Version July 2025
INTRODUCTION —
Obsessive-compulsive disorder (OCD) is characterized by recurrent intrusive thoughts, images, or urges that cause anxiety or distress (obsessions), and by repetitive mental or behavioral acts (compulsions) that the individual feels driven to perform. Compulsions are done in response to the obsession (ie, to decrease distress or prevent a feared consequence from occurring, or according to rules that must be applied rigidly). Almost all individuals with OCD have both obsessions and compulsions.
OCD typically starts in childhood or adolescence, often persists throughout life if not successfully treated, and usually produces substantial impairment in functioning due to the severe and often-chronic nature of the illness.
OCD can be treated with cognitive-behavioral therapy using exposure and response prevention (CBT/ERP), serotonin reuptake inhibitors (selective serotonin reuptake inhibitors [SSRIs] or the serotonin reuptake inhibitor clomipramine), or a combination of the two. Frequently, patients experience a partial response to treatment and warrant augmentation strategies.
This topic and an associated algorithm discuss the initial treatment choice and subsequent management of OCD in adults (algorithm 1). Psychotherapy for OCD is discussed elsewhere. Other topics related to OCD in adults, children, adolescents, pregnancy, postpartum, and treatment of refractory OCD are found elsewhere.
●(See "Obsessive-compulsive disorder in adults: Epidemiology, clinical features, and diagnosis".)
●(See "Obsessive-compulsive disorder in adults: Psychotherapy".)
●(See "Obsessive-compulsive disorder in pregnant and postpartum patients".)
●(See "Deep brain stimulation for treatment of obsessive-compulsive disorder".)
●(See "Obsessive-compulsive disorder in children and adolescents: Treatment overview".)
CHOOSING INITIAL TREATMENT MODALITY —
For many individuals with obsessive-compulsive disorder (OCD), we favor initiating treatment with cognitive-behavioral therapy using exposure and response prevention (CBT/ERP). The initial treatment of OCD may consist of a serotonin reuptake inhibitor, CBT/ERP, or their combination. However, we consider many factors in determining the choice of initial treatment. For example, we consider the severity of symptoms, presence of suicidality, presence of comorbid psychiatric or other medical disorders, treatment history, ability and willingness to participate in CBT/ERP, availability of CBT/ERP, and patient preference.
As examples:
●Severe symptoms – For individuals with severe symptoms, we initiate treatment with combined modality (CBT/ERP plus pharmacotherapy). We typically assess clinical severity with the Clinical Global Impressions Scale (CGI) scale (5 or above) and/or the Yale-Brown Obsessive Compulsive Scale (Y-BOCS; score of 30 to 40 on this 40-point scale) [1]. However, if the severity of symptoms precludes participation in CBT/ERP, we begin with pharmacotherapy, as this can lessen symptoms sufficiently for the individual to be able to engage in CBT/ERP, if needed. (See 'Initiating pharmacotherapy' below.)
●Suicidality – For individuals with suicidality (suicidal ideation or behavior) we initiate treatment with combined modality including CBT/ERP plus pharmacotherapy.
Further discussion of assessment and management of suicidal behavior in adults is found elsewhere. (See "Suicidal ideation and behavior in adults".)
●Co-occurring psychiatric disorder – For individuals with a co-occurring disorder that is typically responsive to selective serotonin reuptake inhibitor (SSRI) treatment (eg, major depression or social anxiety disorder), we often initiate treatment with an SSRI with or without psychotherapy. (See "Major depressive disorder in adults: Approach to initial management" and "Social anxiety disorder in adults: Treatment overview", section on 'Medication management'.)
●Unwilling/unable to participate in CBT/ERP; no response to past CBT/ERP – We initiate pharmacotherapy (ie, SSRI) for individuals who are unwilling or unable to participate in CBT/ERP (eg, due to severity of symptoms or limited ability to critically examine thoughts or feelings) or for those who have previously not responded to cognitive-behavioral therapy (CBT). (See 'Initiating pharmacotherapy' below.)
●Individuals with past response to pharmacotherapy (eg, SSRI, clomipramine) – For individuals with a favorable response to a particular medication (with or without CBT/ERP) in the past, we typically begin treatment with the same medication (with or without CBT). For example, if an individual has successfully responded to clomipramine in the past, we would use clomipramine as our first choice again.
●Patient preference for combined treatment – We initiate treatment with both pharmacotherapy and CBT/ERP in individuals who express a preference for combined treatment and seem likely to benefit from the combined treatment.
Meta-analyses and individual trials have found both CBT/ERP and pharmacologic management with clomipramine or SSRIs to be effective treatments for OCD [2-12].
Treatment with the combination of medication and CBT/ERP has shown greater efficacy than pharmacotherapy alone; however, practical issues (eg, greater cost, increased frequency of appointments, availability of trained therapists) may make this option less appealing [9,13-17]. The addition of exposure and response prevention (ERP) to ongoing pharmacotherapy has been shown to improve rates of response [15-17]. The efficacy of combined modality treatment may be influenced by other factors such as the timing of the treatments (eg, if they begin together or sequentially) and medication dose [3]. (See 'Augmentation with CBT/ERP' below.)
Some studies suggest that CBT/ERP may be more effective than pharmacotherapy; however, most have shown similar effects. CBT/ERP, however, has the advantage of not exposing the patient to potential side effects of medication [3].
