Health services are vital in people's lives, and it's important to evaluate their quality from the patient's perspective as they are the primary beneficiaries. However, due to the extensive range of medical services, this leads to data scattering during data collection because patients are unable to access all services. This paper presents a study that aims to predict the quality of missing services and offer optimal recommendations to patients. The study involved four Iraqi hospitals, the methods of collecting the samples differed, some were manual via paper, and others were electronic, via the website and social media. To obtain the most precise results, the study employed k-means clustering and hybrid collaborative filtering. The accuracy (Acc) achieved was 97.5%, and the F1-measure (F1) was 97.64%. The proposed model was compared to traditional methods, demonstrating its success in predicting missing services and providing suitable recommendations for the patient’s needs.