KOOS score maps were externally validated to inform knee arthroplasty shared decision making

Area of Science:

• Orthopedics
• Rheumatology
• Medical Informatics

Background:

• The Knee injury and Osteoarthritis Outcome Survey (KOOS) is frequently used to assess outcomes in knee arthroplasty (KA).
• Accurate prediction of patient-reported outcomes is crucial for shared decision-making prior to KA.

Purpose of the Study:

• To develop and externally validate score maps for KOOS Pain and Function subscales.
• To create easily understandable, single-page graphical depictions of predicted KOOS scores.

Main Methods:

• Cross-sectional study using presurgical KA data from two independent cohorts.
• Score maps were derived from Osteoarthritis Initiative (OAI) data and validated on an independent cohort.
• Agreement was assessed using Weighted Kappa (Κw) coefficients and score differences.

Main Results:

• Score maps demonstrated moderate to substantial chance-corrected agreement for KOOS Pain and Function subscales.
• KOOS Pain score map agreement (Κw) ranged from 0.43 to 0.73.
• Score maps predicted actual item scores within ±1 point at least 94% of the time.

Conclusions:

• Developed KOOS score maps provide acceptable precision when compared to actual patient scores.
• These user-friendly score maps can enhance clinical decisions and patient-surgeon discussions before knee arthroplasty.