Concierge care and patient reviews

Area of Science:

• Health Services Research
• Computational Linguistics
• Patient Experience

Background:

• Concierge medicine is a growing model of healthcare delivery.
• Understanding patient feedback is crucial for assessing quality of care.
• Online reviews offer a rich source of patient-reported data.

Purpose of the Study:

• To analyze differences in patient review language based on concierge doctor status.
• To identify predictors of concierge status and patient numerical ratings.
• To explore the impact of specific word types on patient scores.

Main Methods:

• Utilized patient review data from Healthgrades.com for physicians in California and Florida.
• Employed machine-learning sentiment analysis to process review text.
• Correlated word usage (technical vs. interpersonal) with concierge status and numerical ratings.

Main Results:

• Reviews for concierge doctors more frequently contained healthcare-specific technical terms (e.g., 'staff,' 'office').
• Reviews for non-concierge doctors were more likely to include interpersonal, bedside-manner words (e.g., 'listen,' 'concerns').
• Interpersonal words had mixed effects on ratings, while technical terms showed a negative correlation with patient scores across all physicians.

Conclusions:

• Language used in patient reviews differs significantly between concierge and traditional physician practices.
• Technical language in reviews may indicate a negative patient experience, impacting scores.
• This study provides initial insights into sentiment analysis for evaluating quality of care in concierge medicine.