Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Cognitive therapy and panic disorder


We usually test whether interventions work or not in randomised trials. We call this a test of efficacy. Less usual is to see an intervention tested as part of a usual package of care, mimicking its use in a real world situation. This is called a test of effectiveness. Knowing whether an intervention works in practice is much more important than knowing that it works in a trial, so a study demonstrating the value of cognitive behavioural therapy in a real world setting [1] is particularly welcome.


University-affiliated primary care clinics in the western USA formed the setting for the study. Patients were those who were aged 18 to 70 years, fulfilled standard criteria for diagnosis of panic disorder. They had to have at least one panic attack in the previous week, and be willing to accept a combined treatment of anti-anxiety medicine with cognitive behavioural therapy. Exclusions were sensible.

Patients were randomised to usual care or the intervention. Those receiving usual care received whatever their primary care physician thought appropriate, usually pharmacotherapy. The physician knew the results of initial screening and diagnosis, so outcomes could not be attributed to non-recognition of panic disorder. They could be referred or could self-refer to mental health resources thought appropriate.

The intervention used non-physicians (recent masters or doctoral degrees with no cognitive therapy experience) who were trained to deliver evidence-based cognitive therapy targeting panic symptoms, as well as depressive and social anxiety symptoms. Patients received six sessions within the first three months plus six brief telephone contacts. Primary care physicians managed patients' medication, using an algorithm that used a dose titration of SSRI or alternate antidepressant.

Assessments were made by telephone questionnaire by interviewers blinded to intervention status, at baseline and at three monthly intervals to 12 months. Lots of things were measured, but the main outcomes were remission (no panic attack in last month, minimal anticipation of panic, low agoraphobia score), and response (score of less than 20 on an anxiety sensitivity index).


The 232 patients enrolled had slightly more women than men, with an average age of 41 years. Over 60% had other chronic medical conditions, with social phobia, posttraumatic stress disorder, generalised anxiety and major depression, common psychiatric conditions, in between 30% and 50% of patients. The average panic attack frequency was about 1.5 a week.

By 12 months, remission had occurred in 29% of patients with the intervention, and in 16% of those receiving usual care. The number needed to treat for one more patient receiving the intervention to be in remission at 12 months was 7.4 (95% CI 4.2 to 34).

Anxiety sensitivity index scores fell in both groups (Figure 1), but significantly more so than those with the intervention. By 12 months, 63% of those with the intervention had a score less than 20, compared with 38% in those receiving usual care. The number needed to treat for one more patient receiving the intervention to have an anxiety score below 20 at 12 months was 4.0 (95% CI 2.7 to 8.0).

Figure 1: Anxiety sensitivity scores with cognitive therapy intervention and usual care

At 12 months, depression and disability scores were also significantly lower with intervention than with control. There was no difference in the proportion of patients receiving appropriate anti-panic medication for more than six weeks.


This is a necessarily complex study excellently reported. No summary can quite do it justice. It may be that the individual contributions of medication and cognitive therapy to the better results with the intervention cannot be discerned, but does that matter?

The point is that the authors have developed a care pathway for treating a common and difficult disorder in primary care, and have shown that it can deliver better results than usual care. Moreover, young graduates with no experience of cognitive therapy delivered the intervention. Clearly this has potential for actual use in primary care by training existing staff.


  1. PP Roy-Byrne et al. A randomized effectiveness trial of cognitive-behavioural therapy and medication for primary care panic disorder. Archives of General Psychiatry 2005 62: 290-298.

previous or next story