To avoid wrong decisions, people often look for as much and reliable information as possible for decision making. This sometimes results in cognitive overload due to complex information, which can lead to adverse effects such as decision paralysis or serious errors in judgment. But in some cases, confusion can also have positive consequences for decisions by increasing attention and cognitive effort. A visible signal of confusion is eyebrow contraction - and we can detect this automatically using our facial expression analysis software. Currently, we have started to empirically investigate how valid this signal is for detecting confusion. In a next step, we try to find out what effects confusion has on performance and quality in decision making and how confusion detection could be used for decision support.