This chapter draws on current major guidelines and the secondary literature commonly cited in board-prep references. Sources include the APA Practice Guideline for the Treatment of Patients with OCD, the NICE clinical guideline CG31, the Canadian clinical practice guidelines from the Anxiety Disorders Association of Canada, the World Federation of Societies of Biological Psychiatry (WFSBP) guidelines, the British Association for Psychopharmacology guidelines, recent Cochrane systematic reviews on SSRIs and on CBT for OCD, and standard reference texts (Stahl's Essential Psychopharmacology, Kaplan and Sadock's Synopsis of Psychiatry, the Maudsley Prescribing Guidelines).^1-2,4-7 Search terms combined "obsessive-compulsive disorder" with "SSRI", "clomipramine", "CBT", "exposure and response prevention", "treatment-resistant", "deep brain stimulation", "transcranial magnetic stimulation", and "DSM-5-TR". Date of citation pass: 10 May 2026. Where guidelines disagreed, both positions are summarised.

Lifetime prevalence of OCD in adults is approximately 2 to 3 percent, with 12-month prevalence around 1 to 2 percent across large epidemiological surveys including the National Comorbidity Survey Replication.⁸ Onset is bimodal: an early peak in late childhood to early adolescence (mean around 10 years), particularly in males, and a second peak in early adulthood (early 20s) that is more common in females. Roughly a quarter of cases begin before age 14, and onset after age 35 is uncommon and should prompt a careful search for a medical or neurological cause.^1,8 Sex distribution is approximately equal in adults, with a male preponderance among childhood-onset cases. Comorbidity is the rule rather than the exception. Lifetime rates of major depressive disorder approach 60 to 80 percent in clinic samples, anxiety disorders co-occur in roughly three-quarters of patients, and tic disorders are seen in 20 to 30 percent — particularly in early-onset, male, and "just-right" symmetry-driven presentations.^1,9 Substance use disorders affect roughly a quarter of patients over the lifetime. The relationship to suicide is clinically important and historically underappreciated: patients with OCD have approximately ten-fold higher rates of suicide attempts and completed suicide than the general population, with risk concentrated in those with comorbid depression and severe symptoms.¹⁰ Risk factors include first-degree family history (heritability estimates 40 to 65 percent in twin studies, higher for childhood-onset cases),¹¹ perinatal complications, childhood streptococcal infection in a small subset (the PANDAS construct, discussed under Controversies), and adverse childhood experiences. There is no robust socioeconomic or ethnic gradient.

The dominant neurobiological model of OCD centres on hyperactivity within cortico-striato-thalamo-cortical (CSTC) loops, particularly a circuit running from the orbitofrontal cortex and anterior cingulate cortex through the caudate nucleus and thalamus and back to cortex.¹² Functional imaging studies — including PET, SPECT, and task-based fMRI — consistently show increased resting metabolism in the orbitofrontal cortex, anterior cingulate, and head of the caudate in patients with OCD, with normalisation following successful treatment with either SSRIs or CBT.^12-13 The ventral striatum and the Default mode network are also implicated. The current working model frames obsessions as outputs of an overactive error-detection signal that fails to be "switched off" by striatal gating, with compulsions emerging as overlearned, habit-like behaviours subserved by dorsal striatal circuits. Serotonergic dysfunction is the longest-standing neurochemical hypothesis, supported by the selective efficacy of serotonin reuptake inhibitors (SSRIs and clomipramine) and the relative ineffectiveness of noradrenergic-only agents such as desipramine.¹⁴ However, response is partial in many patients and serotonin alone cannot explain the disorder; converging evidence implicates glutamatergic dysregulation in CSTC circuits, supported by elevated CSF and MRS glutamate metabolites and by emerging — though still preliminary — evidence for glutamate-modulating agents such as memantine, riluzole, and ketamine.¹⁵ Dopaminergic involvement is suggested by the augmentation efficacy of low-dose antipsychotics, particularly in tic-related OCD. Genetics contribute substantially. Twin heritability is estimated at 40 to 65 percent, with childhood-onset OCD showing the highest familial loading.¹¹ Genome-wide association studies have identified suggestive signals in glutamatergic and serotonergic genes (SLC1A1, HTR2A) but no replicated locus of large effect, consistent with a polygenic architecture shared partially with Tourette syndrome and other anxiety disorders. Environmental contributors include perinatal hypoxia, traumatic life events, and — in a small minority of paediatric cases — group A streptococcal infection in the PANDAS/PANS framework. Cognitive-behavioural models, which inform the dominant psychotherapy, frame OCD as the misinterpretation of normal intrusive thoughts as personally meaningful and dangerous, with compulsions maintained through negative reinforcement (anxiety reduction).¹⁶