Further discussion of psychotherapy for OCD, including specifics on administering psychotherapy, can be found elsewhere. (See "Obsessive-compulsive disorder in adults: Psychotherapy".)
INITIATING PHARMACOTHERAPY
SSRI as preferred initial pharmacotherapy — We use a selective serotonin reuptake inhibitor (SSRI), rather than other antidepressants, as the first option for individuals with obsessive-compulsive disorder (OCD) who are starting pharmacotherapy (with or without cognitive-behavioral therapy [CBT]). Our preference comes from the extensive empirical support for their efficacy for OCD, favorable side effect profile, and ease of use (table 1 and algorithm 1) [2,9,18-30]. (See 'Choosing initial treatment modality' above and 'Efficacy of serotonin reuptake inhibitors' below and "Selective serotonin reuptake inhibitors: Pharmacology, administration, and side effects".)
We choose from among the SSRIs based on each medication’s side effect profile, drug-drug interactions, and the individual’s prior treatment response. In cases where the individual has successfully responded to an agent in the past, we often choose that agent again. We also consider patient preference. (See 'Choosing initial treatment modality' above.)
For example, in an individual who is concerned about medication-related weight gain, we would choose fluoxetine but avoid paroxetine, as paroxetine is associated with weight gain [31]. Due to cardiac conduction concerns, we do not use escitalopram in older individuals (age >65 years).
Indirect comparisons of effectiveness suggest that while SSRIs tested have similar efficacy for symptom reduction in treatment of OCD, they differ in terms of adverse effects [2]. Side effects of commonly used antidepressants used in the treatment of depression and other disorders is discussed elsewhere (table 2). (See 'Efficacy of serotonin reuptake inhibitors' below and "Selective serotonin reuptake inhibitors: Pharmacology, administration, and side effects".)
Starting and titrating SSRIs — To avoid side effects and increase tolerability, we start individuals at the lowest initial starting dose recommended for the individual medication (table 1).
We typically titrate at approximately twice a month intervals to the target dose for OCD. As an example, for an adult in the outpatient setting we often begin with fluoxetine 20 mg/day and titrate by 20 mg/day every other week, as tolerated. However, the pace of dose titration depends on several factors, including the patient’s sensitivity to side effects, frequency of visits, treatment setting, and patient adherence.
Furthermore, if symptoms appear to be improving at a lower dose we often continue at the lower dose and monitor further prior to continuing titration.
For an individual with reported sensitivity to side effects, panic disorder, or illness anxiety disorder, we typically start at a lower starting dose (eg, fluoxetine 10 mg) and titrate in 10 mg increments every 7 to 10 days, as tolerated. A lower starting dose (eg, half of the usual initial starting dose for adults) and a slower titration are recommended for youth and older adults.
We obtain an electrocardiogram (ECG) at baseline, at escitalopram doses of 30 mg/day and higher, and when otherwise clinically indicated for an individual. We also consider getting an ECG when titrating other SSRIs to supratherapeutic doses. Initiation of clomipramine is described elsewhere. (See 'Subsequent treatment' below.)
Starting dose, target dose and supratherapeutic dosing are found on the table (table 1).
Prescribing and titrating SSRIs, including discussion of adverse effects, are discussed further elsewhere. (See "Selective serotonin reuptake inhibitors: Pharmacology, administration, and side effects".)
Therapeutic trial — If a given medication is tolerated by the patient, we ensure that they have reached the target dose prior to considering it ineffective or making further changes (eg, titration to supratherapeutic dose, augmentation strategies) (table 1). (See 'Inadequate response to initial treatment' below.)
The medication trial should be at least 12 weeks, with at least six of these weeks at the target dose for OCD (if this high a dose is needed). We explain this to the individual so that they do not prematurely stop the medication before a therapeutic trial has been completed. The treatment response curve appears to follow a logarithmic shape, with the greatest incremental benefits (but not the full benefit) occurring on average by week 6 [18].
Most fixed-dose trials of SSRIs for OCD found that higher doses led to greater probability of response and/or greater mean degree of improvement, compared with lower doses [27,28,32-34]. However, a meta-analysis that supported this dose-response relationship also found that higher doses are associated with a higher proportion of dropouts due to side effects [29]. (See 'Efficacy of serotonin reuptake inhibitors' below and "Selective serotonin reuptake inhibitors: Pharmacology, administration, and side effects".)
Defining response and frequency of monitoring — We define response as a minimum of 35 percent decrease on the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) score. We define remission as Y-BOCS score ≤12 [35]. However, response to treatment of OCD has been variably defined. A 25 to 35 percent improvement on the Y-BOCS is considered partial response [35].
In a standard outpatient treatment setting, we typically monitor individuals who have been started on pharmacotherapy twice monthly or monthly for one month and then approximately monthly thereafter, with less frequent monitoring an option once substantial improvement or remission is obtained. However, the frequency of the visits depends on the severity of the individual’s symptoms, whether suicidal ideation is present, the complexity introduced by comorbid conditions, and the presence of side effects. We see individuals with suicidal ideation more frequently (eg, every two weeks).
At each visit, we review symptoms and their effect on the individual’s functioning, progress in treatment, side effects of medications, and medication adherence, and we reinforce the need for completion of a full therapeutic trial. Additionally, we administer a Y-BOCS (or the scale’s first three items) at each visit. If time does not permit, the self-report version of the Y-BOCS can be used [36].
