Obsessive-Compulsive Disorder: Duration and Discontinuation of Treatment

Duration and Discontinuation of Treatment

Little systematic research has been performed to guide decisions about continuation, maintenance and discontinuation of treatment. The largest study of extended pharmacologic treatment involved fluoxetine (Tollefson et al., 1994b). In this study, 70 patients who had responded to fluoxetine and 198 patients who had not responded during an acute 13-week trial were given the opportunity to continue on medication for another 6 months. At the end of the 6 months, 74.3% of those who had responded initially experienced further improvement. In the nonresponder group, 91.7% experienced a decrease in symptoms when the medication dose was increased from the previous unsuccessful dose. Only 19% of patients (N − 13) experienced a significant worsening of symptoms during the follow-up period. Similar to previous reports on fluoxetine, this study suggested that symp-tom improvement was maintained over time. Even more impor-tant, further improvement occurred in responders with longer treatment and in nonresponders with continued treatment at higher doses.

In another study, 85 patients who had been treated with a variety of antiobsessional medications were followed up 1 to 3.5 years after initial treatment (Orloff et al., 1994). Ninety-four percent of the patients were still taking medication, and 87% had maintained previous gains or achieved further improvement. Thus, from a clinical point of view, it seems wise to continue medication for an extended period, perhaps for 6 months to a year after initiating treatment, because during this period im-provement is maintained and some patients experience further improvement. Overall, this extended duration of treatment did not result in worsening side effects; in fact, in most cases pa-tients habituated to side effects. In general, the most bothersome side effects that persist with SRI treatment appear to be fatigue, weight gain and sexual dysfunction.

In recent years a number of studies of long-term efficacy have emerged. One study that included a follow-up period of 2 years for 38 OCD patients on sertraline showed continued ef-ficacy with fewer side effects with longer-term treatment. An-other study assessed a larger group of patients and attempted to answer important clinical questions not only about long-term efficacy but about the effects of medication discontinuation after 1 year of treatment. The latter is an important question because few studies of systematic discontinuation have been carried out and in those that have been done relapse rates were quite high, above 90%. This sertraline study (Koran et al., 2002) involved 223 patients who had been successfully treated with single-blind sertraline for 52 weeks who were then ran-domized in a double-blind manner to continue treatment for another 6 months or placebo. One-third of the patients in the placebo group relapsed; this was surprisingly lower than the percentage found in earlier studies. The authors offered several plausible explanations for this, which included the possibility that 1 year of effective treatment may provide sustained benefit for patients, and that patients may have engaged in self-directed behavior therapy, something that was not readily available at the time of the previous discontinuation studies. They also noted that while OCD symptom ratings did not worsen in the placebo-treated group as a whole, quality of life did significantly dete-riorate. This finding points to the need for more sensitive meas-ures of patient improvement and for further studies of long-term treatment efficacy.

Some preliminary data have suggested that in treatment re-sponders it may be possible to decrease the dose of clomipramine over the longer term without subsequent relapse, although this important issue needs further investigation.

The data on discontinuation of behavioral therapy are also encouraging and overall, about 75% of patients continue to do well at follow-up. However, most studies also noted that few pa-tients were symptom-free.