Risk Prediction Model for Exacerbating Phenotype in Patients With Chronic Obstructive Pulmonary Disease

N/ARecruitingNCT06198309

Li An365 enrolled

Overview

This study is planned to be conducted based on the cohort of patients with severe chronic obstructive pulmonary disease in our hospital. Based on gut microbiota, random forest was used to search for potential diagnostic biomarkers in patients with frequent acute exacerbation and controls with non frequent acute exacerbation; Construct a frequent acute exacerbation risk prediction model using random forest, support vector machine, and BP neural network models. The development of this study will provide valuable references for the clinical classification and prognosis evaluation of chronic obstructive pulmonary disease (COPD), and improve the health level of COPD patients by further searching for treatable targets.

Study Type

OBSERVATIONAL

Enrollment

365

Conditions

Chronic Obstructive Pulmonary Disease Severe

Eligibility

Sex: ALLMin age: 40 YearsMax age: 85 Years

Medical Language ↔ Plain English

Inclusion Criteria: 1. Patients who meet the diagnostic criteria for COPD of the global initiative for chronic obstructive lung diseases (GOLD 2022) and GOLD grading Ⅲ - Ⅳ (FEV1/FVC\<70%, FEV1% predicted value ≤ 50% after Bronchiectasis) 2. Age\>40 years old 3. COPD stable for more than 4 weeks 4. Short acting Bronchiectasis was not used within 24 hours before this experiment, long acting Bronchiectasis was not used within 48 hours, and glucocorticoids were not used throughout the body in the past month 5. Patient informed and signed consent form Exclusion Criteria: 1. Asthma, active pulmonary tuberculosis, interstitial pneumonia and severe Bronchiectasis 2. Complicated with serious diseases (acute infection, diabetes, stroke, heart disease, liver and kidney dysfunction, cancer or autoimmune disease) 3. History of chronic diarrhea or constipation 4. History of Gastrointestinal Surgery 5. Using probiotics or antibiotics within the past 4 weeks 6. No history of using oral hormones or traditional Chinese medicine in the past three months 7. Pregnancy or lactation

Outcomes

Primary Outcomes

Evaluate the predictive performance of the COPD frequent seizure risk prediction model based on the area under the ROC curve.

According to the Area Under Curve (AUC) of ROC, the largest one has the best predictive performance. When AUC\>0.5, the closer it is to 1, the better the predictive performance of the model. When AUC=0.5, it indicates poor model fitting and no potential predictive value.

Time frame: A year