Efficacy of serotonin reuptake inhibitors — Our goal is elimination of symptoms of OCD; however, monotherapy (as described so far) typically leads to partial amelioration of symptoms. In clinical trials, treatment with SSRIs generally leads to clinically significant response in 40 to 60 percent of participants. On average, individuals experience a 25 to 35 percent reduction in symptoms within 12 weeks of treatment with monotherapy [26,30].
Therefore, adding cognitive-behavioral therapy using exposure and response prevention (CBT/ERP), augmentation with adjunctive medications, or increasing the SSRI to a supratherapeutic dose is often necessary to achieve fuller response. (See 'Inadequate response to initial treatment' below and 'Subsequent treatment' below.)
Data from clinical trials support the use of serotonin reuptake inhibitors in the treatment of OCD [2,9,18-30]. As an example, in a meta-analysis of 17 randomized trials with a total of 3097 patients with OCD, SSRIs as a group were more effective than placebo in reducing symptoms of OCD at 6 to 13 weeks posttreatment (Y-BOCS weighted mean difference 3.21, 95% CI -3.84 to -2.57) [2]. Furthermore, in a meta-analysis of 13 studies including 2697 participants with OCD, individuals in the SSRI treatment group had a higher rate of response (variably defined as 25 to 35 percent reduction in Y-BOCS score or a Clinical Global Impressions Scale [CGI] score of 1 or 2) than individuals in the placebo arm (relative risk 1.84, 95% CI 1.56-2.17) [2]. The weighted mean difference for the individual SSRI medications included (citalopram, fluoxetine, fluvoxamine, paroxetine, and sertraline) were similar to one another, indicating that they appear to be similarly effective versus placebo in treating symptoms of OCD.
Direct and indirect comparisons between clomipramine and SSRIs have not shown either to be superior to the other for the treatment of OCD [3,9,19-25]. Meta-analysis of randomized trials that did not directly compare SSRIs with clomipramine have suggested that clomipramine has a greater effect size than SSRIs [37]. However, studies of SSRIs generally occurred in a later era than those for clomipramine, and likely included more individuals who had previously had an inadequate response to another agent.
ASSESSMENT OF RESPONSE AND NEXT STEPS
Good response — For individuals with a good response to either cognitive-behavioral therapy using exposure and response prevention (CBT/ERP), or pharmacotherapy, our next steps are as follows. For individuals who have responded to combination treatment we follow the next steps for each modality.
Response to CBT/ERP — For patients with a good response to cognitive-behavioral therapy (CBT), we encourage practicing skills learned during treatment. We offer booster sessions for those whose symptoms worsen or are in need of further treatment. Psychotherapy for obsessive-compulsive disorder (OCD) is discussed elsewhere. (See "Obsessive-compulsive disorder in adults: Psychotherapy", section on 'Frequency and duration'.)
Response to selective serotonin reuptake inhibitors — We continue pharmacotherapy for at least two to three years in most patients who achieve remission with medication.
OCD is usually a chronic disorder. In one long-term follow-up study, only 20 percent of OCD patients who participated in placebo-controlled trials of selective serotonin reuptake inhibitors (SSRIs) experienced remission at 10- to 20-year follow-up, and 49 percent were still experiencing significant OCD symptoms. Initial response to SSRI pharmacotherapy strongly predicted better long-term outcome, but even of those who responded to initial SSRI pharmacotherapy, only 31 percent remained in remission at 10- to 20-year follow-up [37]. However, a substantial proportion of study participants were no longer taking a serotonin reuptake inhibitor at the time of the follow-up assessment. Therefore, long-term relapse prevention is critically important.
The majority of relapse prevention studies of SSRIs show superiority to placebo in preventing relapse, and that discontinuation of maintenance SSRI treatment even after response is associated with significantly elevated risk of relapse [11]. However, the severity of comorbid disorders, suicidality, and other patient-specific factors play a role in the decision of whether to discontinue medication. In individuals with a comorbid disorder or suicidality that is being treated with the medication, we consider the possibility that comorbidity or suicidality could worsen if the medication dose is decreased. Our treatment plan is individualized to address these differences.
The decision to taper off medications after successful treatment is based on several clinical factors, including response to medications, level of psychosocial impairment, history of prior exacerbations upon discontinuation of medications, initial and current severity of symptoms, initial and current severity of comorbid disorders, past and current suicidality, and whether the patient has done CBT/ERP. For example, in an individual with OCD who has responded to medication treatment, but who had significant psychosocial impairment due to the disorder, or in an individual who has had recurrence with tapering of medications in the past, we would continue medications for a longer period of time (eg, three or more years).
If the decision is made to taper medications, we attempt a gradual taper [38]. We typically taper medications by up to 20 or 25 percent every month or two while monitoring for recurrence of symptoms. For example, in an individual who has responded to 300 mg/day of fluvoxamine, we would lower by 50 mg every month or two to 50 mg/day, then by 25 mg every month before discontinuing fluvoxamine. Most clinical trials of medications for OCD are short-term trials. Extended trials of SSRIs, which randomly assigned individuals who responded to medication to continued medication treatment or to placebo, have generally found that patients who continue medication have a lower rate of relapse than patients on placebo [28,39-41]. Relapse rates have varied widely, in part due to methodologic differences among studies. OCD patients who have completed a course of CBT/ERP have a significantly lower rate of relapse after medications are discontinued than do patients treated with pharmacotherapy alone [42].
