Development and validation of a new nomogram for self-reported OA based on machine learning: a cross-sectional study.

Area of Science:

• Rheumatology and Public Health

Background:

• Osteoarthritis (OA) diagnosis and intervention are critical for patient prognosis.
• Early identification of at-risk populations is essential for timely management.

Purpose of the Study:

• To develop and validate a diagnostic prediction model for osteoarthritis (OA).
• To identify key predictors for OA development using a large population dataset.

Main Methods:

• Utilized data from the National Health and Nutrition Examination Survey (NHANES) 2011-2016.
• Employed LASSO, XGBoost, and Random Forest (RF) algorithms to identify significant indicators.
• Developed and evaluated an OA prediction nomogram using AUC, calibration, and decision curve analysis (DCA).

Main Results:

• Identified 5 key predictors of OA: age, gender, hypertension, BMI, and caffeine intake.
• The developed OA nomogram demonstrated strong predictive performance (AUCs of 0.804 and 0.814 in training and validation sets).
• The nomogram showed good consistency, stability, and high net benefit in diagnostic accuracy.

Conclusions:

• The 5-variable OA nomogram is an accurate and convenient tool for clinicians.
• This model facilitates the identification of individuals susceptible to OA for prompt diagnosis and intervention.