scispace - formally typeset
Search or ask a question
Journal Article•DOI•

Efficacy of self-guided internet-based cognitive behavioral therapy in the treatment of depressive symptoms a meta-analysis of individual participant data

Eirini Karyotaki1, Heleen Riper1, Jos W. R. Twisk2, Adriaan W. Hoogendoorn1, Annet Kleiboer1, Adriana Mira, Andrew Mackinnon3, Björn Meyer4, Cristina Botella, Elizabeth Littlewood5, Gerhard Andersson6, Helen Christensen3, Jan Philipp Klein7, Johanna Schröder8, Juana Bretón-López, Justine Scheider9, Kathy Griffiths10, Louise Farrer10, Marcus J.H. Huibers1, Rachel Phillips11, Simon Gilbody5, Steffen Moritz8, Thomas Berger12, Victor J M Pop13, Viola Spek13, Pim Cuijpers1 •
VU University Amsterdam1, University of Amsterdam2, University of New South Wales3, City University London4, University of York5, Karolinska Institutet6, University of Lübeck7, University of Hamburg8, University of Nottingham9, Australian National University10, King's College London11, University of Bern12, Tilburg University13
01 Apr 2017-JAMA Psychiatry (American Medical Association)-Vol. 74, Iss: 4, pp 351-359
TL;DR: In this article, the effect of self-guided internet-based cognitive behavioral therapy (iCBT) in treating adults with depressive symptoms compared with controls and evaluate the moderating effects of treatment outcome and response.
Abstract: IMPORTANCE Self-guided internet-based cognitive behavioral therapy (iCBT) has the potential to increase access and availability of evidence-based therapy and reduce the cost of depression treatment. OBJECTIVES To estimate the effect of self-guided iCBT in treating adults with depressive symptoms compared with controls and evaluate the moderating effects of treatment outcome and response. DATA SOURCES A total of 13 384 abstracts were retrieved through a systematic literature search in PubMed, Embase, PsycINFO, and Cochrane Library from database inception to January 1, 2016. STUDY SELECTION Randomized clinical trials in which self-guided iCBT was compared with a control (usual care, waiting list, or attention control) in individuals with symptoms of depression. DATA EXTRACTION AND SYNTHESIS Primary authors provided individual participant data from 3876 participants from 13 of 16 eligible studies. Missing data were handled using multiple imputations. Mixed-effects models with participants nested within studies were used to examine treatment outcomes and moderators. MAIN OUTCOMES AND MEASURES Outcomes included the Beck Depression Inventory, Center for Epidemiological Studies-Depression Scale, and 9-item Patient Health Questionnaire scores. Scales were standardized across the pool of the included studies. RESULTS Of the 3876 study participants, the mean (SD) age was 42.0 (11.7) years, 2531 (66.0%) of 3832 were female, 1368 (53.1%) of 2574 completed secondary education, and 2262 (71.9%) of 3146 were employed. Self-guided iCBT was significantly more effective than controls on depressive symptoms severity (s =-0.21; Hedges g = 0.27) and treatment response (s = 0.53; odds ratio, 1.95; 95% CI, 1.52-2.50; number needed to treat, 8). Adherence to treatment was associated with lower depressive symptoms (s =-0.19; P =.001) and greater response to treatment (s = 0.90; P <.001). None of the examined participant and study-level variables moderated treatment outcomes. CONCLUSIONS AND RELEVANCE Self-guided iCBT is effective in treating depressive symptoms. The use of meta-analyses of individual participant data provides substantial evidence for clinical and policy decision making because self-guided iCBT can be considered as an evidence-based first-step approach in treating symptoms of depression. Several limitations of the iCBT should be addressed before it can be disseminated into routine care.

Summary (4 min read)

Jump to: [M] – [Eligibility Criteria] – [Study Identification and Selection Process] – [Data Collection and Data Items] – [Risk of Bias Assessment in Individual Studies] – [Key Points] – [IPD Meta-analysis] – [Exploration of Variation in Effects: Participant-Level Moderators] – [Treatment Adherence as a Predictor Within the Treatment Group] – [Study and Participant Characteristics] – [Results of Traditional Meta-analysis] – [One-Stage IPD Meta-analysis: Depressive Symptoms Severity] – [Discussion] – [Strengths and Limitations] and [Conclusions]

M

  • Any studies [1] [2] [3] [4] have found that depressive symptoms can be effectively treated with psychotherapy, pharmacotherapy, or both.
  • Nevertheless, many people with depressive symptoms do not seek help, and even well-resourced health care systems find it difficult to marshal enough qualified therapists to offer psychological interventions.
  • 14, 15 These contradicting findings drew much attention and raised concerns about the benefits of these interventions.
  • 16 Meta-analyses using individual participant data (IPD) estimate aggregate effect sizes using IPD from RCTs.
  • The term self-guided iCBT is defined as CBT delivered via the internet, which may involve automated feedback but does not provide support related to the therapeutic content.

