indexes has high differential diagnosis efficiency for CPE and pneumonia, and can be used in bedside cardio-pulmonary ultrasound practice.