DSM-5-TR defines OCD by the presence of obsessions, compulsions, or both, that are time-consuming (more than one hour per day) or cause clinically significant distress or functional impairment, and that are not better explained by another disorder, substance, or medical condition.³ Obsessions are recurrent and persistent thoughts, urges, or images experienced as intrusive and unwanted, that the person attempts to ignore, suppress, or neutralise with another thought or action. Compulsions are repetitive behaviours (handwashing, checking, ordering) or mental acts (counting, praying, silently repeating words) performed in response to an obsession or according to rigid rules, aimed at reducing distress or preventing a feared outcome — but the act is either not realistically connected to what it is meant to prevent, or is clearly excessive. Specifiers refine the clinical picture and have prognostic and treatment implications. Insight is rated as good or fair, poor, or absent/delusional. Roughly 15 to 30 percent of adults have poor insight, and 2 to 4 percent have absent insight at presentation; poor insight predicts greater symptom severity, more comorbidity, and a more guarded treatment response.^3,17 A separate tic-related specifier is applied when the patient has a current or past tic disorder; this subtype responds preferentially to antipsychotic augmentation.

Two further DSM-5-TR points often missed in training: there is no longer an "egodystonic" requirement, and the "adults recognise the obsessions as excessive or unreasonable" criterion was removed in DSM-5 — this avoided false-negative diagnoses in patients with poor insight. Symptom dimensions are not formal specifiers but are clinically useful: contamination/cleaning, symmetry/ordering, forbidden or taboo thoughts (aggressive, sexual, religious), and harm/checking. These dimensions are stable, partly heritable, and predict differential treatment response.

Most patients describe months to years of subthreshold symptoms before reaching diagnostic threshold, with the average lag from symptom onset to first effective treatment historically reported at over a decade.² Onset is typically gradual; abrupt onset, particularly in a child, should raise concern for an autoimmune or infectious aetiology.

Symptom dimensions

Contamination obsessions with washing or cleaning compulsions are the most prevalent dimension and the easiest to recognise — chapped hands, raw forearms, and avoidance of public surfaces. Symmetry and ordering symptoms involve a sense of something being "not just right" until objects are aligned, items are counted, or movements are repeated equally on both sides. Forbidden or taboo thoughts involve intrusive aggressive, sexual, or religious imagery (so-called scrupulosity) experienced as repugnant; the patient is profoundly distressed precisely because the thought is incongruent with their values, and may engage in mental neutralisation, reassurance-seeking, or avoidance rather than overt rituals. Harm and checking obsessions centre on a fear of having caused or failing to prevent danger — repeated checking of locks, stoves, and driving routes is typical.

Course

Untreated OCD is typically chronic and waxing-waning, with full spontaneous remission uncommon. Severity often fluctuates with stressors, sleep, and depressive comorbidity. A minority of patients have an episodic course with discrete periods of remission, more common in adult-onset cases.

Red flags

Abrupt or explosive symptom onset in a child (consider PANDAS/PANS), new-onset OCD after age 40 (consider basal ganglia lesion, frontotemporal pathology, or substance), OCD-like symptoms accompanied by neurological signs (chorea, tics with cognitive decline), and severe insight loss with bizarre content (re-examine for a primary psychotic disorder). Hoarding behaviour that dominates the clinical picture warrants consideration of hoarding disorder as a separate diagnosis under DSM-5-TR.

