Risk factor analysis and nomogram for predicting poor symptom control in smoking asthmatics

Area of Science:

• Pulmonary Medicine
• Immunology
• Medical Informatics

Background:

• Smoking significantly worsens asthma control and lung function decline.
• Asthma patients who smoke face challenges in managing their condition.
• A predictive model is needed to assess prognosis and guide treatment for smoking asthmatics.

Purpose of the Study:

• To construct a nomogram prediction model for assessing asthma control in smokers.
• To identify independent risk factors associated with poorly controlled asthma in this population.
• To provide a clinical tool for early identification and management of smoking asthmatics.

Main Methods:

• Collected clinical data from 190 smoking asthmatics aged ≥14 years.
• Utilized logistic regression to identify risk factors for poor asthma control (ACQ-5 > 0.75).
• Developed and internally validated a nomogram using R4.2.2 software, evaluating efficacy with C-index, ROC, calibration, and decision curves.

Main Results:

• Identified five independent risk factors: FEV1%pred, smoking index, comorbidities, medication regimen, and adherence.
• The nomogram demonstrated high predictive ability with an AUC of 0.824.
• Model validation confirmed consistency between predicted and actual outcomes, with significant net benefit across a wide risk threshold.

Conclusions:

• FEV1%pred, smoking index, comorbidities, medication regimen, and adherence are key predictors of poor asthma control in smokers.
• The developed nomogram effectively predicts risk, aiding clinicians in early identification and tailored treatment selection.
• The tool can improve medication adherence and patient attention to complications in smoking asthmatics.