An autoethnographic critique of a past report of inpatient psychiatric treatment for gender diverse children

An article by Robert Kosky published in the Medical Journal of Australia in 1987 reported a case series on “inpatient therapy” administered to gender diverse children, concluding that such cases might be “correctable”.1 This work promoted the idea that “gender‐disordered children” could be “treated” to achieve cisgender‐typical behaviours and identity. In present political debates this article is being cited to promote bans on gender‐affirmative support for minors, to uphold so‐called “conversion therapy” for trans and gender diverse young people,2,3,4,5,6,7,8,9,10 and it is being quoted in current clinical guides11,12 and school or parent guidance documents.13,14,15 The report, founded on the unsupported assumption that gender diversity is inherently pathological, is also cited to promote an idea that children's gender diversity represents a parentally induced factitious disorder.8,16 This representation now extends to portraying parental facilitation of their child's social and medical transition as “child abuse”, justifying legal and procedural repercussions such as referrals to the Texas Department of Family and Protective Services17 and activation of NHS England's safeguarding procedures.18

The 1987 report describes eight consecutively treated children who came to clinical attention because they exhibited “markedly inappropriate gender behaviour, or [rejected] their biological sex”. The diagnostic formulation theorised that they were expressing cross‐gendered behaviours to meet a parent's emotional needs. The treatment rationale was that hospitalisation, sometimes described as “milieu therapy”,19 would remove the child from “the location of the primary problem” and “develop the child's social skills”.1 Inpatient stays in an Australian state‐funded psychiatric hospital for children and adolescents extended up to 6 months. No specific details are provided about the treatment program purporting to develop the targeted social skills. However, the report states that “No conscious attempt was made by the staff members to encourage masculine or feminine role behaviours”,1 a point reiterated in an accompanying editorial.20

This 1987 case series1 came to the attention of the first author of this article, Jayne McFadyen (J), after renewed international interest in the report. J, a transgender woman, had received inpatient treatment for cross‐gendered behaviour in the same hospital and during the same period as the cases included in the report. On reading the report, J strongly suspected that she was Case 5 (Supporting Information, item 1); however, some memories conflicted with the published narrative.

Concerned that hers and others’ treatment experiences were being misrepresented and deployed to limit gender‐affirming care and perpetuate conversion therapy, J initiated this research, which aimed to evaluate the 1987 publication1 against her recollection of events and the details of treatment recorded in her childhood medical records. Analysis of the patient account and medical records of desistance‐oriented “treatment” for pathologised gender non‐conformity raises questions about the reliability of the previously published account and provides a valuable long term follow‐up demonstrating inefficacy of conversion practices with transgender young people.

Methods

This study critically analyses three documents: the 1987 MJA article,1 the medical records of J's inpatient treatment, and J's autoethnographic account of her treatment. The qualitative research method of collaborative autoethnography21 was used to produce J's account of her treatment and outcomes. In a series of ten meetings between 22 March 2024 and 5 May 2024, J related her life story to one of the authors with a clinical background, who was experienced in taking case histories (AR). After each meeting, AR wrote up that section of J's story; J reviewed the draft and corrected and elaborated the narrative. After the narrative was finalised, J's archived childhood psychiatric records were requested using a freedom of information request to Western Australia's Child and Adolescent Health Service. The request was for “the full inpatient notes, including nursing records; all medical correspondence or summaries, including the lead‐up correspondence arranging the hospital admission; case‐conference records; results of any investigations; results of psychological tests, and reports of medical and nursing management decisions”. The supplied medical records, which were released on 19 June 2024, included redacted clinical correspondence; reports (social work, psychological and education department); hospital admission notes; weekly information sheets listing “problems with treatment”, and discharge notes (Supporting Information, item 2). Records of case‐conference meetings and daily hospital notes from nurses or doctors were not supplied. The reason proffered for the redactions was to protect the identities of staff and members of J's immediate family. These independently produced accounts of J's treatment were then comparatively analysed by the research team members with the account published in 1987.1 The team concurred with J's view about the high level of similarity between some specific details in the medical record, J's account and the details provided for Case 5 in the 1987 article (Supporting Information, item 3). The research team comprised psychiatric, paediatric, general practice and social research experts, LGBTQ+ as well as heterosexual and cisgender researchers, including people with lived experience of LGBTQ+ change and suppression practices and people without. Ethics approval was obtained from La Trobe University's Human Research Ethics Committee (HEC25045).