The differential is broad because intrusive thoughts and repetitive behaviours occur transdiagnostically. The diagnostic discipline is to distinguish ego-dystonic, anxiety-driven obsessions and rituals from phenomenologically similar but mechanistically different presentations. Generalised anxiety disorder shares worry, but GAD worries concern realistic, future-oriented life domains (finances, health, work) and are not paired with stereotyped neutralising rituals. Body dysmorphic disorder, Hoarding disorder, trichotillomania, and excoriation disorder share the obsessive-compulsive related disorder chapter and overlap mechanistically; co-diagnosis is common but each has a distinct phenotypic core. Obsessive-compulsive personality disorder (OCPD) is frequently confused with OCD on shelf exams and in clinic. OCPD is ego-syntonic, characterised by perfectionism, rigidity, and preoccupation with rules and order experienced as virtuous rather than distressing; there are no true obsessions or compulsions. The two can co-occur but are distinct. Psychotic disorders enter the differential when insight is absent and obsessions take on a delusional quality. The discriminating feature is the presence of compulsions, the egodystonic phenomenology, and the absence of formal thought disorder, hallucinations outside the obsessive content, or negative symptoms. Tic disorders and Tourette syndrome share repetitive behaviours; tics are typically preceded by a premonitory sensory urge rather than by a cognitive obsession, and are stereotyped and motor rather than goal-directed. Autism spectrum disorder includes restricted, repetitive behaviours, but these are typically ego-syntonic, sensorily driven, and present from early childhood without an associated obsession. Medical and substance-induced mimics deserve a structured search, particularly with late or abrupt onset:

• Streptococcal infection in children (PANDAS/PANS): abrupt onset, choreiform movements, emotional lability.
• Basal ganglia disorders: Sydenham chorea, Huntington disease, post-encephalitic parkinsonism, Wilson disease.
• Frontal-lobe pathology: tumour, stroke, or frontotemporal dementia presenting with new-onset perseverative behaviour.
• Autoimmune encephalitis (anti-NMDA receptor and others) — consider with rapid onset, seizures, or autonomic instability.
• Stimulant intoxication or chronic methamphetamine use can produce "punding" — repetitive, purposeless behaviours that mimic compulsions.

Diagnosis is clinical and rests on a structured interview that explicitly screens for obsessions and compulsions. Patients rarely volunteer symptoms because of shame, particularly with taboo content. Direct, normalising questions are essential: ask about repeated washing, checking, counting, ordering, and intrusive aggressive, sexual, or religious thoughts; ask how much time per day rituals consume and what the patient avoids. The Florida Obsessive-Compulsive Inventory and the brief screening item set developed for primary care can be used at intake.

Mandatory history elements

Age and tempo of onset, full symptom inventory across the four dimensions, time consumed, level of insight, avoidance, family accommodation (a major maintenance factor), tic history, prior treatment trials with adequate dose and duration, comorbid depression and substance use, suicide risk, and pregnancy or contraception status if pharmacotherapy is being considered.

Validated rating scales

The Yale-Brown Obsessive Compulsive Scale (Y-BOCS), and its self-report variant the Y-BOCS-SR, is the gold-standard severity measure used in essentially all OCD trials.¹⁸ The 10 items rate time, interference, distress, resistance, and control over obsessions and compulsions on 0–4 anchors, yielding a 0–40 total. Conventional severity bands: 0–7 subclinical, 8–15 mild, 16–23 moderate, 24–31 severe, 32–40 extreme. A 25 to 35 percent reduction from baseline is the standard definition of treatment response in adult trials. The Children's Y-BOCS (CY-BOCS) is the paediatric equivalent. The Dimensional Obsessive-Compulsive Scale (DOCS) and OCI-R are useful self-reports for tracking response between visits.

Physical exam and laboratory workup

Routine evaluation includes a focused neurological examination, vital signs, and a basic metabolic and thyroid panel to exclude common medical mimics — TSH, complete blood count, basic metabolic panel, and a fasting glucose or HbA1c if pharmacotherapy with weight-affecting agents is anticipated. Pregnancy testing in patients of reproductive age before initiating an SSRI. Neuroimaging is not routine but is appropriate with abrupt or late-onset symptoms, focal neurological signs, or marked cognitive decline. EEG is not indicated unless seizure is part of the differential. Anti-streptolysin O and anti-DNase B titres, plus paediatric neurology consultation, are appropriate when PANDAS/PANS is clinically suspected; they should not be ordered reflexively.