Inadequate response to initial treatment
Potential causes — For individuals with an inadequate response (either minimal/nonresponse or partial response) to initial treatment, our first steps are to:
●Confirm adherence to medications – We assess adherence to medication by directly asking the individual at every visit. In some cases, side effects to medications may be limiting adherence. In cases where there is questionable adherence, we recommend approaches to improve adherence, such as use of a pill box, app, daily alarm, or other approach. We may involve family members or other supports to try to enhance adherence. We stress the need to take the medication every day and remain on medication for a full therapeutic trial to adequately assess response to the medication. (See 'Therapeutic trial' above.)
●Review factors affecting psychotherapy – We review the individual’s level of motivation, the length and frequency of therapy sessions, and the capacity for introspection. Additionally, we confirm that the individual is practicing exposures and resisting rituals in-between sessions. Adherence to exposure and response prevention (ERP) is one of the most robust predictors of both acute and long-term outcomes [43-45]. (See "Obsessive-compulsive disorder in adults: Psychotherapy".)
●Confirm diagnosis and assess for comorbid disorders affecting treatment response – We confirm the current diagnosis by review of history and symptoms and assess for the presence of overlooked comorbid conditions (eg, mood disorders, anxiety disorders, personality disorders, posttraumatic stress disorder, substance use disorders, tic disorders) or psychosocial stressors that might be affecting treatment response. (See "Obsessive-compulsive disorder in adults: Epidemiology, clinical features, and diagnosis", section on 'Comorbidities'.)
Adjusting treatment — For individuals with inadequate response to initial management, after addressing potential causes (see 'Potential causes' above), as indicated, our next steps are:
For inadequate response to CBT/ERP monotherapy — For inadequate response to CBT/ERP, we typically augment treatment by adding an SSRI or clomipramine. In most cases we choose an SSRI; however, for individuals who have responded to another agent in the past (eg, clomipramine) and tolerated it well, we would choose the same medication. (See 'Choosing initial treatment modality' above and 'Initiating pharmacotherapy' above.)
Additionally, in more severe cases, we consider referring the individual for more intensive CBT/ERP at a higher level of care (eg, intensive outpatient program, partial hospitalization program, or residential treatment center). Intensive CBT/ERP administered several hours/day, several days/week, has been found to be effective for many individuals with severe, treatment-refractory OCD.
For inadequate response to initial pharmacologic management — For individuals with inadequate response to initial pharmacotherapy, our next steps are increasing the SSRI to a supratherapeutic dose, then augmenting with CBT/ERP if available and if needed. Some providers encourage doing both at the same time.
Increasing the SSRI medication to supratherapeutic dose — After addressing potential causes of inadequate response to SSRI medication (see 'Potential causes' above), we gradually increase the dose to a supratherapeutic dose (often well above the manufacturer’s maximum recommended dose), as tolerated. However, this approach is not recommended for citalopram or the tricyclic antidepressant clomipramine due to the potential to cause QTc prolongation (table 1) [46,47]. We are cautious when doing this in older patients (age >65). (See 'Subsequent treatment' below.)
For example, for an individual who has tolerated fluoxetine for a total of 12 weeks, with four to six of these weeks at a dose of 80 mg/day, we would attempt to increase to 100 mg/day for approximately four weeks and then to 120 mg/day (if necessary and tolerated). Alternatively, titration could be done in smaller increments (eg, 10 mg/day of fluoxetine) when dosing in the supratherapeutic range. We monitor response and side effects at the maximum dose reached in the supratherapeutic range for a minimum of six to eight weeks before determining whether it is effective or ineffective.
Data supporting the efficacy of higher than normal dosing of SSRIs are limited [29,48], although clinical experience supports this approach, as does the American Psychiatric Association practice guideline for OCD and a review by experts in pharmacotherapy for OCD [49]. One trial randomly assigned patients with OCD who had not responded to 16 weeks of sertraline (titrated from 50 mg/day to 200 mg/day) to continue sertraline at 200 mg/day or to continue at an increased dose (averaging 357 mg/day and up to 400 mg/day) [48]. After 12 weeks of treatment, the higher dose group showed significantly greater average improvement than the lower dose group. The difference, on average, was clinically modest, and some patients in both groups continued to have clinically significant symptoms; however, in our clinical experience a supratherapeutic dose can be substantially more effective than a lower dose. Rates of adverse events were similar in the two groups.
Augmentation with CBT/ERP — In all individuals with inadequate response to medication we encourage augmentation with CBT/ERP for OCD, if available. A detailed discussion of psychotherapy including CBT/ERP for OCD can be found elsewhere. (See 'Choosing initial treatment modality' above and "Obsessive-compulsive disorder in adults: Psychotherapy".)
