A Guardian advice column about one 15-year-old with worsening obsessive-compulsive disorder was turned into a claim that teenage OCD rates were rising during failed therapy attempts. The column contained no prevalence survey, case series or time trend. It cannot support that headline.
The parent's account was still important. Her daughter had completed two private therapy courses, but rituals continued to change and interfere with reading, schoolwork and punctuality. The useful question raised by the column was not whether a population rate had increased. It was whether the treatment she received matched the evidence-based approach recommended for OCD.
One Family's Account Is Not Epidemiology
The mother described counting, repeated switching, ordered washing and arranging objects. When one ritual stopped, another could appear. The teenager reread lines until she abandoned reading for pleasure, rewrote notes repeatedly and arrived late to school because of morning rituals.
Those details show functional impairment in an individual case. They do not reveal how common OCD was among teenagers in April 2026, whether prevalence had changed or whether unsuccessful therapy causes rates to rise. A prevalence claim would require a defined population, a consistent diagnostic method, a comparison period and uncertainty estimates. None appeared in the source.
Oxford clinical psychologist Polly Waite told the Guardian that adolescence is one of the peak periods for OCD to develop, a stage marked by change, transition and increased responsibility. A peak age of onset is not the same as evidence that population rates are climbing.
The Prior Treatment Was Not Clearly ERP
The mother said both therapy courses focused on eliminating one compulsion at a time. The Guardian and Waite said they did not know whether the teenager had received cognitive behavioural therapy, or CBT, with exposure and response prevention, known as ERP.
That uncertainty is central. Calling two courses “failed therapy” does not show that a complete evidence-based treatment failed. It may mean the intervention used another method, did not include ERP or was not delivered in a form suited to this teenager. The source did not provide treatment length, clinician qualifications, adherence or outcome measures.
Waite described ERP as learning through action. In the Guardian's account, the young person encounters an intrusive thought or feared situation, refrains from the compulsion and experiences anxiety changing over time. The aim is not to prove that a feared event is impossible; it is to reduce the rule that a ritual must follow the thought.
This is more precise than saying therapy must uncover one hidden fear behind every visible ritual. OCD can include obsessions, compulsions and mental neutralising strategies in different forms. The treatment description must match the patient's assessment rather than an internet summary of a universal mechanism.
NICE Recommends CBT With ERP and Family Involvement
The National Institute for Health and Care Excellence recommends CBT including ERP as the treatment of choice for children and young people with moderate to severe functional impairment. The treatment should involve family or carers and be adapted to the child's developmental age.
NICE also says treatment should collaboratively identify targets, maintain a therapeutic alliance and encourage use of ERP when new or different symptoms emerge. When compulsions interfere with ordinary functioning, liaison with teachers and other professionals may be appropriate.
Family involvement does not mean relatives should improvise exposures or abruptly refuse every request for reassurance. NICE says plans should help relatives reduce involvement in compulsions, avoidance or reassurance seeking in a sensitive and supportive way. The guideline treats this as part of a planned intervention.
The Guardian column mentioned school support and specialist resources, but it did not set a schedule for removing accommodations. Educational decisions depend on the young person's functioning and coordinated plan. Portraying support as surrender to the disorder would go beyond both the source and the guideline.
Nonresponse Has a Defined Review Path
NICE specifies that if a child or young person has not had an adequate response within 12 weeks to a full trial of CBT including ERP with family involvement, a multidisciplinary review should take place. That threshold separates a documented nonresponse from the vague statement that therapy did not work.
The review can consider whether treatment was adequately delivered and whether coexisting conditions, learning difficulties or psychosocial factors affect the response. It is not evidence that the teenager is resisting care, and it does not justify diagnosing additional conditions from a parent's letter.
For young people aged 12 to 18 with moderate to severe impairment, NICE says that after multidisciplinary review, adding an SSRI to continuing psychological treatment should be offered when there has not been an adequate CBT-and-ERP response. It also says prescribing and dose decisions for minors should involve a child and adolescent psychiatrist, with careful monitoring. That is guideline context, not a medication recommendation for the teenager described by the Guardian.
The Evidence Gap Begins With the Headline
The column responsibly asked whether a specific teenager had received the right type of help. The original report converted that narrow question into a rising-rate story and then supplied treatment conclusions the source could not verify.
A credible account must keep three things separate: one family's experience, clinical guidance for young people with OCD and population evidence about prevalence. The first can illuminate the burden of symptoms. The second can show what an adequate treatment pathway looks like. Neither proves the third.
The hard conclusion is that failed or poorly specified therapy cannot be used as a surveillance system. If teenage OCD rates are rising, that claim needs diagnostic and population data. This source had none. Its real lesson was more disciplined: before declaring treatment failure, establish what treatment was delivered, whether ERP and family work were included, how long it lasted and how response was measured.