Results

Characterising the “therapy”, the 1987 article,1 in the diagnostic formulation and summary of therapeutic goals put forward, blamed childhood gender non‐conformity on parental “fantasy” and argued that “treatment” should distance such children from “the primary problem in the family interactions and the unwitting, but pathological, influences of the parents”. Therefore, children were removed from the home by hospitalisation. In J's instance, the representation that her gendered behaviours were induced by a parental symbiotic fantasy conflicts with both the medical record and autoethnographic account.

Beyond separating children from their parents, specific treatment goals are not detailed. Instead, there are ambiguous descriptions like: “develop the child's social skills so that he or she could make friends and cultivate resources outside the home” or “all children were encouraged to join in games with other children”. However, there are other claims within the article itself that contradict these assertions. For example, the statement “No conscious attempt was made by the staff members to encourage masculine or feminine role behaviours” is made immediately before noting that “Age‐appropriate behaviours were encouraged by the nursing staff members to replace the stereotyped inappropriate and isolating cross‐gender behaviours” indicating that “age‐appropriate behaviours” cannot be “cross‐gender”, but must conform to gender norms. This disambiguated conclusion is confirmed on several occasions in the medical record (Supporting Information, item 3).

The 1987 report1 did not specify the details of “therapy” administered to the child inpatients over the course of (allegedly) 10–23 weeks or 10–28 weeks. However, several comments in the article about the early responses of children and their parents to hospitalisation, and also later developments during follow‐up, confirmed the use of coercive and aversive practices. These include: overriding the developing agency and voice of the child; ongoing detention against the child's and, later, parents’ wishes (medical staff advising parents against hospital discharge); hospital discharge conditional on established gender stereotypical behaviours; and fear of rehospitalisation for “recurrence of … cross‐dressing”.

These insights align with the autoethnography, which contained additional details conflicting with the published claim that staff were not focused on encouraging gendered behaviours. J remembered staff actively discouraging her from wearing girls’ clothes; being steered away from grouping and playing with girls; being prevented from playing with dolls; and being encouraged to play boisterously and associate with boys, and play with traditionally masculine toys. J also remembered continuing to acquire and wear feminine articles of clothing, as well as an escalating response from staff. The staff response progressed from frequent checks of clothes and belongings; to random checks of underwear; to staff accompanying J to the toilet on all occasions to check underwear, and eventually insisting that J toilet with the door open to prevent her from urinating while seated, and forcing her to urinate while standing. She recalls frequently feeling embarrassed and shamed. J reported that the punishment for wearing a feminine article of clothing was lengthy locked seclusion, followed up by a formal meeting with an authoritative figure, after which she would leave the room feeling distressed. These coercive and aversive techniques explicitly directed at changing J's gender expression and suppressing her gender identity are omitted in both the provided medical records and the published report.

The 1987 article1 also claimed that “by the time of discharge from hospital the children were functioning socially and educationally at approximately age‐appropriate levels.” However, 18 days before discharge, the following was entered in J's medical record: “Seems to have no attachment to other Staff [sic] or children”. Further, there was nothing in the medical record to suggest that social and educational functioning had been formally evaluated before hospital discharge. The discrepancies between the three sources regarding aetiology, treatment and immediate health outcomes are significant and concerning.

Life course after “inpatient treatment”

The 1987 article states1 that the findings indicate “treatment of cross‐gender behaviour by means of inpatient therapy seems effective” and that “Long‐term clinical evaluations suggest that initial changes can be maintained.” The table goes so far as to describe Case 5 as “happy”.