Eligibility Criteria

  • Studies were included if the participants were adults (aged >18 years) with elevated symptoms of depression based on any diagnosis or any self-report scale of depression.
  • Only those RCTs in which self-guided iCBT was compared with a control condition (usual care, waiting list, or attention control) were included.
  • No language or publication status exclusions were applied.

Study Identification and Selection Process

  • The analysis was completed in compliance with the Preferred Reporting Items for Systematic Review and Metaanalyses IPD Statement.
  • The authors used an existing database on psychological treatments for depression 18 that is updated annually by a systematic literature search in the bibliographic databases of PubMed, Embase, PsycINFO, and Cochrane Library (from inception to January 1, 2016).
  • Two researchers (P.C. and E.K.) independently examined titles and abstracts of 13 384 articles.
  • In case of disagreement regarding inclusion, consensus was sought through discussion.
  • The authors also asked key researchers in the field whether they knew of unpublished trials.

Data Collection and Data Items

  • Authors of eligible articles were contacted for permission to use their data sets.
  • Reminders were sent after 2 weeks and if necessary after 1 month.
  • Finally, the authors combined all individual data sets into a merged data set, using a generic standardized protocol for integrating IPD sets.
  • The authors also used study-level variables, which were available from the full reports (type of comparator condition, recruitment, level of support).
  • The selection of moderator variables has been based on previous literature related to moderators of face-to-face CBT or iCBT.

Risk of Bias Assessment in Individual Studies

  • The authors examined the risk of bias in the included studies using the criteria of the Cochrane Collaboration risk of bias assessment tool.
  • 20 Two independent reviewers (E.K., P.C.) evaluated the included studies to determine whether there was a risk for bias related to selection, performance, detection, attrition, and outcome reporting.
  • In case of unclear risk of bias for 1 or more key domains, the authors contacted the first authors of the included studies for clarifications.

Key Points

  • Internet-based cognitive behavioral therapy was more effective compared with controls.
  • Adherence predicted better treatment outcomes within the experimental condition.
  • Meaning Self-guided internet-based cognitive behavioral therapy may be a viable alternative to current first-step treatment approaches for symptoms of depression, particularly in those individuals who are not willing to have any therapeutic contact.

IPD Meta-analysis

  • Studies included in this IPD meta-analysis used measures such as the Center of Epidemiologic Studies-Depression Scale, 22 the Beck Depression Inventory I 23 or II, 24 (hereafter referred to as Beck Depression Inventory) or the 9-item Patient Health Questionnaire 25 to monitor change in depressive symptoms severity.
  • The authors also conducted sensitivity analyses using only participants with complete data after treatment to examine whether there was a difference between those who dropped out of the RCTs and those who provided posttreatment data.
  • 27, 28 We calculated the standardized β coefficient for the examined comparisons.the authors.the authors.
  • Second, the authors analyzed the effects of the interventions on treatment response (defined as a 50% reduction in baseline depressive symptoms scores) at the posttreatment assessment using a multilevel mixed-effects logistic regression (using a random intercepts model with a random effect for each trial and fixed effects for the intervention and the depressive symptoms severity, using STATA's melogit command).
  • Two-stage IPD meta-analysis facilitates analysis standardization across the included studies and estimation of outcomes that are not available in the published reports, such as treatment response.

Exploration of Variation in Effects: Participant-Level Moderators

  • The authors tested whether available demographic and clinical characteristics moderated the effect of self-guided iCBT on depression outcomes (depressive symptoms severity and treatment response).
  • Not all included studies reported data on the examined moderators (for precise numbers regarding the missing data, see Table 1 and Table 2 ).
  • To examine moderators, the authors added the interaction between each potential moderator and treatment outcome on depression severity into the multilevel mixed-effects linear regression model.
  • The authors similarly added the interaction between each potential moderator and treatment response into the multilevel mixed-effects logistic regression model.

Treatment Adherence as a Predictor Within the Treatment Group

  • The authors examined whether adherence to treatment predicted within treatment group effect size for the experimental condition only, using a linear mixed model, which regressed posttreatment depressive symptoms severity on treatment adherence and baseline depressive symptoms severity (fixed effects) and using random intercepts for the studies.
  • Treatment adherence was defined as the total number of sessions that each participant completed divided by the total number of treatment sessions.

