COVID-19 has accelerated the challenges of cardiovascular complications among patients, leading to the increased urgency of safe, trustworthy, and smart healthcare information management systems. Electrocardiogram (ECG) records are very crucial in detecting COVID-related cardiac abnormalities, but due to their sensitivity and distributed use between mobile and cloud environments, they are vulnerable to privacy and integrity breaches and unauthorized access. In order to overcome these difficulties, the current study offers the triple-tier secure mobile–cloud architecture that incorporates AES-DES hybrid cryptosystem, multi-hash integrity verification based on using the SHA-3 algorithm, and blockchain-supported immutability to protect the COVID-related cardiovascular ECG data. The framework uses morphological signal reconstruction and fractal dimension-directed adaptive tokenization to increase the ECG image representation and reduce cardio-pulmonary signal noise. Non-stationary cardiac dynamics indicative of myocardial stress due to COVID-19 with wavelet-entropy and recurrence quantification analysis Rare feature learning detects non-stationary cardiac dynamics linked to COVID-19-induced myocardial stress. A hierarchical priors Bayesian risk assessment model is introduced to offer probabilistic cardiovascular risk forecasting and attempt to address patient variability and uncertainty. The opportunistic edge offloading with the help of cloudlet decreases the processing latency and power usage, thus allowing an analysis of real-time in resource-starved mobile resources. Vast experimental assessment with a publicly available ECG data shows that the suggested system could attain the prediction equal of 97.4% with greater confidentiality, tamper resistance, and attack identification in comparison with traditional mobile-cloud healthcare systems. The findings validate that the suggested framework provides a privacy-constrained, scalable, and clinically valid system of safety management and smart analytics of COVID-related cardiovascular data.