Multiple routing metrics have been proposed for maximizing throughput or guaranteeing reliable delivery of the data. Some of the most common used metrics, neglect the bursty behavior of the channel. A new metric improve this by taking burstiness into account. This creates a burst estimation of all links and allocates the maximum number of required slots in worst-case. Even though this is a solution for the worst-case scenario, this pessimistic approach cannot achieve the performance for applications with low deadline applications. In order to close the gap between these two metrics we propose a new routing metric called Joint LATency (JLAT). We use a better understanding of the bursts with respect to specific application deadlines via joint probability distributions of the link latencies. We show that the proposed metric achieves the expected trade-off for maximal throughput and burst consideration in order to deliver a routing decision depending on the application deadline requirement. We support this claim via simulations in an exemplary scenario.