Summary of evidence
Internet delivered therapies (self-guided or assisted) based on CBT or ERP are effective in reducing OCD symptoms in adults (level of evidence A). Evidence for other therapeutical approaches (MT, PMR, AS, CMT, ATT) or technologies (bibliotherapy, telephone, video conference) or adolescents and children is insufficient.
The results are generalizable to Finnish population and health care settings for adults.
Note: Remission rate or clinically significant change are rarely reported, and when so, approximately 1/3 of patients achieve remission or clinically significant change. The evidence for adolescent and children is very scarce. Harmful side-effects are rarely reported (<1%) as well as deterioration (1-5% of patients). Adding minimum therapist support may improve the effectivity.
ATT: Attention Training, AS: Association Splitting, CBT: Cognitive Behavioural Therapy, CMT: Competitive Memory Training, ERP: Exposure and Responsel Prevention, MT: Meridian Tapping, PMR: Progressive Muscle Relaxation,
| Reference | Study type | Population | Intervention and comparison | Outcomes | Risk of bias |
|---|---|---|---|---|---|
| Y-BOCS, Yale-Brown Obsessive Compulsive Scale
OCI-R, Obsessive Compulsive Inventory - Revised DOCS, Dimensional Obsessive-Compulsive Scale OCCWG, Obsessive-Compulsive Cognitions Working Group WSAS, Working and social adjustment scale HAMD, Hamilton Depression Rating Scale MADRS, Montgomery-Asberg Depression Rating Scale PHQ-9, Patient Health Questionnaire OBQ-44, Obsession Beliefs Questionnaire-44 WHOQOL-BREF WHO, Quality of Life Questionnaire I-8, Impulsive Behavior Scale-8 RSES, Rosenberg Self-Esteem Scale IQ-24, Insecurity Questionnaire-24 |
|||||
| 1. | MA | N=420 adults (n=389) and children (n=31) OCD-symptoms or OCD-diagnosis Australia, Sweden, UK, US, Germany and Norway |
Technology delivered CBT (including self-help books, telephone and videoconnected) vs. control (wait-list attention relaxation control, active face-to-face CBT therapy) |
OCD-symptoms:Y-BOCS, DOCS, OCI-R Depression symptoms: HAMD, MADRS, PHQ-9 |
Low risk of bias (Cochrane criteria) Syst.search on PsychINFO, ScienceDirect, Medline + hand search (incl. 23 previous syst.revs on different T-CBT) |
| 2. | MA | N=1570 Adults with OCD-symptoms or OCD-diagnosis Australia, Sweden, UK, US, Germany and Norway |
self-help interventions (bibliotherapy, computerized, and internet-based). Grouped by the amount of therapist support
|
OCD-symptoms: Y-BOCS, DOCS, OCI-R, Other: WSAS |
Low/Unclear risk of bias. (Cochrane criteria) |
| 3. | RCT | England. Adults. On-line recruitment on social media platforms for self-help for perfectionism with clinical levels of perfectionism Mean age 28.9 yrs female 82% studying 62% previous treatment for mental health disorder 39,2% current treatment 28,3% Baseline OCI-R 27.88 (13.71) |
I-CBT for perfectionism, 8 modules, weekly therapist guidance | OCD-Symtoms: OCI-R |
Low risk Concealed randomization. ITT-analysis (REML), 39% (n=47) lost in follow-up. |
| 4. | RCT | Australia. Adults Recruited from primary care, mental health care or self- referral via online advertisement. Clinician confirmed DSM-IV-TR dg for OCD Mean age 33.4, Female 65.7% ongoing psychological treatment 33.6% On medication 69.3% Comorbidity 76.8% YBOCS at baseline 21.94 (SD 0.49) |
Therapist assisted iCBT vs therapist assisted progressive relaxation iPRT, 12 modules(weeks) in each. | OCD symptoms: YBOCS |
Low risk Stratified, automated, blind randomization. Posttreatment assessment was blind to randomization. |
| 5. | RCT | Germany. Adults, 18-65yrs Anonymous online recruitment in social media and health care facilities. OCD-symptoms on Y-BOCS >7 Mean age 40.3 Female 76% Current psychotherapy 27.3% On medication 32.8% Mean baseline YBOCS 20.19 (SD 5.99) |
unguided iCBT&CAU vs. CAU (Care as Usual) 8weeks program. |
OCD-symptoms: YBOCS OCI-R OBQ-44 OCCWG Other outcomes: WHOQOL-BREF PHQ-9 I-8 RSES IQ-24 |
Low risk Sealed, automated randomization. High drop-out rate, 75% participated in post-assessment. Only 58% logged in the treatment application at least once. Mean use of application 172min (Median 122min) |
| 6. | RCT | Australia, UK, US, Canada. Adults >18yrs Email recruitment of subjects who had previously expressed interest in program + social media advertisement. Scored >7 in DOCS and >14 in YBOCS-SR Exclusion criteria: psychosis or bipolar disorder, serious self-harm, severe depression (>20p inPHQ-9) Mean age 33.4-34.0 Female 81% on psychotropic medication 47.7-52% |