Study and Participant Characteristics

  • The included studies were conducted in 6 countries: Australia, Germany, Spain, Switzerland, the Neth-erlands, and the United Kingdom (eTable 1 in the Supplement presents a summary of study characteristics).
  • The mean baseline depressive symptoms scores were 25.7 on the Center of Epidemiologic Studies-Depression Scale, 28.3 on the Beck Depression Inventory, and 14.1 on the 9-item Patient Health Questionnaire in their respective studies.
  • (eTable 2 in the Supplement provides a summary of participants' characteristics.) .17 a This is a sensitivity analysis that was conducted including only participants who completed posttreatment depression questionnaires.

Results of Traditional Meta-analysis

  • Sixteen studies examined the comparison between selfguided iCBT and control groups.
  • There was no significant difference between the outcome findings of studies included in the present IPD meta-analysis and studies with unavailable data (P = .95) .
  • There was some indication of publication bias.

One-Stage IPD Meta-analysis: Depressive Symptoms Severity

  • Table 1 presents the main findings of the 1-stage IPD metaanalysis on depressive symptoms severity after testing (ranging from 6 to 16 weeks after randomization).
  • None of the participant-level variables (sociodemographic and clinical characteristics) significantly moderated outcome after treatment (Table 1 ).
  • Adherence to treatment predicted significantly better outcomes within the self-guided iCBT group (β = −0.19; P = .001).

Discussion

  • The authors examined the effects of self-guided iCBT on severity and treatment response.
  • The authors found that self-guided iCBT had lower depressive symptom severity and greater treatment response compared with control conditions after testing.
  • These findings were robust in complete case analyses.
  • None of the examined participant-and study-level variables significantly moderated the treatment effect.
  • The role of treatment adherence in outcomes has been identified by a previous review in the field conducted by Donkin and colleagues.

Strengths and Limitations

  • Among the strengths of the present study was its high power to detect small statistically significant differences between intervention and controls and to yield more precise and robust evidence compared with traditional meta-analyses.
  • Moreover, the included RCTs had high methodologic quality, which allows us to be confident that the present analysis is relatively free of critical biases.
  • This repeated administration of symptom inventories might yield lower mean scores with each wave of measurement (completer biases related to self-report ratings).
  • 45 Moreover, the included studies did not report on recruitment issues related to large-scale, fully unguided internetadministered interventions, including factors such as repeated registration attempts by individuals who did not meet inclusion criteria or who were dissatisfied with their intervention allocation.

Conclusions

  • Self-guided iCBT produces results that are encouraging.
  • The absence of a significant difference in treatment outcomes associated with clinical and sociodemographic characteristics implies that self-guided iCBT can be used by most individuals with depressive symptoms regardless of the severity of their symptoms or their sociodemographic background.
  • Currently, antidepressant medications are widely used in the treatment of depressive symptoms, whereas psychotherapeutic interventions are provided to a lesser degree, despite many individuals with depressive symptoms preferring psychotherapy to antidepressants.
  • 46 However, the high treatment costs and the limited number of trained clinicians hamper the implementation of psychotherapy in practice.
  • Unguided iCBT has several limitations that should be addressed before it is disseminated as part of routine care (eg, high dropout rates, small effects compared with face-to-face and guided internet interventions, and possible participant selection bias).

Did you find this useful? Give us your feedback

Figures (4)

