This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
Previous studies on Internet-based treatment with minimal therapist guidance have shown promising results for several specific diagnoses.
To (1) investigate the effects of a tailored, therapist-guided, Internet-based treatment for individuals with reoccurring panic attacks, and (2) to examine whether people in different age groups (18–30 years and 31–45 years) would respond differently to the treatment.
We recruited 149 participants from an online list of individuals having expressed an interest in Internet treatment. Screening consisted of online questionnaires followed by a telephone interview. A total of 57 participants were included after a semistructured diagnostic interview, and they were randomly assigned to an 8-week treatment program (n = 29) or to a control condition (n = 28). Treatment consisted of individually prescribed cognitive behavior therapy text modules in conjunction with online therapist guidance. The control group consisted of people on a waitlist who later received treatment.
All dependent measures improved significantly immediately following treatment and at the 12-month follow-up. The between-group effect size on the primary outcome measure, the Panic Disorder Severity Scale, was
Tailoring an Internet-based treatment can be a feasible approach in the treatment of panic symptoms and comorbid anxiety and depressive symptoms. Younger adults benefit as much as adults over 30 years and up to 45 years of age.
Clinicaltrials.gov NCT01296321; http://www.clinicaltrials.gov/ct2/show/NCT01296321 (Archived by WebCite at http://www.webcitation.org/65wddsqlL)
Internet-based cognitive behavior therapy (iCBT) [
Panic attacks are common across psychiatric conditions, but far from all persons with panic attacks develop panic disorder [
We recruited participants via an online list (www.studie.nu) among individuals who had expressed an interest in participating in research on iCBT for panic disorder and generalized anxiety disorder via email. The list consisted of 1459 individuals who had been on the list for a minimum of 12,648 days, maximum 14,652 days, and mean of 13,298 days with a standard deviation of 505 days. They were presented with the project website, which contained information about the trial, how to register, and how to submit written informed consent. Screening consisted of the following questionnaires via the Internet: Montgomery-Åsberg Depression Scale-Self-rated (MADRS-S) [
Demographic description of the participants at pretreatment.
Treatment group |
Control group |
Total |
||
|
||||
Male | 8 (28%) | 12 (43%) | 20 (35%) | |
Female | 21 (72%) | 16 (57%) | 37 (65%) | |
|
||||
Mean (SD) | 32.3 (7.4) | 32.5 (6.5) | 32.4 (6.9) | |
Minimum–maximum | 20–45 | 21–44 | 20–45 | |
18–30 | 13 (45%) | 12 (43%) | 25 (44%) | |
31–45 | 16 (55%) | 16 (57%) | 32 (56%) | |
|
||||
Single without children | 6 (21%) | 6 (21%) | 12 (21%) | |
Single with children | 1 (3%) | 0 (0%) | 1 (2%) | |
Living apart without children | 1 (3%) | 2 (7%) | 3 (5%) | |
Living apart with children | 1 (3%) | 1 (4%) | 2 (4%) | |
Married/living together without children | 5 (17%) | 6 (21%) | 11 (19%) | |
Married/living together with children | 15 (52%) | 13 (46%) | 28 (49%) | |
|
||||
9-year compulsory school | 2 (7%) | 2 (7%) | 4 (7%) | |
Secondary school | 11 (38%) | 4 (14%) | 15 (26%) | |
Vocational (completed) | 1 (3%) | 3 (11%) | 4 (7%) | |
Collage/university (not completed | 5 (17%) | 7 (25%) | 12 (21%) | |
Collage/university (completed) | 10 (35%) | 12 (43%) | 22 (39%) | |
|
||||
No experience | 10 (35%) | 6 (21%) | 16 (28%) | |
Previous experience | 19 (66%) | 22 (79%) | 41 (72%) | |
|
||||
Ongoing | 15 (52%) | 12 (43%) | 27 (47%) | |
Completed | 4 (14%) | 5 (18%) | 9 (16%) | |
No experience | 10 (35%) | 11 (39) | 21 (37%) | |
|
||||
Self-employed | 2 (7%) | 0 (0%) | 2 (4%) | |
Employed | 14 (48%) | 14 (50%) | 28 (49%) | |
Unemployed | 0 (0%) | 4 (14%) | 4 (7%) | |
Student | 5 (17%) | 6 (21%) | 11 (19%) | |
On parental leave | 4 (14%) | 2 (7%) | 6 (11%) | |
Sick leave | 4 (14%) | 2 (7%) | 6 (11%) | |
|
||||
Panic disorder | 2 (7%) | 2 (7%) | 4 (7%) | |
Panic disorder + agoraphobia | 24 (83%) | 23 (82%) | 47 (83%) | |
Generalized anxiety disorder | 2 (7%) | 9 (32%) | 11 (19%) | |
Social phobia | 1 (3%) | 8 (29%) | 9 (16%) | |
Anxiety disorder not otherwise specified | 1 (3%) | 0 (0%) | 1 (2%) | |
Major depression | 2 (7%) | 3 (11%) | 5 (9%) | |
Comorbid disorders | 5 (17%) | 13 (46%) | 18 (32%) |
a Anxiolytic and/or antidepressant. Needed to be stabilized for 3 months.
b Structured Clinical Interview for DSM-IV Axis I Disorders.