self-guided iCBT vs. wait list. 5 lesson treatment. | OCD-symptoms: YBOCS-SR DOCS Depression: PHQ-9 |
Medium risk for bias. Not blinded randomization. 4 of the 69 participants randomized to the intervention group were not included in analyses as they did not commence using the materials. Last lesson (Nr.5) participated 40% (n=26) 19-32% missing at post-treatment 46% missing at follow-up |
| Reference | Number of studies and number of patients (I/C) | Follow-up time | Results as effect size | Relative effect (95% CI) |
|---|---|---|---|---|
| Level of evidence: high/moderate/low/very low Assess the risk of bias and delete irrelevant sources of bias: The quality of evidence is downgraded due to study limitations, inconsistency, indirectness, imprecision, publication bias. I=intervention; C=comparison; CI=confidence interval *SA=self administered: no therapist contact, PSH=Predominantly Self-help: very limited written contact with the therapist, at some occasions short phone calls. MC=Minimal Contact Self-help: Live therapist contact (call or therapist meetings) |
||||
| 1. | Studies N=8 patients n=420 of which children (13-18yrs) n=31 |
post-treatment, T-CBT vs. control condition | Effect size OCD d=0.82 Depression d=0.33 |
0.55-1.08 -0.01-0.67 |
| 2. | N(studies)=18 N(patients)=1570
|
3-17 weeks (post-treatment) | Effect size* for OCD symptoms when heterogeinity was taken into account. Hedges' g overall g=0.55 SA g=0.33 PSH g=0.68 MC g=1.08 *only RCT studies from the meta-anlysis (ie. Not observational studies) |
overall 0.44-0.66 SA 0.18-0.47 PSH 0.40-0.96 MC 0.79-1.37 |
| CBT interventions only N(studies=6) N(patients=845) |
d=0.38-1.53 | all statistically significant (>0) | ||
| 3. | N=120 Intervention (n=62), wait list control group (n=58). |
post-intervention (12 weeks), and at follow-up (24 weeks). | between group effect size d (ITT) 12 weeks d=-0.55 24 weeks d=-0.6 27,4% of intervention group completed no treatment modules, 58.1% completed 1-4 modules (half or less), 14.5% completed 5-8 modules (Mean number of modules completed was 2.48 (SD 2.37) |
(-0.92- -0.42) (-0.97- -0.24) |
| 4. | N=179 Interventio n=89 Control n=90 |
post-intervention | YBOCS Cohen d (ITT) Within iCBT d=1.05 Between groups d=0.55 Reliable recovery (YBOCS<14 and >6p change) 18% vs 6% (p=0.048) (YBOCS <16p and >6p change) 33% vs 11% |
(0.72-1.37) (0.18-0-91) |
| 5. | N=128 Intervention n=64 Control n=64 |
post-intervention | Cohens d (ITT) YBOCS d=0.3 RSES d=0.3 other secondary outcomes nonsignificant |
p=0.176 p=0.017 |
| 6. | N=190 Intervention n=69, but 4 were not included in the ITT analyses as they did not commence using the materials Control n=75 |
post-intervention 3months follow-up within-group effect |
Cohens d (ITT) YBOCS-SR d=1.05 DOCS d=0.84 PHQ-9 d=0.58 YBOCS d=1.23 PHQ-9 d=0.42 38% showed clinically significant change 3m f-u. 4% showed deterioration no adverse events Responders (³30% reduction in YBOCS) post-treatment I 27% C 2% At follow-up I 32% (Intervention) |
0.89-1.21 0.69-1.00 0.43-0.73 0.79-1.68 0.01-0.84 24-55% 16-42% 0-11% 19-50% |
| Reference | Comments |
|---|---|
| 1. | Following studies were not included in Pearcy et al 2016 meta-analysis. Lovell (n=72). Reason for not inclusion: face-to-face vs telephone administered therapy (i.e. therapist administered) Storch (n=31). Reason for not inclusion: only children 7-16yrs Note: Dettore et al. included only CBT-therapies and RCT-studies, Pearcy et al. all frameworks and also quasi experimental studies. |
| 2. | Included different kinds of theoretical backgrounds. Of 18 studies 6 were CBT, 7 ERP,
5 mixed (1 meridian tapping, 1 competitive memory training, 1 association splitting,
1 metacognitive training, 1 Attention training) The following studies were included in the meta-analysis of Dettore et al 2015: Andersson et al 2012 (N=101), Griest et al 2002 (n=183) Kenwright et al 2005 (n=36) Mahoney et al 2014 (n=86) Tolin et al 2007 (n=41) Wootton et al 2013 (n=52) Limitations of the study: No RCT's using evidence-based self-help treatments (CBT and ERP) were conducted within the self-administered self-help category. Some risk of bias remains, even after adjustments. Inexistent or short follow-up times. Characteristics of the studies in this meta-analysis, which were not included in the meta-analysis by Dettore et al: 1 i-CBT (Herbst et al 2014) 1 self-help bibliotherapy (Vogel et al 2014) 8 self-administered self-help programs, of which none was a CBT based RCT-study. |