Introduction: There are currently no published studies using patient reviews of pain physicians to quantitatively assess patient preferences for pain physician attributes. The aim of the present study was to use natural language processing to quantitatively analyze patient reviews of pain physicians by determining the effect of physician demographics and word frequency on positive review outcomes.

Methods: Using a peer-reviewed algorithm, online Healthgrades reviews of pain physicians practicing in the United States were scored according to their positive sentiment from -1 to 1. These sentiment scores and star ratings were used to compare physicians by age, gender and region of practice. Frequency analysis of words and bigrams was performed for all reviews.

Results: 15101 reviews among 1275 pain physicians showed male physicians received higher star ratings and review sentiment scores than did female physicians. Pain physicians younger than 55 years received higher star ratings and sentiment scores than those of 55 years and older. Frequency analysis revealed that words most commonly used in the more positive patient reviews included “care”, “professional”, “patient”, “help”, and “kind”; the words most commonly used in less positive reviews included “pain”, “back”, “office”, “time”, and “years”.

Conclusions: Male and/or younger pain physicians receive more positive reviews. Patients highly rate pain physicians who are perceived as personable. Patients lowly rate physicians who are perceived as providing ineffective treatment of their pain as well as when they experience barriers to their access to care.