Flowchart of study participants, point of random allocation, and dropouts at each stage of the trial. MADRS-S = Montgomery-Åsberg Depression Scale; OCD = obsessive–compulsive disorder; PTSD = posttraumatic stress disorder; SCID-I = Structured Clinical Interview for DSM-IV Axis I Disorders.
The treatment consisted of 19 CBT modules derived from previous iCBT trials on panic disorder [
Number of prescribed treatment modules.
Module | Treatment group |
Control group |
Total (n = 57) |
Introduction | 29 (100%) | 28 (100%) | 57 (100%) |
Cognitive restructuring 1 | 26 (90%) | 22 (79%) | 48 (84%) |
Cognitive restructuring 2 | 26 (90%) | 22 (79%) | 48 (84%) |
Panic 1 | 29 (100%) | 28 (100%) | 57 (100%) |
Panic 2 | 29 (100%) | 28 (100%) | 57 (100%) |
Agoraphobia | 21 (72%) | 22 (79%) | 43 (75%) |
Social anxiety 1 | 1 (3%) | 3 (11%) | 4 (7%) |
Social anxiety 2 | 1 (3%) | 3 (11%) | 4 (7%) |
Generalized anxiety 1 | 3 (10%) | 6 (21%) | 9 (16%) |
Generalized anxiety 2 | 3 (10%) | 6 (21%) | 9 (16%) |
Generalized anxiety 3 | 3 (10%) | 6 (21%) | 9 (16%) |
Behavioral activation 1 | 0 (0%) | 2 (7%) | 2 (4%) |
Behavioral activation 2 | 0 (0%) | 2 (7%) | 2 (4%) |
Relaxation (applied) | 15 (52%) | 15 (54%) | 30 (53%) |
Sleep management | 1 (3%) | 1 (4%) | 2 (4%) |
Mindfulness | 3 (10%) | 2 (7%) | 5 (9%) |
Setting boundaries | 3 (10%) | 1 (4%) | 4 (7%) |
Solving problems | 1 (3%) | 0 (0%) | 1 (2%) |
Preventing relapse | 29 (100%) | 28 (100%) | 57 (100%) |
The control condition consisted of a waitlist group. Participants were informed that they would receive the treatment after 10 weeks when the treatment group had completed their treatment. They were also informed how to contact the study team if they had any questions during this time frame.
We used a randomized controlled design to compare the effects of the treatment against waiting. The participants were divided into two groups so that the two predetermined age groups 18–30 years (young adults) and 31–45 years(adults) were equally represented in each condition. The blocked randomization process was conducted through an online true random number-generation service (random.org) independent of the investigators and therapists. The therapists in this trial were not aware of the participants’ age group. The project was approved by the regional ethics committee in Linköping and registered at ClinicalTrials.gov (NCT01296321). We obtained written informed consent through regular mail, which was sent to the study team. At posttreatment participants were instructed via email to complete the follow-up questionnaires and to participate in a semistructured telephone interview carried out by a blinded assessor who had no earlier contact with the participants. The same procedure was repeated at 12 months after treatment completion. We used the PDSS as the primary outcome measure in the trial.
To examine whether the randomization process had succeeded in generating a balanced distribution across the two conditions, we used independent
Of the 29 participants in the treatment group, 7 (24%) completed all prescribed modules (6–8) within the 8-week treatment period. A total of 17 participants (59%) completed 50% of the prescribed modules and 14 (48%) completed 75% of the prescribed modules. The mean number of completed modules for the whole group was 5.0 (SD 2.6). The mean number of completed modules in the young adult group was 5.15 (SD 2.34) and the corresponding number in the adult group was 4.19 (SD 3.16). This difference was not statistically significant (
As is evident in
Immediate results with intention-to-treat analysis using mixed-effect model estimated means (ES) (n = 57), observed means (OM) (n = 50 for the main outcome measurea and n = 49 for the secondary outcome measures), and standard deviation (observed) at pre- and posttreatment for the measures of panic, anxiety, depression, and quality of life.
