Do Popular Physician-Rating Websites Provide Consistent Reviews for Academic and Community Hematology-Oncology Physicians?
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: Physician-rating websites are frequently used by patients and health systems seeking information about clinicians, but the reliability of these data for hematology-oncology physicians is unclear. OBJECTIVE: To evaluate review volume, rating scores, and physician-level associations on popular physician-rating websites for academic and community hematology-oncology physicians. METHODS: We conducted a retrospective cross-sectional analysis of 210 US hematology-oncology attending physicians, including 105 academic and 105 community/private physicians. Academic physicians were sampled from 21 academic cancer-center sampling blocks across five US regions, including four blocks each from the Northeast, Southeast, Midwest, and Southwest, and five from the West. Community/private physicians were sampled from nearby non-academic hospitals or oncology practices in the same geographic areas. Fellows, residents, advanced practice providers, non-attending physicians, and pediatric-only oncologists were excluded. Physician-rating data were collected from Healthgrades, Vitals, RateMDs, and WebMD. Continuous variables were summarized using means with standard deviations and medians with interquartile ranges. Academic and community physicians were compared using Mann-Whitney U tests for skewed continuous variables and chi-square tests for categorical variables. Associations between years in practice, review volume, and ratings were assessed using Spearman correlation. A blinded second reviewer abstracted a subset of physicians to assess interrater reliability. RESULTS: Among 210 hematology-oncology physicians, the mean total number of reviews across all four websites was 40.3 ± 99.7, with a median of 14.0 reviews (interquartile range (IQR): 4.0-38.8) and a range of 0-1030. Nine physicians (4.3%) had zero reviews, 63 (30.0%) had fewer than five reviews, and 83 (39.5%) had fewer than 10 reviews. Community/private physicians had higher review volume than academic physicians, with 56.0 ± 126.8 reviews vs. 24.6 ± 58.3 reviews, and median review counts of 24.0 (IQR: 5.0-55.0) vs. 10.0 (IQR: 3.0-25.0) (p = 0.0014). Academic physicians were more likely than community/private physicians to have fewer than 10 total reviews: 52/105 (49.5%) vs. 31/105 (29.5%) (p = 0.003). Among physicians with at least one available platform rating, the mean rating score was 4.39 ± 0.78 on a 1-5 scale. Mean rating scores were 4.51 ± 0.56 for community/private physicians and 4.28 ± 0.93 for academic physicians, but rating distributions did not differ significantly by cohort (p = 0.47). No meaningful gender-based differences were identified. Longer time in practice was associated with greater total review volume (Spearman's ρ = 0.379, p < 0.001) and lower mean rating scores (ρ = -0.255, p < 0.001). Interrater agreement was near-perfect across review presence, number of reviews, rating presence, average rating scores, and narrative comment presence, with kappa values ranging from 0.94 to 0.96. CONCLUSION: Popular physician-rating websites offer limited, highly variable review volume for hematology-oncology physicians, with many clinicians receiving few or no reviews despite generally high average ratings. Community/private physicians had more reviews than academic physicians, while rating distributions were similarly high across cohorts. These results indicate that online physician ratings should be interpreted cautiously by patients, clinicians, and health systems.