Randomized trials have demonstrated the benefit of adding CBT/ERP to medication treatment of patients with OCD who have partially responded to a serotonergic antidepressant [15-17]. As an example, in one trial, 108 individuals with a partial response to 12 weeks of treatment with an SSRI or clomipramine continued the medication and were randomly assigned to ERP or stress management training (17 90-minute sessions over eight weeks) [16]. After eight weeks, patients in the ERP plus medication group, as compared with the stress management plus medication group, experienced a greater decrease in OCD symptoms (mean decline in Yale-Brown Obsessive Compulsive Scale [Y-BOCS] score of 11.2 versus 3.6, respectively), a higher rate of treatment response (74 versus 22 percent), and a greater proportion with mild to minimal symptoms (33 versus 4 percent). In another trial, 100 individuals with OCD who had obtained some benefit from at least 12 weeks of a serotonin reuptake inhibitor treatment were randomly assigned to receive eight weeks of the addition of ERP, risperidone, or pill placebo to the serotonin reuptake inhibitor [17]. A greater percentage of individuals in the ERP group responded to treatment (Y-BOCS score decrease ≥25 percent) than those in the risperidone or placebo groups (80 versus 23 versus 15 percent). Additionally, a greater percentage of patients in the ERP group reached remission (Y-BOCS score of ≤12) as compared to the risperidone or placebo group (43 versus 13 versus 5 percent).
SUBSEQUENT TREATMENT —
Our subsequent pharmacotherapy for individuals who have not responded to treatment to this point (initial management, addressing potential causes, adding/intensifying cognitive-behavioral therapy using exposure and response prevention [CBT/ERP], increasing selective serotonin reuptake inhibitors [SSRIs] to supratherapeutic doses) is informed, in part, by the level of response to treatment (eg, minimal/no response or partial response) (algorithm 1).
However, as minimal data support switching versus augmentation of an ineffective medication in the treatment of obsessive-compulsive disorder (OCD), clinical judgment, history, and patient preference are factors at all decision points. (See 'Assessment of response and next steps' above and 'Inadequate response to initial treatment' above.)
Minimal or nonresponse — For individuals with minimal or no improvement of OCD symptoms or in those who could not tolerate the side effect of the medication, we typically switch to a second SSRI with a different side effect profile. However, we are extremely cautious about switching to a second SSRI in individuals with co-occurring disorders (eg, depression, anxiety) or suicidality that has been effectively treated with the SSRI.
The choice between augmentation versus switching medications is based on clinical experience and patient preference.
●Change to a different SSRI – Our preferred method is to cross-titrate from the initial SSRI to a new medication. This can usually be accomplished over one to two weeks and may avoid side effects or discontinuation symptoms from particular SSRIs. Discussion of switching antidepressant medications in adults is discussed elsewhere. (See "Switching antidepressant medications in adults", section on 'Between SSRIs'.)
It has been estimated that less than one-half of patients will benefit from switching from one SSRI to another, and the likelihood of response diminishes as the number of failed adequate trials increases [38].
If switching to a second SSRI is ineffective, we switch to clomipramine. While clomipramine is an effective treatment for OCD, some patients may be reluctant to try it due to possible side effects and need for laboratory and electrocardiogram (ECG) monitoring.
●Switch to clomipramine – We treat individuals who have not responded to at least two therapeutic trials of SSRIs with clomipramine (see 'Therapeutic trial' above). Clomipramine appears to have similar efficacy to SSRIs in the treatment of OCD [3,9,19-25]. (See "Tricyclic and tetracyclic drugs: Pharmacology, administration, and side effects".)
•Pretreatment assessment – Prior to starting treatment with clomipramine, we review the individual’s past medical history with attention to a prior history of heart disease (including congenital or acquired long QT syndrome or conduction system disease).
A detailed discussion about baseline testing and periodic monitoring for safety in individuals being treated with clomipramine can be found elsewhere. (See "Tricyclic and tetracyclic drugs: Pharmacology, administration, and side effects", section on 'Baseline testing and monitoring for safety'.)
•Starting and titrating clomipramine – We typically start clomipramine at 25 mg every night at bedtime. If tolerated, we then increase by 25 mg every three to four days to a target dose of 150 to 250 mg nightly. We check levels of clomipramine and its active metabolite, desmethylclomipramine, when we reach 100 mg and then at 200 mg total daily dose. We keep combined plasma levels below 500 mg/dL to minimize the risk of cardiac conduction delay or seizures. We check an ECG before beginning clomipramine and upon reaching the maximum dose.
Although clomipramine is often well tolerated, particularly by younger patients, it is associated with side effects including anticholinergic effects (eg, dry mouth, constipation, urinary hesitancy), sedation, weight gain, cardiac conduction delay, and lowering of the seizure threshold. Seizures can occur in individuals treated with higher than maximum doses and in those with previous seizure disorder or underlying condition predisposing to a seizure disorder. (See "Tricyclic and tetracyclic drugs: Pharmacology, administration, and side effects", section on 'Seizures'.)
•Efficacy – Meta-analyses and randomized trials have found clomipramine to be superior to placebo in the treatment of OCD [3,50-53]. In a meta-analysis of seven trials including 392 individuals with OCD, treatment with clomipramine led to greater improvement as measured by a decrease in the 40-point Yale-Brown Obsessive Compulsive Scale (Y-BOCS) compared with placebo (pooled difference -8.2, 95% CI -10.5 to -5.9) [3]. Additional evidence comes from a collaboration of two studies totaling 520 individuals with OCD. In each of these studies, individuals treated with clomipramine had greater declines in Y-BOCS scores compared with individuals treated with placebo (38 versus 3 percent in one study and 44 versus 5 percent in the other) and higher rates of response with clomipramine compared with placebo (51 versus 8 percent and 60 versus 7 percent) [52].