Treatment is anchored by two evidence-supported modalities: CBT incorporating exposure and response prevention, and SSRIs (or clomipramine). Either is acceptable as monotherapy in mild-to-moderate disease; combination therapy is commonly recommended for severe disease and after partial response to monotherapy.^1-2,4

Pharmacotherapy

First-line pharmacotherapy is an SSRI. All SSRIs appear roughly equivalent in efficacy in head-to-head trials and meta-analyses; choice is driven by side-effect profile, drug interactions, and patient factors.^5-6 Fluoxetine, fluvoxamine, sertraline, and paroxetine carry FDA approvals for OCD; escitalopram and citalopram are commonly used off-label and recommended by major guidelines. Two principles distinguish SSRI dosing in OCD from MDD: doses are typically higher (often near the upper end of the licensed range), and onset of meaningful response takes longer — typically 6 to 12 weeks at adequate dose, with continued gains for up to 6 months.^1,5 Typical adult target ranges, drawing on FDA labels and the Maudsley Prescribing Guidelines: fluoxetine 40 to 80 mg, fluvoxamine 200 to 300 mg, sertraline 150 to 200 mg, paroxetine 40 to 60 mg, escitalopram 20 mg (with up to 30 mg sometimes used by specialists, balanced against QTc considerations), and citalopram 40 mg (with the FDA QTc-related 40 mg ceiling in adults under 60).⁶ Start low to mitigate early activation and titrate over weeks. Clomipramine, a tricyclic with potent serotonin reuptake inhibition, has the longest evidence base and may be marginally more effective in head-to-head meta-analyses, though SSRIs are typically preferred first because of tolerability.¹⁴ Target dose is 150 to 250 mg daily; pretreatment ECG and monitoring for anticholinergic effects, orthostasis, weight gain, sexual dysfunction, and seizure risk (dose-related, particularly above 250 mg) are standard. Plasma level monitoring (clomipramine plus desmethylclomipramine) is appropriate at higher doses.

An adequate SSRI trial is defined as 10 to 12 weeks at the maximum tolerated or licensed dose.¹ Roughly 40 to 60 percent of patients meet response criteria with a first SSRI; of non-responders, switching to a second SSRI or to clomipramine yields response in another 20 to 40 percent. Patients are commonly only "partial responders" — symptoms attenuate but remain disabling — which is the entry point for augmentation. Augmentation strategies, in approximate order of evidence: low-dose antipsychotic augmentation (risperidone 0.5–2 mg, aripiprazole 5–15 mg) has the strongest meta-analytic support, with response in roughly one-third of SSRI non-responders and a particularly favourable signal in tic-related OCD.¹⁹ Adding CBT with ERP to an SSRI is at least as effective as antipsychotic augmentation and is preferred when accessible. Glutamate modulators (memantine, riluzole, N-acetylcysteine, topiramate) have lower-quality evidence and are reserved for treatment-resistant cases under specialist care. Lamotrigine and ondansetron have suggestive but not definitive data. SSRI plus clomipramine, intravenous clomipramine, and high-dose SSRI strategies (e.g. escitalopram 30 mg, sertraline 400 mg) are options used by specialist services.

Psychotherapy

Exposure and response prevention is the highest-evidence psychological treatment for OCD and the psychotherapy benchmark against which others are judged.^20-21 In ERP, the patient is systematically and repeatedly exposed to feared stimuli or thoughts (a contaminated doorknob, an intrusive blasphemous image) while abstaining from the compulsive response. The therapeutic mechanism is inhibitory learning — the patient learns that feared outcomes do not occur and that anxiety habituates without the ritual. Effective ERP requires a graded exposure hierarchy, in vivo and imaginal exposures, homework between sessions, family work to reduce accommodation, and typically 13 to 20 sessions delivered weekly or more intensively. Effect sizes in trials are large (Cohen's d typically 1.0 to 1.5 vs. waitlist) and superior to non-ERP CBT.²¹ Cognitive therapy elements (challenging inflated responsibility, thought-action fusion, and overestimation of threat) are commonly integrated. Acceptance and commitment therapy and inference-based CBT have growing but less mature evidence bases. Internet-delivered CBT with therapist support, as well as group-format ERP, broaden access where individual therapists are scarce.

