Integrating artificial intelligence (AI) and machine learning technology in healthcare systems has revolutionized predictive analytics, enabling more accurate diagnoses, treatment recommendations, and patient care. However, increasing reliance on these technologies also requires an examination of their accuracy and the potential risks associated with safe deployment. This paper explores Edge AI (i.e., Edge Intelligence) applications for healthcare cyber-physical systems (CPS), including drug ratings and reviews, etc. Advances in technological development, such as wearable devices, the Internet of Things (IoT), fifth-generation networks, the advancement of distributed computing, machine learning, and others, have significantly enhanced healthcare CPS. These advancements, part of the Healthcare 4.0 trend, aim to deliver precision medicine. Specifically, this paper proposes an Edge AI architecture for healthcare CPS and conducts an evaluation using publicly available real-world healthcare data to validate our approach. Our evaluation results reveal that combining a TF-IDF vectorizer with a passive-aggressive classifier yields the best results while achieving a training accuracy of 99.8% and maintaining high performance with a testing accuracy of 93.8%. The precision rates are 99.7% during training and 91.3% during testing. Moreover, the recall rates of 99.7% during the training phase and 91.3% during the testing phase illustrate the model’s efficacy in accurately identifying true positive cases within the evaluation metrics. Additionally, the F1-scores, balancing precision and recall, stand at 99.7% during training and 92.1% during testing, further confirming the robustness and reliability of the model.