ASR has progressed a long way since its beginnings and is nowadays integrated in many commercial applications. Recently, it has started to be used by many International Organizations for the transcription of meetings. There are certain elements of spontaneous speech that can contribute to a poor performance in ASR and one of them are disfluencies. Since to our knowledge there are no studies on the impact of DFs on ASR transcriptions of international conferences, we decided to analyse the impact and the phenomenon of DFs in ASR and, particularly, its impact on the WIPO S2T tool. For this, we carried out an automatic and a manual evaluation. We established that WIPO S2T provides good quality transcripts, although there is a decrease in the quality of the transcripts from 2019 to 2021; that the impact of DFs is not as high as expected and that native speakers have a lower DF rate than non-native ones.