Subsequent medication trials for continued poor or minimal response to treatment include the following:
●Switch to another SSRI or venlafaxine – In individuals who do not respond to or do not tolerate treatment with clomipramine, a third SSRI or the SNRI venlafaxine can be tried.
Limited data support the use of SNRIs in the treatment of OCD [54-59].
In a trial, 150 individuals with OCD were randomly assigned to receive either venlafaxine 300 mg daily or paroxetine 60 mg daily [55]. Similar improvements (as measured by the mean decrease in the Y-BOCS) and response rate (>35 percent decrease in Y-BOCS) were noted for each. Additionally, the incidence of adverse effects was comparable. A small trial of 47 individuals with OCD compared venlafaxine 350 mg daily with clomipramine up to 225 mg daily. No significant difference in efficacy was found, but those treated with venlafaxine had fewer side effects [56]. However, the response rate to venlafaxine was low, and the lack of a significant difference between medications could reflect type II error. Larger, placebo-controlled trials are needed to establish the efficacy of venlafaxine in the treatment of OCD.
Duloxetine, another SNRI, has shown promise in case reports and open trials [60]. In one randomized trial, it has been shown to be as effective as sertraline as an agent of augmentation of current anti-OCD treatment for individuals with refractory OCD, although this study had multiple methodologic limitations [61].
Side effects associated with SNRIs are discussed elsewhere. (See "Serotonin-norepinephrine reuptake inhibitors: Pharmacology, administration, and side effects".)
Partial response — For individuals with a partial response (ie, continue to have clinically significant symptoms of OCD) to treatment despite addressing potential causes and adjusting treatment (eg, combined psychotherapy and supratherapeutic dosing), we typically augment the SSRI with another medication. (See 'Potential causes' above and 'Adjusting treatment' above.)
However, switching medication to another agent is a reasonable alternative choice (eg, to avoid polypharmacy) at all decision points. Strategies for switching medication are discussed above. (See 'Minimal or nonresponse' above.)
When augmentation of initial agent is the chosen next step, our choice of agent is based on several factors, including the quality of the evidence in support of the treatment, potential side effects of the medication, past history, and patient preference. For most patients, we augment with antipsychotic medication as a first option and, if not effective, we typically try augmentation medications sequentially (as listed by bulleted section directly below). Atypical antipsychotics are the class of medications that have the strongest evidence of efficacy for refractory OCD [9,11,62,63].
●Antipsychotic augmentation – In most cases, a second-generation antipsychotic is our first choice in augmentation. Our preference is based on this strategy having the strongest evidence of efficacy from randomized trials [62-72] and their relatively favorable side effect profile, including a lower likelihood of extrapyramidal symptoms than first-generation antipsychotics. (See "Second-generation and other antipsychotic medications: Pharmacology, administration, and side effects".)
•Starting and titrating antipsychotic medication – We typically start with a very low dose of an atypical antipsychotic and titrate as tolerated. As an example, we often start augmentation with aripiprazole at 1 or 2 mg/day and gradually titrate to 5 to 10 mg/day over several months. When using risperidone, we begin at 0.5 mg/day and gradually titrate to 2 to 4 mg over several months.
Effective target doses of adjunctive antipsychotics are generally lower than when these medications are used for treatment of psychotic disorders or bipolar disorder. For example, doses of adjunctive aripiprazole of 5 to 10 mg/day, risperidone 2 to 4 mg/day, and olanzapine 5 to 10 mg/day have been found to be effective for serotonin reuptake inhibitor-refractory OCD in placebo-controlled trials [65,66,68].
For individuals who respond to antipsychotic augmentation, we typically continue the medication for at least 12 months before attempting a gradual taper over several months. If symptoms recur or worsen during the taper, we increase the medication back to the dose that was effective and continue for another six months to one year or even longer, especially if the relapse was severe, provided that the individual is not experiencing long-term adverse effects to the antipsychotic (eg, extrapyramidal symptoms, tardive dyskinesia, significant weight gain, metabolic abnormalities, sedation). In individuals with evidence of extrapyramidal symptoms, dyskinesias, or metabolic dysregulation, we consider other options for augmentation.
If the individual does not show a clear and clinically significant benefit to adjunctive treatment with a low-dose atypical antipsychotic, the dose can be increased as tolerated. If higher doses are not tolerated or are ineffective, it may then be useful to switch to a different adjunctive antipsychotic. We are cautious to keep antipsychotic exposure to a minimum to avoid side effects such as tardive dyskinesia and metabolic effects. (See "Psychosis in adults: Initial management", section on 'Antipsychotic therapy'.)
•Efficacy – Meta-analyses and individual trials investigating augmentation methods for individuals with OCD treated with serotonin reuptake inhibitors have generally shown efficacy for adjunctive antipsychotics [62-77]. However, not all trials have supported this [78,79].
