Mobile healthcare is an evolving area that typically employs cloud-connected wireless biomedical gadgets for diagnosis, monitoring, and prognosis. In such environment, data compression, transmission, and processing are key issues. We propose a new method for the automated diagnosis of arrhythmia in an efficient and effective manner. The goal is to reach an efficient method by achieving a real-time data size reduction, computationally efficient signal preconditioning and a lower latency accurate classification. The Electrocardiogram (ECG) signal is sampled by using the level-crossing concept. The segmentation and adaptive-rate denoising with lower tap filters are realized. The denoised segments, without any handcrafted features extraction, are classified with one dimensional (1-D) deep Convolutional Neural Network (CNN). Five clinically important classes of arrhythmias, collected from the Beth Israel Hospital (MIT-BIH) dataset, are used to examine its applicability. Our experimental results show a 4.2-times reduction in the count of acquired samples on average compared to fix-rate counterparts. Similarly, data dimension reduction results in a more than 7.2 fold computational improvement of the post denoising stage over the conventional counterparts. Moreover, classification latency is also significantly reduced while still achieving an accuracy rate of 99%.