What do we know about the relative efficacy of current medication and [cognitive-behavioral] therapy (CBT) for OCD? Dr H. Blair Simpson answers that question and discusses future directions and causes in the study of OCD.
Dr Simpson is funded by the National Institute of Mental Health (NIMH).
In collaboration with Dr Edna Foa at the University of Pennsylvania, our clinic did a study to exactly answer the question. The data from that study showed very clearly that the therapy—when delivered by highly trained therapists and well supervised with patients who adhere to the treatment—is at least as good if not better than the medication. So that is surprising for psychiatrists to know. That's why either the medication or the therapy are first-line treatments for OCD.
Many people make the assumption [and ask] what happens when you put [medication and therapy] together [and wonder if] that [would be] better still. The interesting thing about that first study is that's not what we found. We found that the medication and the therapy together were no better than the therapy alone. There are a couple of important caveats. In that study, the medication and the therapy were started at the same time [and medication can take up to 6 weeks to work]. The therapy in that study was delivered not twice weekly, but 5 days a week for 3 weeks within a week of home visits so the therapy [was intensive over a month's duration]. So that combination cell in essence was like a month of intensive therapy and the ["rocket booster" at the end] was medication. So what I would say is that no, you do not need to start people . . . all people do not need to start on combination. They can start with medication, or start with the therapy.
If people get a partial response to medication, we have now done 2 studies, again in collaboration with Dr Edna Foa at the University of Pennsylvania that show definitively that adding CBT on top of people who have gotten the most that they are going to get out of their medication will help more people get well. In those subsequent studies, everyone came in on a stable dose of their serotonin reuptake inhibitor (SRI) medication. They had been on it for at least 12 weeks on the maximum dose. They weren't going to get anything more from their medication.
One study randomized them to exposure and ritual prevention (ERP) treatment versus a control therapy where we were able to control for attention and support and coming to our clinic with a research staff. We could show clearly that the control therapy didn't do much. By adding ERP treatment on top of the medication, many people (70%) responded and a third got well. That study was comparing adding exposure therapy versus adding antipsychotic medication, which is right now the only other treatment we have for people who have a partial response to SRI medication. Again, people came in on their SRI medication; all of them had been in on their highest dose for at least 12 weeks and they weren't going to get anything more from their SRI. This is the situation that psychiatrists are going to [encounter].
The question becomes what do you do? This was a study where we compared adding ERP treatment versus adding antipsychotic medication—in the study we used risperidone because the evidence was best for risperidone in OCD—versus a placebo. What we found in that study is that once again with ERP treatment 80% of people responded and 43% were well after 17 sessions. Moreover, we have also shown that if you get well, you are very likely to maintain your response out to 6 months if you have gotten well from ERP.
The big picture is there are SRI medications as a first-line treatment. There is this very particular type of CBT as a first-line treatment. When you get a partial response to SRI medication, the first thing you should be thinking about is adding the CBT on top. With all the caveats I've given you, that if you have a patient who cannot adhere to the treatment, then you're not going to get the good results that we can get in our studies. If the therapist isn't giving the right treatment, you're not going to get the good results that you get in our studies. That's where as a psychiatrist, once you've tried those options, you [use treatments] that are less evidence-based to try to help your patients.
We [are] interested in who the people are who get well once they have been given an SRI medication and then CBT/ERP added on top? As I said, about 70% to 80% of people will respond to that combination and about 35% to 40% of people will get well at the end of that added-on 17 sessions. And if you give some people a little bit more, maybe as high as 60% will get well.
Dr Simpson is Professor of Clinical Psychiatry, Columbia University, and Director of the Anxiety Disorders Clinic and the Center for OCD and Related Disorders at New York State Psychiatric Institute in New York City,