Content maybe subject to copyright    Report

Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/313891168
EfficacyofSelf-guidedInternet-Based
CognitiveBehavioralTherapyintheTreatment
ofDepressiveSymptomsA...
ArticleinJAMAPsychiatry·March2017
DOI:10.1001/jamapsychiatry.2017.0044
CITATIONS
2
READS
333
25authors,including:
Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:
PositiveTechnology:Designingdigitale-experiencesforpositivepersonalchangeViewproject
Frequencyandchangemechanismsofpsychotherapyamongdepressedpatients:amulticenter
randomizedtrialcomparingtwice-weeklyversusonce-weeklysessionsofCBTandIPTViewproject
HeleenRiper
VUUniversityAmsterdam
183PUBLICATIONS4,285CITATIONS
SEEPROFILE
AnnetKleiboer
VUUniversityAmsterdam
66PUBLICATIONS821CITATIONS
SEEPROFILE
GerhardAndersson
LinköpingUniversity
773PUBLICATIONS21,386CITATIONS
SEEPROFILE
PimCuijpers
VUUniversityAmsterdam
767PUBLICATIONS26,409CITATIONS
SEEPROFILE
AllcontentfollowingthispagewasuploadedbyHeleenRiperon22February2017.
Theuserhasrequestedenhancementofthedownloadedfile.Allin-textreferencesunderlinedinblueareaddedtotheoriginaldocument
andarelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.
Copyright 2017 American Medical Association. All rights reserved.
Eff icacy of Self-guided Internet-Based Cognitive Behavioral
Therapy in the Treatment of Depressive Symptoms
A Meta-analysis of Individual Participant Data
Eirini Karyotaki, MSc; Heleen Riper, PhD; Jos Twisk, PhD; Adriaan Hoogendoorn, PhD; Annet Kleiboer, PhD;
Adriana Mira, PhD; Andrew Mackinnon, PhD; Björn Meyer, PhD; Cristina Botella, PhD; Elizabeth Littlewood, PhD;
Gerhard Andersson, PhD; Helen Christensen, PhD; Jan P. Klein, PhD; Johanna Schröder, PhD;
Juana Bretón-López, PhD; Justine Scheider, PhD; Kathy Griffiths, PhD; Louise Farrer, PhD;
Marcus J. H. Huibers, PhD; Rachel Phillips, MSc; Simon Gilbody, PhD; Steffen Moritz, PhD; Thomas Berger, PhD;
Victor Pop, PhD; Viola Spek, PhD; Pim Cuijpers, PhD
IMPORTANCE
Self-guided internet-based cognitive behavioral therapy (iCBT) has the
potential to increase access and availability of evidence-based therapy and reduce the cost of
depression treatment.
OBJECTIVES To estimate the effect of self-guided iCBT in treating adults with depre ssive
symptoms compared with controls and evaluate the moderating effects of treatment
outcome and response.
DATA SOURCES A total of 13 384 abstracts were retrieved through a systematic literature
search in PubMed, Embase, PsycINFO, and Cochrane Library from database inception to
January 1, 2016.
STUDY SELECTION Randomized clinical trials in which self-guided iCBT was compared with a
control (usual care, waiting list, or attention control) in individuals with symptoms of
depression.
DATA EXTRACTION AND SYNTHESIS Primary authors provided individual participant data from
3876 participants from 13 of 16 eligible studies. Missing data were handled using multiple
imputations. Mixed-effects models with participants nested within studies were used to
examine treatment outcomes and moderators.
MAIN OUTCOMES AND MEASURES Outcomes included the Beck Depression Inventory, Center
for Epidemiological Studies–Depression Scale, and 9-item Patient Health Questionnaire
scores. Scales were standardized across the pool of the included studies.
RESULTS Of the 3876 study participants, the mean (SD) age was 42.0 (11.7) years, 2531
(66.0%) of 3832 were female, 1368 (53.1%) of 2574 completed secondary education, and
2262 (71.9%) of 3146 were employed. Self-guided iCBT was significantly more effective than
controls on depressive symptoms severity (β = −0.21; Hedges g = 0.27) and treatment
response (β = 0.53; odds ratio, 1.95; 95% CI, 1.52-2.50; number needed to treat, 8).
Adherence to treatment was associated with lower depressive symptoms (β = −0.19;
P = .001) and greater response to treatment (β = 0.90; P < .001). None of the examined
participant and study-level variables moderated treatment outcomes.
CONCLUSIONS AND RELEVANCE Self-guided iCBT is effective in treating depressive symptoms.
The use of meta-analyses of individual participant data provides substantial evidence for
clinical and policy decision making because self-guided iCBT can be considered as an
evidence-based first-step approach in treating symptoms of depression. Several limitations of
the iCBT should be addressed before it can be disseminated into routine care.
JAMA Psychiatry. doi:10.1001/jamapsychiatry.2017.0044
Published online February 22, 2017.
Supplemental content
Author Affiliations: Author
affiliations are listed at the end of this
article.
Corresponding Author: Eirini
Karyotaki, MSc, Department of
Clinical Psychology and EMGO
Institute for Health and Care
Research, VU Amsterdam,
Van der Boechorststraat 1, 1081 BT
Amsterdam, the Netherlands
(e.karyotaki@vu.nl).
Research
JAMA Psychiatry | Original Investigation | META-ANALYSIS
(Reprinted) E1
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/0/ by a Vrije Universiteit User on 02/22/2017
Copyright 2017 American Medical Association. All rights reserved.
M
any studies
1-4
have found that depressive symp-
toms can be effectively treated w ith ps ycho -
therapy, pharmacotherapy, or both. Nevertheless,
many people with depressive symptoms do not seek help, and
even well-resourced healthcare systemsfindit difficult to mar-