Neuromodulation

Repetitive transcranial magnetic stimulation (rTMS) targeting the medial prefrontal cortex and anterior cingulate via deep H7-coil TMS received FDA clearance for OCD in 2018; effect sizes are modest (Y-BOCS reduction of approximately 6 points vs. sham in the pivotal trial), and patient selection and protocol fidelity matter.²² Other targets (dorsolateral prefrontal cortex, supplementary motor area) have suggestive evidence. Deep brain stimulation received an FDA Humanitarian Device Exemption for severe, treatment-resistant OCD in 2009; targets include the ventral capsule/ventral striatum and the subthalamic nucleus. Open-label and small randomised series report response in roughly 50 to 60 percent of carefully selected, severely refractory patients.²³ Use is restricted to specialist centres with multidisciplinary review. Stereotactic ablative procedures (anterior capsulotomy, anterior cingulotomy) remain available at very few centres for the most refractory cases. Electroconvulsive therapy is not effective for OCD itself and is not indicated unless severe comorbid depression, catatonia, or psychosis warrants it.

Adjunctive

Education of the patient and family, exercise, sleep regularisation, and treatment of comorbid depression and substance use are routinely incorporated. Self-help workbooks built around ERP principles and supported online platforms can extend gains between sessions. Peer support groups reduce shame, particularly around taboo-content obsessions. Hospitalisation is rarely needed except for severe self-neglect, suicide risk, or for intensive residential ERP programmes when outpatient treatment has failed.

The table below summarises the principal interventions for adult OCD, certainty graded with the GRADE wording convention applied throughout this chapter.

Common SSRI adverse effects include gastrointestinal upset, headache, insomnia or somnolence, transient activation early in treatment, and sexual dysfunction (rates 30 to 60 percent depending on agent and ascertainment).^5-6 Sexual side effects are commonly underreported and a leading cause of non-adherence; routine direct enquiry is appropriate. Increased risk of bleeding (particularly upper GI in patients also on NSAIDs or anticoagulants) and hyponatraemia (more common in older adults) are dose-related class effects. Discontinuation symptoms (dizziness, paraesthesiae, flu-like malaise, mood instability) can be substantial with short-half-life agents (paroxetine, fluvoxamine) and warrant slow tapers, often over weeks to months. Clomipramine adds the tricyclic harm profile: anticholinergic load (dry mouth, constipation, urinary retention, cognitive blunting, particularly in older adults), orthostatic hypotension, cardiac conduction effects with QTc prolongation, weight gain, and a dose-related seizure risk that becomes meaningful above 250 mg.¹⁴ Lethality in overdose is substantially higher than for SSRIs; prescribing in patients at acute suicide risk requires care. Antipsychotic augmentation introduces metabolic risk (weight gain, dyslipidaemia, glucose intolerance), extrapyramidal effects, prolactin elevation (with risperidone), and sedation. Long-term tardive dyskinesia risk is real and must be balanced against benefit; reassess the need for augmentation periodically rather than continuing indefinitely by default. ERP is highly effective but not benign: exposures are deliberately distressing in the short term, and dropout in intensive protocols is meaningful. Therapist availability is the dominant practical limitation in most healthcare systems; quality varies, and ERP delivered without proper response prevention is not ERP. Limitations of the evidence base include relatively short trial durations (typically 10 to 16 weeks), exclusion of patients with prominent suicidality and severe comorbidity from many trials, modest representation of older adults and ethnic minorities, and publication bias favouring positive industry-sponsored studies for novel agents. Long-term comparative-effectiveness data beyond two years remain sparse for most augmentation strategies and for neuromodulation.

Children and adolescents

Paediatric OCD is common, with onset peaking around age 10. CBT with ERP is first-line monotherapy for mild-to-moderate paediatric OCD and is at least as effective in this age group as in adults; the Pediatric OCD Treatment Study (POTS) showed combined CBT plus sertraline superior to either monotherapy in moderate-to-severe disease, with CBT alone superior to sertraline alone.²⁴ FDA-approved SSRIs in paediatric OCD include fluoxetine (≥7 years), sertraline (≥6 years), fluvoxamine (≥8 years), and clomipramine (≥10 years). Black-box warnings for suicidality apply; informed consent and close monitoring during the first weeks of treatment are mandatory. Family involvement to address accommodation is a critical adjunct.

