Objective:

To develop and validate a practical prehabilitation-enhanced nomogram for predicting early postoperative pulmonary recovery failure within 7 days after lung tumor surgery.

Approach:

• Study Design: Multicenter retrospective study with a development cohort from one center and an external test cohort from another.
• Data Collection: Routine perioperative variables and quantified 14-day prehabilitation process indicators were collected.
• Model Comparison: A routine-variable model was compared with a prehabilitation-enhanced model using multivariable logistic regression.
• Model Evaluation: Model performance was assessed through discrimination, calibration, and clinical utility.

Key Findings:

• The addition of prehabilitation indicators improved discrimination in all cohorts.
• The area under the curve (AUC) increased significantly with the prehabilitation-enhanced model.
• The final nomogram included five predictors: intraoperative blood loss, DLCO% predicted, preoperative resting SpO2, incentive spirometry target-achieved days, and breathing training target-achieved days.
• The nomogram showed good discrimination with AUCs of 0.972, 0.863, and 0.883 in training, internal validation, and external test cohorts, respectively.
• Calibration was acceptable across cohorts with Brier scores indicating good predictive accuracy.

Interpretation:

The prehabilitation-enhanced nomogram allows for individualized early risk stratification for postoperative pulmonary recovery failure within 7 days.

Limitations:

• The study is retrospective and may have inherent biases.
• External validation was limited to one additional center.

Conclusion:

The nomogram may support targeted prevention for patients at high risk of early postoperative pulmonary recovery failure.