*4.2. Timing*

The same two studies used di fferent criteria for starting CBT. Denys et al. started CBT after partial response, defined as a substantial reduction of on average 6 points on the Y-BOCS, without further decrease during three consecutive visits [13,14]. In the study by Tyagi, CBT was started after 36 weeks in every patient, irrespective of the amount of improvement achieved by DBS [15]. In clinical practice the question of when to best start CBT is an important one. It does not seem sensible to start CBT immediately postoperatively after electrode implantation and activating the DBS system. The mental state of the patient has not changed yet, and because of that there is no reason to expect that treatments that were ine ffective before DBS would now be e ffective. It also makes no sense to start CBT if the response to DBS is very large, since there may not be any relevant treatment goals left to work towards. The best time to start CBT is probably when there is a partial response to DBS. The altered mental state of the patient, with not only some reduction in obsessive-compulsive behavior, but usually also reduced anxiety and improved mood, provides a di fferent starting position for CBT, and the patient may be more able and motivated to comply with therapy, as was also described by Greenberg et al., and by Van Westen et al. in their qualitative study [16,17]. The question then is when to start CBT in

case of partial response? Some clinicians routinely start after a certain period (e.g., after 8 or 12 weeks). Other clinicians start CBT when it is assumed that optimal stimulation settings have been achieved. Whereas this may be preferable in a research context, in order to separate the di fferential contribution of DBS and CBT to the response, clinically this is debatable. On one hand, reaching optimal stimulation settings may take a long time in many patients, which would lead to an unacceptable delay for therapy and loss of momentum; on the other hand, these patients have already experienced non-e ffective CBT and it is important to spare them another failure because of starting CBT to soon, as this would decrease their motivation for another attempt when stimulation parameters are optimal. One option is to assess the "readiness" for CBT, as mentioned above. Another option may be to look for improvement of cognitive measures that may increase the likelihood of successful CBT. A recent intervention study showed that the e ffect of VC/VS DBS is explained in part by enhancement of cognitive control by the prefrontal cortex. In this study, DBS improved the patients' performance on a cognitive control task and increases theta (5–8 Hz) oscillations in both medial and lateral PFC, which predicts the clinical outcome [25]. Perhaps such indicators could be made to clinical use and help to indicate the best time to start CBT after DBS.
