Our paper aimed to highlight the marked increase in referrals seen locally within recent years. We accept that, as Dr Shaw notes, some people, particularly those from the student population who also have families of origin outside Oxfordshire, may have been referred directly to gender identity clinics and so were not captured by the data. This would mean that our findings likely represent an underestimate of the total increase in numbers of people seeking assistance from specialist clinics, which makes the increase we did find even more striking.

We also appreciate Dr Shaw's points regarding the referral pathway in Oxfordshire and the need to minimise the barriers transgender people experience to accessing services. The clinical pathway has indeed been reviewed and altered since the period described in the study, with service users now being referred directly by general practitioners, generally to the specialist clinic in Northamptonshire.

The clinic did not assess people under the age of 17, so we did not include data on this age group in our study, but it is certainly notable, as both Dr Shaw and Dr Clyde's letters highlight, that referrals to gender identity disorder services (GIDS) for children and adolescents have risen dramatically over the past five years. Dr Clyde in her letter draws attention to the high rates of referral to GIDS for children and adolescents since 2010, and in particular the increase in rates of referral of those assigned female at birth. This has also been our experience, although the increase was less marked in our adult population than in the child and adolescent population Dr Clyde describes, and it is interesting to consider possible reasons for this. In our data, we found a marked trend towards more natal females being referred over the six-year period from 2011 to 2016, with a ratio of approximately 1:2 compared with natal males requesting transition. In our previous audit published in 2011 and covering the period 2006–2011, this ratio was 1:3 (Saunders and Bass, 2011). However this apparent trend did not reach statistical significance in our data. We also identified 8.5% of people presenting with non-binary gender identities, which were not noted in our previous audit. The finding of increased rates of autism spectrum disorder is also of considerable interest, and our finding of 7.8% is almost certainly an underestimate. We agree that clinics are being overwhelmed, and that there is an urgent need for both more research and discussion in this rapidly changing field, in order to best meet the needs of transgender young people and adults.