In a meta-analysis of 14 trials, 491 individuals with OCD were treated with first- or second-generation antipsychotics as augmentation (quetiapine, risperidone, aripiprazole, olanzapine, paliperidone, haloperidol) versus placebo [71]. The primary outcome was mean change in the Y-BOCS total score. Pooled effect size for antipsychotic treatment was superior to placebo (Hedges’ g -0.64, 95% CI -0.87 to -0.41). Aripiprazole, risperidone, and haloperidol showed the strongest evidence for effect compared with placebo; however, the results for haloperidol are based on one trial only. Olanzapine, paliperidone, and quetiapine showed similar effect to placebo. Furthermore, response rate (defined as Y-BOCS reduction ≥35 percent) was higher in the treatment versus placebo group (pooled response rate 30 versus 13 percent). Methodologic differences limit the interpretation of data supporting antipsychotic augmentation of SSRIs for OCD. However, a systematic review concludes that augmentation with some antipsychotics is supported by clinical trials, and that antipsychotics as a group appear to benefit approximately one-third of individuals with serotonin reuptake inhibitor-refractory OCD [62].
Efficacy trials of antipsychotic augmentation for OCD have generally examined short-term use only. In a follow-up to a clinical trial (discussed above) that compared addition to an SSRI of risperidone, exposure and response prevention (ERP), or pill placebo in patients with OCD who had not remitted on SSRIs, the eight individuals who acutely responded to risperidone and remained on risperidone maintained their treatment gains out to six months [73]. In one small retrospective study, 13 out of 15 patients had a return of OCD symptoms within two months of their antipsychotic being stopped [74]. It is unclear whether varying responses found across antipsychotic trials reflect true differences in efficacy or methodologic issues specific to the trials (eg, differences in the proportion of individuals with comorbid tic disorder, as they may be more likely to respond to antipsychotic augmentation of an SSRI).
Despite evidence supporting antipsychotic augmentation for individuals with partial response to SSRIs, some practitioners may be reluctant to begin treatment with an antipsychotic due to potential metabolic dysregulation and other potential side effects. Treatment with other augmenting agents may be an acceptable alternative to antipsychotic agents in individuals with existing metabolic disturbances [38].
Side effects, monitoring, and other considerations when prescribing antipsychotic medications are discussed elsewhere. (See "Second-generation and other antipsychotic medications: Pharmacology, administration, and side effects" and "First-generation antipsychotic medications: Pharmacology, administration, and comparative side effects".)
●Augment with clomipramine – In individuals with partial response to initial treatment with an SSRI, we sometimes augment with clomipramine (usually up to 50 to 150 mg/day) as a next choice.
When we use this approach, we typically add 25 mg nightly of clomipramine to the SSRI and increase by 25 mg per week to 50 to 150 mg nightly, as tolerated. We check an ECG prior to beginning clomipramine and upon reaching the maximum dose of clomipramine. We also check plasma levels of clomipramine upon reaching 50 or 75 mg/day and after further dose increases because SSRIs can substantially increase levels of clomipramine, which has a low therapeutic index. Combined plasma levels of clomipramine and desmethylclomipramine should be below 500 mg/dL. Clomipramine doses are usually lower when clomipramine is combined with an SSRI than when it is prescribed as monotherapy. (See "Tricyclic and tetracyclic drugs: Pharmacology, administration, and side effects", section on 'Baseline testing and monitoring for safety'.)
Augmentation of SSRIs with clomipramine has been found to be effective for treatment of OCD in case series and open trials [76,80,81]. Combining clomipramine with some SSRI medications increases the risk of serotonin syndrome, QTc prolongation, and other side effects due to each interacting with cytochrome p450 metabolic enzymes and subsequent increasing levels of each. Symptoms of toxicity include tremor, tachycardia, myoclonic jerks, QTc prolongation, seizures, and serotonin overstimulation. (See "Serotonin syndrome (serotonin toxicity)" and "Acquired long QT syndrome: Clinical manifestations, diagnosis, and management".)
●Augmentation with buspirone or a glutamate modulator or antagonist – Controlled trials of buspirone as an serotonin reuptake inhibitor augmenting agent have not shown clear benefit, but some experts recommend it [49]. Evidence for efficacy of glutamate modulators or antagonists as serotonin reuptake inhibitor augmenters is also limited. Our preference from among glutamate modulators/antagonists is based on the quality of the evidence in support of the agent, efficacy for comorbid disorders (eg, bipolar disorder), avoidance of side effects from antipsychotic medication (eg, weight gain, metabolic syndrome, extrapyramidal symptoms), and familiarity and with comfort using the agent. For example, for a patient with OCD and comorbid bipolar disorder, we might choose adjunctive lamotrigine. For a participant with comorbid binge eating or migraine disorder, we might choose adjunctive topiramate.
Altered glutamate neurotransmission in cortico-striatal-pallidal-thalamic brain circuits may contribute to symptoms of OCD. In some but not all studies, glutamate modulators or inhibitors appear to be effective as serotonin reuptake inhibitor augmenting agents for those individuals with OCD without sufficient response to serotonin reuptake inhibitors. However, the studies are small and further research is warranted.
As examples:
•Lamotrigine [82]
•Topiramate [86,87]
•Pregabalin [88]
•Amantadine [89]
●Pharmacologic options with limited supporting data – We use the following agents only after inadequate response to the above treatments. We seek expert consultation in the management of individuals who are refractory to treatment or in whom we are considering use of the following agents.
