Statistical veto methods are commonly used to reduce the list of candidate gravitational wave (GW) events which are detected as transient (burst) signals in the main output of GW detectors. If a burst event in the GW channel is coincident with an event in a veto channel (where the veto channel does not contain any GW signal), it is possible to veto the event from the GW channel with a low 'false-veto' rate. Unfortunately, many promising veto channels are interferometer channels which can, at some level, contain traces of any detected GW signal. In this case, the application of a 'standard statistical veto' could have a high false-veto rate. We will present an extension to the standard statistical veto method that includes an 'amplitude consistency check'. This method allows the application of statistical vetoes derived from interferometer channels containing GW information with a low false-veto rate. By applying a statistical veto with an amplitude consistency check to data from the GEO 600 detector, veto efficiencies between 5 and 20%, together with a use-percentage of up to 80%, were obtained. The robustness of this veto method was also confirmed by hardware injections. The burst triggers were generated using the mHACR detection algorithm.