Radiologists fare well overall in online physician reviews posted by patients at RateMDs.com, although the reviews betray evidence of the “halo effect”—the doc can either do no wrong or gets almost nothing right—according to a study running in the May edition of the American Journal of Roentgenology.

Luke Ginocchio, MD, and Andrew Rosenkrantz, MD, of New York University and Richard Duszak Jr., MD, of Emory further found that patients’ perceptions of their radiologists tend to be affected by their experience with the related facility and its staff.

RateMDs.com uses a 5-star system to let patients post grades in four categories—staff, punctuality, knowledge and helpfulness—and invites free-text comments.

Focusing on the 297 U.S. cities that have a population of at least 100,000, the researchers analyzed 1,891 patient reviews of 1,259 radiologists.

The team drew additional information on reviewed radiologists from Medicare files and assessed usage of common words in the free-text comments.

Across all four categories, the most common score was 5 stars for excellent (62.7 percent to 74.3 percent).

Evidencing the halo effect, the second most common score was 1 star for terrible (13.5 percent to 20.4 percent).

Fair to good-but-not-great scores—i.e., 2 to 4 stars—were far less frequent (1.9 percent to 11.6 percent), and scores for all four categories highly correlated with one another (r = 0.781–0.951).

The team additionally found that radiologists in the Northeast scored significantly lower (p < 0.001) than those elsewhere for both staff and punctuality.

The most common words in free-text comments for positive reviews were “caring,” “knowledgeable” and “professional.”

The negative reviews were most commonly colored with the words “rude,” “pain” and “unprofessional.”

Among the limitations in study design the authors acknowledge are its questionable generalizability to rural markets and its use of unsolicited reviews, which may self-select patients who are highly satisfied, deeply dissatisfied or motivated to complete a review for unknown reasons (e.g., personal affection or animus).

They note their selection of RateMDs over similar sites, such as HealthGrades, HealthcareReviews and Yelp, for several reasons, not least its straightforward design, inclusion in earlier studies and high daily traffic.

Ginocchio et al. conclude that their observations “indicate the importance for radiologists to recognize both facility- and radiologist-related factors in affecting patients’ impressions and online public ratings of radiologists.”

Accordingly, radiologists “should be mindful of the influence of such factors on patient perceptions and strive to promote excellent service at the entire practice level, rather than solely at an image interpretation level, to ensure a positive patient experience during all phases of each patient’s imaging encounter.”