I could ask an experimental astrophysicist but I don't know myself. As a guess, though, it would be something like this: measure the amount of energy received from the quasar over a certain amount of time; separately, measure the distance from the source (probably involving a great deal of uncertainty as quasars are some distance away, and relying on some sort of parallax measurement); and then one can scale the energy we received from the quasar based on that distance to estimate the total energy output from the quasar. That estimate will probably depend greatly on certain assumptions about the way quasars emit energy, eg no preferred direction, so may not be totally accurate, but is probably a reasonable order of magnitude estimate.
So my guess is that it will have to rely on distance calculations in some way.