Provider attributes correlation analysis to their referral frequency and awards

Area of Science:

• Health Services Research
• Health Informatics
• Medical Economics

Background:

• Health provider search portals are increasingly common, using filters and patient ratings to rank physicians.
• Previous research identified intuitive attributes linked to provider quality via user surveys.

Purpose of the Study:

• To investigate the association between provider quality indicators and a comprehensive set of attributes using a data-driven approach.
• To analyze how factors like medical school, experience, patient reviews, and location correlate with provider quality.

Main Methods:

• A data-driven methodology was employed, analyzing individual providers (not organizations).
• Quality indicators included referral frequency and peer-nominated designations.
• Data sources comprised Centers for Medicare and Medicaid Services (CMS) and provider rating websites.

Main Results:

• Several attributes correlated with referral volume and peer recognition, varying significantly by locality.
• The frequency of an attribute (e.g., number of reviews) was more impactful than its value (e.g., average rating).
• Accurate classifiers for referral frequency and quality designation were developed, achieving over 85% accuracy.

Conclusions:

• A uniform provider ranking approach is insufficient; portals must calibrate rankings based on location and specialty.
• Traditional search filters may be inadequate; patients need awareness of local quality factors.
• Findings can help provider search portals better guide patients and balance patient distribution among physicians.