Published on
In an analysis of 1 million online Yelp reviews across 138,605 U.S. healthcare providers, researchers identified patterns and words linked to patient satisfaction by leveraging machine learning. More than 46% of the reviews were negative (1-2 stars), and 50.1% were positive (4-5 stars). The mean (SD) rating was 3.1 stars, according to the study published in JAMA Network Open. Negative feedback often cited administrative problems (such as unfair payment and poor phone interactions), and positive reviews emphasized staff attributes (such as kindness and professionalism), according to the authors. The mean positive review length was 520.8 characters, and the mean negative review length was 785.8 characters—suggesting that patients with bad experiences have more to say. In terms of practical relevance, healthcare organizations that wish to address negative reviews in a timely manner could potentially focus on online reviews that include the commonly used negative words and longer text.
Urgent care well represented: What’s interesting is that the article’s supplemental material shows that the largest proportion of reviews the authors examined came from patients who visited urgent care centers (233,266 reviews). The median star rating was a middling 3.0. The words most associated with negative reviews for urgent care were: “told,” “not,” “said,” and “rude.” While the words most associated with positive reviews were: “and,” “very,” “staff,” “friendly,” and “great.”
Read More