For sensor networks deployed in unattended, harsh environments, the knowledge of sensor statuses such as liveness, node density and residue energy, is critical for maintaining the normal operation of the network. In this paper, we propose a poller-pollee based architecture to monitor the sensor status, focusing on two important issues: false alarm and energy efficiency. To reduce the false alarm rate, each pollee is monitored by multiple pollers. However, this approach will increase the power and bandwidth consumption. To address this issue, we propose a novel solution where the monitored sensor sends the status reports to different pollers in a round robin manner. In this way, bandwidth and power can be saved, and the false alarm rate can be reduced. We further propose a simple randomized algorithm to select the optimal number of pollers to stochastically minimize the expected total energy consumption due to monitoring. Theoretical analyses and simulations are used to demonstrate the effectiveness of the proposed techniques.