These claims conflict with details in the autoethnographic account. J's felt sense of female gender/sex remained stable. She never felt male. Rather, her experiences — including but not limited to the inpatient “treatment” — led her to develop self‐protective strategies to conceal her gendered feelings, alongside a distrust of adults in authority. She refined coping strategies directed towards suppressing her feelings and her ongoing desire to express her feminine sense of self. J's sense of gendered discord, and associated fear and guilt, persisted into adulthood. She steadfastly maintained secrecy about these feelings and tried extinguishing them with “masculine pursuits” — predominantly alcohol‐fuelled socialising in male company. Looking back, J describes these efforts as attempts to prove to all, including herself, that she was unequivocally male. And yet both her gendered feelings and desire for self‐expression were unquenched. She continued to cultivate and occasionally wear a small collection of feminine garments, which was stored in a hidden cache.

At age 23 years, prompted by a third party's discovery of the cache, J spoke for the first time of her gendered feelings and associated experiences since she was discharged as a 10‐year‐old. Subsequently, J's sense of wellbeing improved for a time. As the years passed, however, her mood declined, which was later accompanied by recurring suicidal ideation. By 1997, at age 32 years, a nadir was reached and J was encouraged to move into shared accommodation with a friend. J's recollection in the autoethnographic account was that this encouragement came with a suggestion that the move may give her space to reflect and “work out who you are”.

Soon after moving, J began disentangling her longstanding felt sense of gender and the numerous negative associations she had carried since her inpatient experiences. It was at this stage that J began to discern their lasting effects on her sense of self, self‐expression, and the quality of her interpersonal relationships. A new phase of self‐acceptance emerged. J resolved that she had no choice but to become “who I know myself to be” — something she'd known was possible since 8 years of age. She remembered this moment of childhood self‐realisation, which occurred after reading a brief article in the daily newspaper about an adult who had hormone therapy and surgery to change from male to female. As well as comprehending that “I'm not the only one”, she recalled relief from knowing that “it can be fixed”. Twenty‐two years after her release from hospital, J, with support from her general practitioner, initiated social and medical transition in accord with procedures available at the time. While J continues to report occasional struggles with mental health, her depression is now well managed, and the improvement compared with her state before transition remains.

Now, 28 years on from when she finally accepted “who I am” and began steps towards transition, J remains comfortably aligned with her unerring feminine sense of self. Although her past feelings of gendered dissonance are a faded memory, and it is now 50 years since her experience of “inpatient treatment”, J still has traumatic childhood memories. Recognising a close portrayal of herself in the 1987 publication was distressing. We do not have access to any of the study details or approvals; however, the experience highlights several possible failures in medical ethics oversight, most notably the publication of what seemed to be a distorted representation of personal details of potentially non‐consenting individuals and families. However, the processes of preparing the autoethnographic account, obtaining and reviewing the medical records and then preparing this article have yielded new helpful insights. J's wish is that her story may protect other LGBTQA+ young people and adults from similar harms.

Discussion

Analysis of details in the 1987 publication,1 autoethnography, and childhood medical records reveal that J's months‐long “inpatient treatment” was based on an unsupported belief that the root cause of the children's gender expansive behaviours, disdain for genitalia and declarations that “they wanted to be of the opposite sex” was parental insecurity and anxiety about the child's natal sex. The attitude expressed towards parental support of gender expansive youth — real or imagined — was deeply negative. The administered “therapy” included psychologically coercive and aversive practices with the goal of gender‐typical, heterosexual outcomes. We recognise that J's experience, which occurred five decades ago, was framed by both the scientific knowledge and the social values of the time. It is notable, however, that the notion of childhood gender expansive behaviours being the equivalent of a parentally induced factitious disorder was as unfounded then as it remains today. Further, it recapitulates an equivalent periodically recycled and repeatedly discredited narrative about the aetiological role of parents for same‐sex attracted young people.22,23

