Objective:

To construct a risk prediction nomogram model for pulmonary infection in patients with acute exacerbation of chronic obstructive pulmonary disease (AECOPD) complicated with type II respiratory failure.

Approach:

• Study Design: A retrospective analysis of clinical data from 361 patients with AECOPD and type II respiratory failure.
• Data Analysis: Used restricted cubic spline analysis, single factor and multiple factor logistic regression analyses to identify independent risk factors and construct the predictive model.
• Model Evaluation: Evaluated predictive performance using ROC curve, calibration curve, and clinical decision curve.

Key Findings:

• Age showed no significant nonlinear relationship with pulmonary infection (P for nonlinearity > 0.05).
• Length of hospital stay, APACHE II score, and duration of antibiotic use had significant nonlinear relationships with pulmonary infection (P for nonlinearity < 0.05).
• Independent risk factors identified include age, diabetes, hospital stay > 15 days, APACHE II score > 15 points, and antibiotics combination time > 14 days.
• Albumin was identified as an independent protective factor (P < 0.05).
• The nomogram model demonstrated AUCs of 0.917 and 0.884 in training and validation cohorts, respectively.

Interpretation:

The nomogram model may assist in identifying pulmonary infection risk in AECOPD patients with type II respiratory failure.

Limitations:

• Retrospective design may introduce bias.
• Single-center study limits generalizability.

Conclusion:

The constructed nomogram risk model may demonstrate predictive performance for pulmonary infection in AECOPD patients with type II respiratory failure.