Older adults

OCD persists into later life, and a small proportion of cases first present after 60 — these warrant particularly careful neurological evaluation. Polypharmacy, hyponatraemia risk with SSRIs, and anticholinergic burden with clomipramine make sertraline and escitalopram common first choices. Antipsychotic augmentation in older adults is approached cautiously given fall, stroke, and mortality signals in dementia populations.

Perinatal

Pregnancy and the postpartum period commonly worsen OCD, with postpartum-onset obsessions about harm coming to the infant being especially distressing and frequently mistaken for postpartum psychosis or maternal infanticidal ideation. The discriminating feature is that postpartum OCD obsessions are ego-dystonic, accompanied by avoidance and reassurance-seeking rather than intent. SSRIs are generally compatible with pregnancy and breastfeeding; sertraline has the most reassuring breastfeeding data. Paroxetine carries small absolute risk signals for cardiac malformations and is generally avoided in pregnancy.²⁵ Risk-benefit discussion should be explicit and documented.

Comorbid medical illness and substance use

Cardiac disease warrants ECG before clomipramine and consideration of QTc-aware SSRI choice. Hepatic and renal impairment require dose adjustment for some agents; sertraline is generally a safe default. Comorbid alcohol or stimulant use disorders should be addressed in parallel; OCD rarely remits while active substance use continues, and ERP requires sobriety to consolidate inhibitory learning.

Cultural considerations

Symptom content tracks cultural and religious context. Scrupulosity in observant religious patients (Catholic, Orthodox Jewish, Muslim) requires care to distinguish pathological doubt and reassurance-seeking from valued religious practice; collaboration with culturally informed clergy is often productive. Stigma and disclosure norms shape help-seeking and adherence and should be addressed openly in the formulation.

Untreated OCD is typically chronic, with full remission rates in the absence of treatment estimated at under 20 percent over decades of follow-up.^2,17 With current first-line treatment, 40 to 60 percent of patients meet response criteria on Y-BOCS within 10 to 12 weeks, and roughly half of treated patients achieve remission (commonly defined as Y-BOCS ≤12 with minimal functional impairment) over 1 to 2 years of combined pharmacotherapy and ERP. Relapse on SSRI discontinuation is high, with rates around 25 to 50 percent in the year after stopping medication; current guidelines recommend maintaining an effective medication for at least 1 to 2 years after response, and in many patients indefinitely.^1,4 CBT-derived gains are more durable than pharmacological-only gains after treatment ends. Predictors of poorer outcome include earlier age of onset, longer duration of untreated illness, hoarding-dimension symptoms, poor insight, severe comorbid depression, schizotypal personality features, and high family accommodation. Functional impairment is substantial: WHO data place OCD among the leading causes of years lived with disability among non-fatal mental disorders. Suicide risk is elevated approximately ten-fold over the general population; risk is concentrated in patients with comorbid depression, severe symptoms, hopelessness, and substance use, and warrants explicit and repeated assessment.¹⁰ Mortality from suicide and from medical complications of compulsive behaviours (severe self-injury through cleaning, exhaustion-related events) is meaningfully elevated relative to the general population.

Most OCD is managed in outpatient settings, but several presentations warrant urgent attention. Suicide risk in OCD is concentrated in patients with comorbid depression, severe symptoms, hopelessness, and a history of attempts; ask explicitly about suicide at every visit and document the assessment.¹⁰ Severe self-neglect from compulsive behaviour — dehydration from contamination-driven food avoidance, dermatologic injury from washing, exhaustion from prolonged rituals — can rise to a level requiring inpatient stabilisation and intensive ERP. Acute agitation in OCD is typically driven by an unresolvable obsession or by being prevented from completing a ritual. The clinical priority is to reduce environmental triggers, contain the patient safely, and avoid reinforcing rituals through reassurance or accommodation in the acute setting. Benzodiazepines may transiently relieve distress but do not treat OCD and should not be used as a maintenance strategy. New-onset psychotic features, abrupt symptom escalation in a child, or a dramatic change in baseline OCD pattern warrant evaluation for a medical or substance-induced cause.