Measure | Time | Group | Age group | ES | OM | SD | 95% confidence interval |
PDSSa | Pre | Treatment | Total | 12.54 | 12.75 | 4.95 | 10.62–14.47 |
Young adults | 12.46 | 13.17 | 4.97 | 9.60–15.32 | |||
Adults | 12.63 | 12.33 | 5.12 | 10.05–15.21 | |||
Control | Total | 13.71 | 13.77 | 5.25 | 11.74–15.68 | ||
Young adults | 13.67 | 13.67 | 6.79 | 10.69–16.65 | |||
Adults | 13.75 | 13.86 | 3.74 | 11.17–16.33 | |||
Post | Treatment | Total | 6.39 | 6.54 | 4.97 | 4.30–8.48 | |
Young adults | 5.57 | 6.08 | 4.03 | 2.55–8.60 | |||
Adults | 7.21 | 7.00 | 5.91 | 4.33–10.09 | |||
Control | Total | 13.79 | 13.81 | 5.49 | 11.71–15.85 | ||
Young adults | 14.00 | 14.00 | 6.58 | 10.92–17.09 | |||
Adults | 13.57 | 13.64 | 4.62 | 10.80–16.33 | |||
CORE-OMb | Pre | Treatment | Total | 1.69 | 1.69 | 0.60 | 1.51–1.87 |
Young adults | 1.65 | 1.60 | 0.61 | 1.38–1.91 | |||
Adults | 1.73 | 1.78 | 0.59 | 1.49–1.97 | |||
Control | Total | 1.83 | 1.83 | 0.39 | 1.65–2.02 | ||
Young adults | 1.84 | 1.84 | 0.43 | 1.56–2.12 | |||
Adults | 1.83 | 1.82 | 0.37 | 1.59–2.07 | |||
Post | Treatment | Total | 1.14 | 1.13 | 0.56 | 0.92–1.36 | |
Young adults | 0.96 | 0.93 | 0.44 | 0.65–1.27 | |||
Adults | 1.32 | 1.36 | 0.60 | 1.01–1.63 | |||
Control | Total | 1.69 | 1.69 | 0.53 | 1.49–1.89 | ||
Young adults | 1.65 | 1.65 | 0.68 | 1.35–1.96 | |||
Adults | 1.72 | 1.72 | 0.40 | 1.46–1.99 | |||
BAIc | Pre | Treatment | Total | 28.20 | 28.43 | 10.06 | 24.94–31.46 |
Young adults | 29.15 | 28.36 | 11.64 | 24.31–34.00 | |||
Adults | 27.25 | 28.50 | 8.63 | 22.88–31.62 | |||
Control | Total | 27.97 | 27.50 | 8.91 | 24.63–31.31 | ||
Young adults | 31.25 | 31.25 | 10.15 | 26.21–36.29 | |||
Adults | 24.69 | 24.69 | 6.92 | 20.32–29.06 | |||
Post | Treatment | Total | 17.89 | 17.86 | 8.50 | 14.16–21.63 | |
Young adults | 15.21 | 14.82 | 4.69 | 9.98–20.45 | |||
Adults | 20.58 | 21.20 | 10.60 | 15.25–25.90 | |||
Control | Total | 23.03 | 23.04 | 9.38 | 19.65–26.42 | ||
Young adults | 23.00 | 23.00 | 12.09 | 17.89–28.12 | |||
Adults | 23.06 | 23.06 | 7.16 | 18.63–27.49 | |||
MADRS-Sd | Pre | Treatment | Total | 16.24 | 15.81 | 7.35 | 13.38–19.11 |
Young adults | 15.92 | 15.00 | 8.05 | 11.67–20.18 | |||
Adults | 16.56 | 16.70 | 6.80 | 12.73–20.40 | |||
Control | Total | 18.13 | 17.93 | 8.38 | 15.20–21.05 | ||
Young adults | 19.50 | 19.50 | 7.48 | 15.07–23.93 | |||
Adults | 16.75 | 16.75 | 9.06 | 12.92–20.58 | |||
Post | Treatment | Total | 11.44 | 11.10 | 6.65 | 8.34–14.54 | |
Young adults | 10.57 | 9.91 | 4.37 | 6.14–15.01 | |||
Adults | 12.30 | 12.40 | 8.57 | 7.96–16.64 | |||
Control | Total | 16.82 | 16.93 | 8.47 | 13.89–19.76 | ||
Young adults | 16.08 | 16.08 | 8.10 | 11.65–20.52 | |||
Adults | 17.56 | 17.56 | 8.94 | 13.72–21.40 | |||
QOLIe | Pre | Treatment | Total | 0.83 | 0.96 | 1.92 | 0.18–1.48 |
Young adults | 0.93 | 1.08 | 2.44 | –0.04–1.90 | |||
Adults | 0.72 | 0.83 | 1.25 | –0.15–1.60 | |||
Control | Total | 0.74 | 0.76 | 1.68 | 0.08–1.41 | ||
Young adults | 0.64 | 0.64 | 1.90 | –0.37–1.65 | |||
Adults | 0.84 | 0.84 | 1.55 | –0.03–1.72 | |||
Post | Treatment | Total | 1.56 | 1.67 | 1.47 | 0.93–2.20 | |
Young adults | 1.95 | 2.06 | 1.44 | 1.04–2.86 | |||
Adults | 1.17 | 1.24 | 1.46 | 0.28–2.05 | |||
Control | Total | 0.71 | 0.70 | 1.70 | 0.10–1.31 | ||
Young adults | 0.76 | 0.76 | 1.69 | –0.15–1.68 | |||
Adults | 0.65 | 0.65 | 1.76 | –0.14–1.44 |
a Panic Disorder Severity Scale (PDSS).
