The explosion of big medical data brings with it opportunities and challenges for the healthcare provider who is seeking actionable information and personalized advice. Classical recommendation systems that work in other fields do not fully consider the challenges posed by vast amount of medical data. To tackle the aforementioned weaknesses, we present a hybrid recommendation system by fusing content-based and trust-aware methods based on weighted sum approach. Content based filtering explores the intrinsic characteristics of medical data and finds patterns, relationships from user’s historical record to make recommendations. The trust-aware part contains user trust thresholds which increase the reliability and decrease bias or disturbing proposals. The proposed hybrid system has the lowest RMSE (1.8441) and MAE (1.6293), which is substantially better than other models in terms of prediction performance. These results indicate that combining several recommendation approaches can exploit their advantages and dodge their deficiencies. The hybrid method decreases prediction error, and the recommendations are more accurate and reliable for overcoming problems such as cold start, sparsity and complexity in large-scale medical data. In summary, the system provides personalized and reliable healthcare solutions compared to solo methods.