Barlow, Gorman, Shear, & Woods, 2000

Design: Randomly assigned 312 clients with panic disorder (PD) to either CBT, imipramine (IMI), placebo, CBT + placebo, or CBT + IMI groups. Clients in either placebo or IMI alone received 30 min medical management sessions. Clients in CBT alone received 50 min. sessions. Clients in combined treatment groups received 75 min sessions. The study was conducted in 3 phases: acute, maintenance, and follow-up. Acute treatment was conducted weekly over 3 mo. Responders to acute treatment received monthly sessions that continued over the 6 mo. maintenance phase. Most responders to maintenance treatment were assigned to a discontinuation of treatment, and were re-evaluated at a follow-up 6 additional months after maintenance treatment (15 mo. after the start of treatment).

Measures: Clients were primarily evaluated on the Panic Disorder Severity Scale (PDSS), a clinician rating. “Improvement” on the PDSS was dependent on a 40% reduction from baseline. The Clinical Global Impression scale (CGI) (comprised of 7 pt. Ratings on severity and overall improvement) was also used to evaluate “Responder” status. To be a “responder” the client had to achieve a score of 2 (much improved) or better while being rated as 3 (mild) or less on CGI severity. Clients were predominantly white (90%), female (62%), and 27% of clients had co-morbid major depression. Intent to Treat (ITT) analyses were conducted on all clients entering a phase of treatment, thus ITT included dropouts whose baseline scores were carried forward in the analyses. Intent to Continue (ITC) analyses included those clients that had completed the phase of treatment.

Results: Acute completion rates did not differ across treatments; however only 33% of placebo clients completed maintenance versus 82-90% of the rest of maintenance clients assigned to active conditions. CBT alone and IMI alone were superior to placebo in both the acute and maintenance ITT analyses. Both treatments had sig. Fewer dropouts for lack of efficacy than did the placebo group. IMI had significantly greater dropouts for adverse effects than the placebo group. There were no differences in improvement between clients treated with CBT and IMI, although fewer clients dropped out of treatment for adverse effects in IMI than CBT. Follow-up analyses show trends p < .1 favoring CBT over IMI. The combined treatment (CBT + IMI) was superior to all relevant single treatment comparisons only on 1 of 3 ITT analyses and 1 of 3 completer analyses. In the maintenance ITT analysis the combined treatment was better than all 3 comparison treatments on the PDSS (meeting the authors criteria for superiority of treatment), and better than CBT and CBT + placebo on the ITC analyses.

Responders to IMI, with or without CBT, did significantly worse during no-treatment follow-up than those getting prior CBT alone or prior CBT + placebo. In secondary analyses restricted to responders only, IMI responders had lower PDSS scores than CBT alone responders; indicating a higher quality of response to IMI. Also, IMI + CBT responders had lower PDSS scores than CBT only responders.

Conclusions: Both IMI and CBT > placebo in treatment of PD. IMI produces a higher quality of response, but CBT (4% relapse during maintenance) than was somewhat more durable than IMI (25% relapse) in its effects and CBT was also better tolerated. PDSS was a more sensitive measure than CGI. There were limited benefits to combined treatment during acute and maintenance phases, and CBT + IMI was never more effective than CBT + placebo. Although CBT + IMI produced the lowest PDSS scores at the end of maintenance; this combination produced the highest relapse rate at follow-up. That is, adding CBT to IMI did not improve follow up responding compared to IMI alone. The authors have a hard time accounting for this. At the end the authors comment that their study was conducted before the use of SSRIs as a “first-line” treatment (rather than tricyclics) for PD. Research is cited showing no consistent superiority of SSRIs over tricyclics in treating PD.