b Clinical Outcome in Routine Evaluation-Outcome Measure.
c Beck Anxiety Inventory.
d Montgomery-Åsberg Depression Rating Scale-Self-rated.
e Quality of Life Inventory.
Change in observed mean scores in the Panic Disorder Severity Scale (PDSS) for the treatment and waitlist conditions and the two age groups pre- and posttreatment.
Participants in the treatment condition were contacted 1 year posttreatment and asked to complete the measures again in order to evaluate the long-term effects. A mixed-models analysis showed significant time effects for the primary outcome measure PDSS (
In the treatment group, 16 of the 24 (67%) participants who took part in the posttreatment telephone interview fulfilled the criteria of a 40% improvement on PDSS. The corresponding number in the control condition was 3 of 27 (11%) participants. Chi-square test showed a significant difference between the two conditions (χ2
1 = 16.1,
The aim of this study was to investigate the effects of individually tailored iCBT for panic symptoms along with comorbid anxiety and depressive symptoms on young adults and adults. We found significant treatment effects for all dependent measures immediately following treatment and significant time effects at the 12-month follow-up, showing that a majority of the participants remained stable after completing their treatment. The between-group effect size on the primary outcome measure PDSS was
We found no interaction effect between age group and the treatment condition over all measures; there was, however, an interaction between time and age group for the BAI. While this could indicate that the young adults as a group are more likely to improve spontaneously in the short time frame, it is more probable that this interaction is a chance finding with somewhat lower scores in the adult control group. The effect sizes across all measures showed a tendency for larger effects among the young adults but, due to the small sample size, this requires further investigation for any conclusions to be drawn. Perhaps most important in light of the problems with adherence reported in two previous iCBT trials on social anxiety disorder in high school students [
This study has limitations. First, the prescription of treatment modules in the study may be unreliable because it was based on a structured diagnostic procedure (SCID-I) and clinical impression by relatively inexperienced clinicians. A more comprehensive clinical assessment, such as a functional analysis, may have resulted in the prescription of other modules. Worth mentioning is also that the treatment modules consisted of modules derived from diagnosis-specific trials and were mainly structured after each diagnosis (eg, panic disorder and not panic symptoms). We aim to address this in future trials on individually tailored treatment for anxiety symptoms, making the modules symptom specific and less difficult, or less specific for particular diagnoses. Second, the use of a waitlist control condition is a limitation of the present trial. The use of a passive control group means that the effect of nonspecific factors cannot be determined. The lack of a comparison group at the 12-month follow-up makes it impossible to conclude that the improvements over the follow-up period were due only to the effects of the individually tailored treatment. A third limitation is the size of the study, which in turn affects the generalizability of the results. The forth limitation is that the participants in this study had expressed an interest in iCBT for their problems and therefore the participants may have been highly motivated to undergo treatment. However, we did not measure how highly motivated the participants were. As a fifth limitation, the study was underpowered to detect age differences, and it is possible that small but clinically meaningful differences would have been detected with a larger sample. We do, however, note various clinical implications of the findings in the present trial. Individually tailored iCBT can broaden the use of iCBT to address comorbidity in anxiety disorder, allow for more flexibility, and emphasize the role of the clinician.
The tentative conclusion drawn from these results is that tailoring iCBT can be a feasible approach in the treatment of panic symptoms, comorbid anxiety, and depressive symptoms. Future trials should directly compare individually tailored iCBT for anxiety versus traditional face-to-face CBT, preferably in a clinical setting with larger samples, along with examining how different age groups respond to this treatment format.
Alcohol Use Disorders Identification Test
Beck Anxiety Inventory
Clinical Outcomes in Routine Evaluation-Outcome Measure
Internet-based cognitive behavior therapy
Montgomery-Åsberg Depression Scale-Self-rated
Panic Disorder Severity Scale
Quality of Life Inventory
Structured Clinical Interview for DSM-IV Axis I Disorders
This trial was funded by a grant from the Swedish Council for Working Life and Social Research. The authors would like to thank the participants in this trial for their involvement.
None declared.