In this paper, we investigate the use of cognitive radio systems in medical data transmission. Therefore, based on the state-of-the-art, we propose a complete cognitive-radio communication architecture for eHealth systems. The goal is to point out significant technical challenges required for more flexible and effective transmissions in eHealth contexts. After that, we study connectivity, interference management and performance criteria in cognitive-radio based eHealth systems. For connectivity, we show that cognitive-radio technology allows exploring opportunistically frequencies to find available channels. Then, to avoid interference, we propose a strategy-based management plan to reduce disruption risks. An example of strategy, frequency hopping, is then implemented and tested. Finally, we analyse the performance of our proposal for some categories of medical data. We propose therefore a function that reinforces the cognitive radio decision-making and spectrum sharing modules by controlling some quality parameters.