Abstract
BACKGROUND: Online reviews influence how patients select primary care providers. However, little is known about the language used in such reviews, regional differences in tone or content, and reviewer behavior. OBJECTIVE: This study aims to analyze Google reviews of general practices in four selected German regions - two urban (Lübeck, Dresden) and two rural (Main-Tauber District, Mecklenburg Lake Plateau) - to identify linguistic patterns, reviewer characteristics, and implications for quality improvement. METHODS: We extracted all publicly available Google Maps entries for “Hausarzt [City], Deutschland” using an Application Programming Interface (API). Natural Language Processing (NLP) was applied to analyze the content and metadata of reviews. Descriptive statistics and ordinal logistic regression identified word combinations associated with high and low ratings. Results were compared across regions. RESULTS: A total of 9307 reviews from 1854 practices were analyzed. Ratings were predominantly positive (78.7% 5-star) with no regional differences. Reviewers who wrote positive reviews tended to have written more reviews overall. Common terms in positive reviews included “attentive”, “competent,” and “takes time”; negative reviews featured “unfriendly” and “long wait”. Lexical patterns and predictors were consistent across all regions. CONCLUSION: Google reviews provide some insights into patient perceptions of primary care. Key expectations, such as friendliness and attentiveness, appear stable across urban and rural settings, with minimal regional variation in review language. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s43999-025-00080-2.