Behavioural suppression and change were achieved during J's months‐long stay in the psychiatric facility. As J progressed through adolescence, an accentuating sense of gendered discord and dysphoria was calmed by snatched secretive moments of expressing femininity, accompanying feelings of guilt, and ongoing avoidance — even of the hospital itself and the surrounding area. Although the 1987 article1 concluded that J's case, among others, provided evidence that such therapy can “help” similar children, J's account indicates that all this treatment ultimately did was cause trauma and suppress her self‐acceptance for 22 years. Such discrepancies between patient reports and provider accounts of what amount to suppression practices are far from uncommon across psychology and psychiatry.24,25,26,27,28 Too often, the appeal to authority fallacy is exercised and providers’ accounts prevail. In this instance, major discrepancies between the published account and medical record have served to privilege the patient's account. Our findings suggest that when patient accounts raise questions about the accuracy of published details in the peer‐reviewed literature, they should be privileged a priori, at least until outcomes of triangulated assessments of the published and patient accounts and contemporaneous details in the medical record are known. The passage of five decades reveals also the importance of long term health record retention backed by legislated access rights.

Study limitations

Some of J's autoethnographic accounts of coercive and aversive treatment could not be confirmed by the medical records. Further, we could not definitively confirm that J was among the cases in the 1987 article. Notwithstanding, statements in both the publication and the medical records are evidence that the therapeutic intent was to discourage gender expansive behaviours with the goal of decreasing the likelihood of a “transsexual” outcome, meeting the definition of change or suppression (conversion) practices, or sexual orientation and gender identity and expression change or suppression efforts (SOGICE).29 Another potential limitation is that this account of “therapeutic” inefficacy and iatrogenic harm comes from a single autoethnographic report. As we do not know the experiences of other individuals who have undergone similar “treatment” protocols, most notably those reported in the 1987 case series,1 it is impossible to know whether J's experiences are unique.

Clinical and policy implications

Whether these experiences were unique or not, they have significant implications that must be considered by those entrusted to ensure the safety and quality of mental health services supporting gender diverse children, adolescents and their families. In keeping with the Royal Australian and New Zealand College of Psychiatrists’ recently updated position statement on the role of psychiatrists in the mental health care of gender diverse people,30 which supports outcomes research, J's experience of what some would describe as “non‐affirmative interventions”31 or “exploratory therapy”32 confirms the importance of outcome monitoring. J's story highlights that similar “treatment” of children may cause long term psychological harm requiring trauma‐informed mental health care.

The digital age has made it easier for young people to find relevant information and connect with others. The principles of gender‐affirming care are supported by most mainstream medical organisations,33,34 and gender identity change or suppression (“conversion”) practices are now illegal in many jurisdictions.35 However, some professionals and lawmakers still advocate for practices discouraging social affirmation, withholding or delaying access to long requested gender‐affirming medical treatments, and engaging trans and gender‐diverse young people in protracted “exploratory” psychotherapies without a clear endpoint, and without published evidence of efficacy or safety.36 While denying that these strategies equate to suppression or conversion practices, many of these individuals also actively oppose “conversion therapy” bans.

The 1987 publication1 should not be considered reputable evidence in any debate over transgender policies — either by direct or indirect citation. And yet, this article1 and publications of a similar age and theoretical background continue being referenced as if they were factual and reliable, by actors ranging from the Endocrine Society37 and the Cass review38 to a parent support group;39 all cite the 1987 MJA article indirectly via a review.40 Further, these false representations are now disseminating in other languages.12,41,42

Conclusion

The balance of evidence indicates that efforts to suppress or change gender identity and sexuality are ineffective, as well as actively harmful.43,44 In addition, SOGICE exercised on children, as reported in the 1987 article,1 appears to transgress several articles in the United Nations Convention on the Rights of the Child, including the associated rights and duties of parents and legal guardians.45,46 Unequivocal messaging from leaders in psychoanalysis, psychology and psychiatry is required now to enable the public to recognise misinformation, which in the instances of the 1987 publication1 and other Australian public hospital‐ and university‐sourced reports47,48,49 remain influential.50,51 It is abundantly apparent that these and other similarly detrimental52 articles in scholarly journals are fuelling the current escalation in anti‐trans rhetoric, which is now actively harming gender diverse children and their families.53

Notice

Any trans and gender‐diverse people reading this article who have endured similar experiences are encouraged to connect with trusted friends, family, and/or professionals for support if distressed. QLife (1800 184 527), Lifeline (13 11 14) and Kids Helpline (1800 551 800, for children as young as 5 years and young people up to age 25 years) are helpful supports.