I was taught it as "threat and error management", but it's probably the same thing. As I understood it, it's sort of a framework for identifying connections between human performance and operational safety in aviation. It consists of threats, errors (things that are assumed to be present on every flight) and the progress from those to "undesired aircraft states". The idea is basically that threats can come from a number of places (environmental, organizational, etc); errors can be a number of human-related mistakes (procedural, communications, actual flying); and failure to identify and address them at some level can lead to an undesirable aircraft state (navigation errors, mechanical issues, weather situations, etc).
The company I worked for started teaching this in recurrent, but it can be applied systemwide as well as on a flight-by-flight basis. Our training department started really looking at our procedures and determining where the sequence we did things posed threats, and would change the FOM accordingly. I thought they did a pretty good job of it, honestly.