These agents have shown some promise in the treatment of OCD as monotherapy or augmentation. However, data are limited to small, randomized trials, case reports, or open trials. Further trials with greater numbers of individuals are needed to establish efficacy.
•Vortioxetine [96]
•Stimulants (ie, caffeine, dextroamphetamine, methylphenidate ER) [97-101]
•Anti-inflammatory agents (ie, celecoxib [102], tramadol [103], minocycline [104,105]; risk of misuse and possible serotonin syndrome should be considered with tramadol)
•D-Cycloserine [106-109]
•5-HT3 antagonists (ie, ondansetron, granisetron, tropisetron [112-115])
Surgical and neuromodulatory treatment — Surgical intervention is used only when the individual has been refractory to pharmacologic management and CBT/ERP and still presents with significant distress or reduced quality of life. Discussion of deep brain stimulation, transcranial magnetic stimulation, and other surgical and neuromodulatory treatments for individuals with treatment refractory OCD is found elsewhere. (See "Deep brain stimulation for treatment of obsessive-compulsive disorder".)
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Obsessive-compulsive disorder and related disorders".)
SUMMARY AND RECOMMENDATIONS
●Choosing initial treatment modality – Initial treatment of obsessive-compulsive disorder (OCD) may consist of a selective serotonin reuptake inhibitor (SSRI), cognitive-behavioral therapy using exposure and response prevention (CBT/ERP), or the combination. We consider many factors such as severity of symptoms, presence of suicidality, presence of co-occurring disorders, history, willingness and ability of the patient to participate in CBT/ERP, availability of CBT/ERP, and patient preference in choosing initial treatment modality. (See 'Choosing initial treatment modality' above.)
•Mild to moderate symptoms – For individuals with mild to moderate symptoms, we suggest initiating treatment with CBT/ERP rather than pharmacotherapy or combined therapy (Grade 2C). CBT/ERP monotherapy does not expose the patient to medication side effects and may be somewhat more effective than pharmacotherapy. (See 'Choosing initial treatment modality' above.)
However, pharmacotherapy is a reasonable alternative for patients who are unwilling or unable to participate in CBT/ERP or who have been previously unresponsive to CBT/ERP. It is also reasonable to initiate treatment with pharmacotherapy if the patient has responded to a medication in the past.
Pharmacotherapy (with or without CBT/ERP) is often preferred as initial treatment for patients who have a co-occurring psychiatric disorder that is typically responsive to SSRIs. (See 'Choosing initial treatment modality' above.)
Combined treatment is a reasonable alternative for patients who prefer this approach. (See 'Choosing initial treatment modality' above.)
•Severe symptoms or suicidality – For individuals with severe symptoms or suicidality we suggest initiating treatment with combined modality including CBT/ERP and pharmacotherapy rather than either treatment alone (Grade 2C). Combined treatment is more effective than pharmacotherapy or CBT/ERP alone. (See 'Choosing initial treatment modality' above.)
●Pharmacotherapy – For individuals who will be starting treatment with pharmacotherapy either with or without CBT/ERP, we suggest initial treatment with an SSRI rather than other medications (Grade 2C). While there is limited comparative efficacy data, SSRIs have the most robust evidence of efficacy, are easy to use, and have limited side effects (algorithm 1). (See 'Initiating pharmacotherapy' above.)
●Therapeutic trial – We consider a therapeutic trial of medication to be at least 12 weeks, with at least six to eight of these weeks at the target dose for OCD, if needed. We explain this to the individual so that they do not prematurely stop the medication before a therapeutic trial has been completed (table 1). (See 'Therapeutic trial' above.)
●Response to initial treatment – For individuals with inadequate response to an initial therapeutic trial, we first address potential causes, including poor adherence to medication, fidelity of exposure and response prevention (ERP), and inaccurate diagnosis or the presence of undiagnosed co-occurring disorders. (See 'Potential causes' above.)
After addressing potential causes, if symptoms persist, our preference is combined modality with CBT/ERP and SSRI including further titration of the SSRI to supratherapeutic dosing and increasing intensity of CBT/ERP, as tolerated. (See 'Augmentation with CBT/ERP' above and 'Increasing the SSRI medication to supratherapeutic dose' above.)
●Subsequent pharmacotherapy options – Our preference is to try augmentation methods for those with partial response and changing agents for no response. However, minimal data support one method over another in individuals with inadequate response to initial treatment for OCD. Our decision is based on patient preference, history of response to medications, and clinical judgment.
•Patients with minimal to no response – For individuals with minimal or no response to pharmacotherapy (with or without CBT/ERP), we often switch to a different SSRI. If this is ineffective, we switch to clomipramine. Subsequent trials include a third SSRI or venlafaxine. However, augmentation of the ineffective agent is an alternative option. (See 'Partial response' above and 'Minimal or nonresponse' above.)
•Patients with partial response – For individuals who have had partial response to treatment, we typically augment the SSRI with a second-generation antipsychotic. Other options include augmentation with clomipramine, buspirone, or a glutamate modulator or inhibitor. However, changing the agent to another SSRI is a reasonable alternative option. (See 'Minimal or nonresponse' above and 'Partial response' above.)
ACKNOWLEDGMENTS —
The UpToDate editorial staff acknowledges Sanjaya Saxena, MD, DFAPA, FACNP, and Helen Blair Simpson, MD, PhD, who contributed to earlier versions of this topic review.