PANDAS and PANS

The Paediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal infections (PANDAS) construct, and its broader successor PANS (Paediatric Acute-onset Neuropsychiatric Syndrome), describe abrupt-onset OCD or tic symptoms in children, hypothesised to result from post-streptococcal autoimmunity affecting basal ganglia. The construct remains contested. Specialty consensus statements support targeted antibiotic and immunomodulatory treatment in defined cases; sceptics point to inconsistent biomarker findings and the difficulty of distinguishing PANDAS from typical childhood-onset OCD with infection-coincident worsening.²⁶ Most major paediatric guidelines recommend that abrupt-onset cases be evaluated by a specialist team but do not endorse routine antibiotic prophylaxis or IVIG outside research protocols.

Cannabis and cannabinoids

Cannabidiol and the THC analogue nabilone have generated patient and media interest but lack adequate randomised trial data for OCD. Heavy cannabis use commonly worsens anxiety and obsessions over time and is not recommended.

Psychedelics

Psilocybin has shown preliminary signals in small open-label studies, prompting active randomised trials. Evidence is currently insufficient to make a clinical recommendation, and regulatory status varies by jurisdiction.

Ketamine

Intravenous ketamine produces rapid but short-lived reduction in obsessions in small trials, with effects typically waning within a week. Whether repeated dosing or oral ketamine analogues offer durable benefit in OCD remains an open question. Esketamine is not approved for OCD.

Hoarding as a separate disorder

DSM-5 separated hoarding disorder from OCD in 2013, reflecting evidence that hoarding has a distinct phenotype, family history, neuroimaging signature, and treatment response. Some experts continue to argue that hoarding sits on an OCD spectrum; the practical consequence is that hoarding-dominant presentations respond poorly to standard OCD pharmacotherapy and require specialised CBT protocols.

Long-term medication strategy

Guidelines disagree on duration of maintenance. APA and NICE recommend at least 1 to 2 years of continued effective treatment after response; some specialist groups argue for indefinite continuation given high relapse rates on discontinuation. The case for tapering after sustained remission and successful ERP exists but is supported mainly by expert opinion rather than long-term RCT data.

• DSM-5-TR moved OCD out of the anxiety disorders chapter into its own Obsessive-Compulsive and Related Disorders chapter; criteria require obsessions, compulsions, or both, that consume more than one hour daily or cause significant distress or impairment.
• Lifetime prevalence is approximately 2–3 percent. Onset is bimodal: childhood (male predominance) and early adulthood (slight female predominance). Onset after age 35 should prompt a search for a medical cause.
• The Y-BOCS (10 items, 0–40) is the gold-standard severity scale; a 25–35 percent reduction defines treatment response in trials.
• First-line pharmacotherapy is an SSRI at higher doses than for MDD, with a 10–12 week trial at maximum tolerated dose required before declaring failure. Clomipramine is comparably effective but less tolerable.
• First-line psychotherapy is CBT with exposure and response prevention (ERP); it is at least as effective as SSRI monotherapy and produces more durable gains.
• For partial response to an SSRI, evidence supports either adding ERP or augmenting with a low-dose antipsychotic (risperidone or aripiprazole). Tic-related OCD responds preferentially to antipsychotic augmentation.
• Insight specifier matters: poor insight predicts worse outcome; absent/delusional insight does not by itself convert the diagnosis to a psychotic disorder.
• Suicide risk in OCD is approximately ten-fold the general population, concentrated in those with comorbid depression and severe symptoms.
• Postpartum harm obsessions are ego-dystonic and respond to OCD treatment; they must be distinguished from postpartum psychosis.
• Distinguish OCD from OCPD: OCD is ego-dystonic with true obsessions and compulsions; OCPD is ego-syntonic, characterised by perfectionism and rigidity without obsessions.
• Deep TMS (H7 coil) received FDA clearance for OCD in 2018; DBS is FDA-approved under a Humanitarian Device Exemption for severe refractory OCD.
• ECT is not effective for primary OCD.
• Family accommodation maintains symptoms and predicts worse outcome; assess and address it routinely.
• Maintenance pharmacotherapy after response should continue for at least 1–2 years; relapse rates are 25–50 percent in the year after SSRI discontinuation.

No external funding. No conflicts of interest declared. Drafted with Claude (Anthropic) and human-edited by the Psychpedia editorial team. Peer-review status: pending.