<p>Medical IoT emerged as a result of the rapid advancement of communication technologies and medical IoT devices. This led to the creation of several monitoring applications periodically generating massive amount of data ranging from simple to complex data. The IoT network is greatly burdened by the transmission of this enormous amount of data to the cloud.We thus outline our plan to use fog computing to make informed decisions about which and how much data to send to the cloud. We propose and implement data management system that employs accurate and lightweight classification algorithm to classify the data into normal and abnormal categories and then filters the abnormal data and forwards this in compressed form to cloud. Our experimental findings evaluate our suggested system’s performance in terms of latency, transmission time, data size reduction, and classification accuracy. As demonstrated by the findings, the suggested system performs better than traditional cloud-based medical IoT systems that disregard data processing at fog by (i) gaining 97–100% accuracy for different datasets (ii) reducing transmission time and latency by 95.48% for ECG dataset, 80–81% for different samples of Pulse Oximeter data and 72% for image dataset.</p>