Missing Data in the Unified Dysksinesia Rating Scale (UDysRS)

Area of Science:

• Neurology
• Clinical Assessment
• Data Science

Background:

• Missing data is common in Parkinson's disease (PD) rating scales.
• Incomplete data can compromise analysis validity and lead to score exclusion.
• The Unified Dyskinesia Rating Scale (UDysRS) is frequently affected by missing values.

Purpose of the Study:

• To determine the maximum number of allowable missing scores in the UDysRS.
• To establish thresholds for valid surrogate score calculation despite missing data.
• To ensure data integrity in Parkinson's disease research.

Main Methods:

• Utilized a large dataset of 3313 complete UDysRS scores.
• Sequentially removed item scores (consistently or randomly) from subjective and objective sections.
• Employed Lin's Concordance Correlation Coefficient (CCC) to compare prorated and complete scores, with CCC > 0.95 indicating validity.

Main Results:

• Allowable missing items: 3 (subjective, consistent), 5 (objective, consistent).
• Allowable missing items: 7 (subjective, random), 4 (objective, random).
• These thresholds maintain the validity of surrogate UDysRS scores.

Conclusions:

• Provides clear guidelines for handling missing data in UDysRS.
• Offers specific thresholds for retaining or rejecting scores based on missing values.
• Facilitates more robust data analysis in Parkinson's disease studies.