While the primary focus of this editorial is on the UK, many of the concerns arising from the consequences of COVID-19, economic circumstances, increased referrals and changes in diagnostic practices may be relevant to other countries.

Like other parts of the world, the UK is facing changes in the conceptualisation of autism. The most recent diagnostic manuals (DSM-5-TR and ICD-11) reflect a broader understanding of autism and the experiences of autistic people. In addition, there has been growing debate about whether autism should be understood as a ‘condition’ or ‘difference’ rather than a ‘disorder’, with some rejecting the medical model and challenging the necessity of formal diagnostic assessments. While self-identification may provide some benefits, there are important considerations. ^{Reference David and Deeley1}

Referral rates for autism assessments have grown dramatically in the UK in recent years. This has many likely causes, including increased awareness, reduced stigma, greater access to diagnostic services, the aforementioned broadening of the diagnostic criteria and recognition of more subtle phenotypes or presentations. The National Health Service (NHS) has struggled to manage demand and, in response, introduced the UK NHS Long Term Plan (2019) and the National Strategy for Autistic Children, Young People and Adults (2021), committing to reducing assessment waiting times and improving assessment pathways. However, no agreed models of practice were provided, leaving services to adapt in isolation. The effectiveness of these changes varied between services because of differences in funding, service structures and resources. To address this, the Autism Assessment Framework and Operational Guidance was published in 2023. The overall success of these policies remains unclear.

Despite these innovations, demand for autism assessments continues to increase. For example, in September 2023, there were 161 827 young people and adults with an open referral for autism assessments through NHS services. By June 2024, this had risen to 193 203 and by March 2025, to 224 382. ² Despite increasing demand, many services have reported reduced funding, incomplete multidisciplinary teams, fewer key professionals and an inability to complete comprehensive assessments in a timely manner. ^{Reference Parr, Wigham, Farr, Male, Isard and Lees3}

It is important to acknowledge that autism services were significantly affected during the COVID-19 pandemic, with many services undergoing substantial restructuring and some services being suspended altogether. ⁴ COVID-19 also altered the way in which autism assessments were conducted. Face-to-face assessments, such as the Autism Diagnostic Observation Schedule, Second Edition (ADOS-2), were replaced by non-standardised, non-validated remote assessments. Although standardised assessments such as the Brief Observation of Symptoms of Autism (BOSA) and TELE-ASD-PEDS (TAP) were developed and have shown promise as valid diagnostic assessment tools, ^{Reference Dow, Holbrook, Toolan, McDonald, Sterrett and Rosen5,Reference Wagner, Corona, Weitlauf, Marsh, Berman and Broderick6} during COVID-19, many services developed their own ‘ADOS-2 informed’ assessments. After COVID-19 restrictions were lifted, many services decided not to return to face-to-face assessments. This decision may have been driven by practical and financial reasons rather than clinical evidence.

While the National Institute for Health and Care Excellence ⁷ recommends that autism diagnostic assessments should begin within 12 weeks of referral, this is not the experience of the vast majority of individuals. For example, in March 2025, 95.5% of individuals had not started their assessment within the recommended 12-week period (or 13 weeks according to NHS statistics). ² The average waiting time for an autism assessment is, in fact, more than 12 months and, in some areas, considerably longer. This evidences that the 2023 Autism Assessment Framework and Operational Guidance have not been as effective as hoped in the context of increasing demand for assessments.

Given national policy expectations and the current limitations of NHS services in meeting such demand, the NHS and individuals and families have explored other diagnostic assessment options. Some NHS services have sought to reduce waiting lists through the government-funded Waiting List Initiative. These initiatives allow private organisations to bid for multi-million-pound contracts to complete hundreds, if not thousands, of autism assessments across NHS Trusts. As with all government contracts, cost is a key consideration in the procurement process. In addition, the introduction of the Right to Choose framework ⁸ has allowed families and individuals to choose their own NHS-funded private assessor but under certain conditions and cost restrictions. Alternatively, some individuals and families may choose to pay privately for an assessment.

However, there is considerable variation in the quality of assessments and associated reports within and between NHS services and private practice. ⁹ For example, to reduce costs, some services may undertake only a brief autism-specific developmental history or use non-standardised, non-validated online observations. They may also employ less experienced professionals to administer complex assessment tools. Such clinicians may have less clinical experience and may not be trained in differential diagnosis. Other factors that may contribute to variation in diagnostic decisions include inconsistent determination of impairment as a component of diagnosis, perceived pressure to provide a diagnosis when individuals or families are paying high fees for assessments and differences in the weight given to self-report or family report compared with observable features required for a diagnosis. Furthermore, clinicians differ in their use and interpretation of the concept of ‘camouflaging’, which is broadly understood as behavioural or cognitive strategies used to adapt to or cope within a predominantly non-autistic social world. The extent to which camouflaging influences diagnostic decision-making remains unclear.

Following assessment, comprehensive formulations and reports are recommended to inform support needs. However, to reduce time and associated costs, diagnostic letters with non-individualised recommendations may be provided as an alternative. Robust individual formulations are essential not only for those diagnosed with autism but also for those who do not meet diagnostic criteria, as subthreshold autistic characteristics or alternative diagnoses may be more relevant to their current or future mental health and adaptive functioning. Ultimately, given recent changes to assessment methods in response to increasing demand, the clinical quality of many autism assessments may be at risk.

Variation in the quality of autism assessments means that, in some cases, autism may be missed or incorrectly diagnosed. This inaccuracy can have a profound effect on the trajectory of an individual’s life and may misdirect them away from the most appropriate support. Inadequate assessments and/or incorrect diagnostic conclusions may lead service users to disagree with the diagnostic outcome, either immediately after the assessment or at a later date, and request a review or second opinion. In cases where a diagnosis of autism has been made and the individual disagrees with this outcome, they may request the removal of the diagnosis from their medical records. This may be particularly difficult for children who were not given the choice to complete an autism assessment and disagree with the outcome later in life.

The absence of a detailed report may also mean there is insufficient information to provide a second opinion based on a review of the existing information and a clinical update. Instead, it may be necessary to undertake an entirely new assessment, including a detailed developmental history interview. The burden for the cost of future reassessments will be borne either by families and individuals who opt to pay for a private second opinion assessment or by the government, should the reassessments be conducted via the NHS. Therefore, the NHS may be faced with a growing demand for reassessments alongside the continuing rise in referrals for initial assessments. Currently, neither the NHS nor private services are set up to manage the future demand for review and second-opinion assessments. In other words, although quicker and cheaper assessments may help NHS Trusts meet government targets by reducing waiting lists in the short term, they may also increase future demand for reassessments and place additional pressure on services, with the potential for unmet needs arising from inaccurate diagnosis

This emphasises the importance of ensuring that comprehensive and robust assessments, including a suitable screening and triage process, are conducted by suitably trained, qualified and experienced professionals. Adequate time should be allocated to ensure that comprehensive reports are produced, differential diagnoses are considered and bespoke care plans are developed. While recommendations and guidelines exist, quality control procedures could be introduced to prevent future harm to individuals and negative implications for services.