shal enough qualified therapists to offer psychological inter-
ventions. Access barriers to psychotherapy include limited
availability of trained clinicians, high cost of treatment, and
fear of stigmatization.
5-8
As a consequence, a significant num-
ber of individuals with depressive symptoms remain
untreated.
9,10
Self-guided internet-based cognitive behavioral therapy
(iCBT) without therapist support can allow physicians, such
as general practitioners, to provide easy and affordable ac-
cess to psychological treatments and reduce the cost of such
treatments. A meta-analysis
11
found a small but significant ef-
fect size of self-gui d e d iCBT comp a re d w ith control condi -
tions. However , recent large trials found a range of effects, vary-
ing from small to moderate effect sizes
12,13
to no effect .
14,15
These contradicting findings drew much attention and raised
concerns about the benefits of these interventions.
Randomized clinical trials (RCT s) and study-level system-
atic reviews often lack adequate power and precision in their
estimates. Statistically underpowered samples also preclude
identification of clinically useful moderators or predictors of
treatment outcome.
16
Meta-analyses using individual partici-
pant data (IPD) estimate aggregate effect sizes using IPD from
RCTs. The IPD maximize power to detec t a true effect while
allowing the exploration of study variability (eg, level of sup-
port, treatment adherence, setting) and participant character-
istics as moderators of treatment outcome. The present study
reports the results of an IPD meta-analysis of trials on self-
guided iCBT for adult depressive symptoms compared with
control conditions. The term self-guided iCBT is defined as CBT
delivered via the internet, which may involve automated feed-
back but does not provide support related to the therapeutic
content.
Methods
Eligibility Criteria
Studies were included if the participants were adults (aged >18
years) with elevated symptoms of depression based on any di-
agnosis or any self-report scale of depression. Only those RCTs
in which self-guided iCBT was compared with a control con-
dition (usual care, waiting list, or attention control) were in-
cluded. No language or publication status exclusions were
applied.
Study Identification and Selection Process
The analysis was completed in compliance with the Pre-
ferred Reporti ng Items for Sys tematic Rev iew an d Meta-
analyses (PRISMA) IPD Statement.
17
We used an existing da-
tabase on psychological treatments for depression
18
that is
updated annually by a systematic literature search in the bib-
liographic databases of PubMed, Embase, PsycINFO, and Coch-
rane Library (from inception to January 1, 2016). In these
searches, various index and free terms of psychotherapy and
depression are used in different combinations (full search
strings for PubMed are provided in the eMethods in the Supple-
ment). Two researchers (P.C. and E.K.) independently exam-
ined titles and abstracts of 13 384 articles. The full text of stud-
ies that possibly met the inclusion criteria according to 1 of the
2 reviewers was retrieved. In case of disagreement regarding
inclusion, consensus was sought through discussion. We also
asked key researchers in the field whether they knew of un-
published trials.
Data Collection and Data Items
Authors of eligible articles were contacted for permission to
use their data sets. Reminders were sent after 2 weeks and if
necessary after 1 month. If no response was received, we ex-
cluded the trial. Authors were asked to provide data on so-
ciodemographic, clinical,and intervention characteristics, in-
cluding information regarding randomized group, baseline and
follow-up total scores of depressive symptoms, treatment ad-
herence information (total number of sessions completed di-
vided by total number of treatment sessions), age, sex, edu-
cational level (primary, secondary, and tertiary education),
employment status (employed or unemployed), relationship
status (in a relationship or not), and comorbid anxiety symp-
toms at baseline (yes or no; based on a clinical interview or el-
evated anxiety symptoms ratings on self-report measures). Fi-
nally, we combined all individual data sets into a merged data
set, using a generic standardized protocol for integrating IPD
sets.
1
We also used study-level variables , whic h were avail-
able from the full reports (type of comparator condition, re-
cruitment, level of support). The selection of moderator vari-
ables has been based on previous literature related to
moderators of face-to-face CBT or iCBT.
16,19
Risk of Bias Assessment in Individual Studies
We examined the risk of bias in the included studies using the
criteria of the Cochrane Collaboration risk of bias assessment
tool.
20
Two independent reviewers (E.K., P.C.) evaluated the
included studies to determine whether there was a risk for bias
related to selection, performance, detection, attrition, and out-
come reporting. In case of unclear risk of bias for 1 or more key
domains, we contacted the first authors of the included stud-
ies for clarifications.
Key Points
Questions Is self-guided internet-based cognitive behavioral
therapy effective in treating depressive symptoms and which
variables moderate treatment outcome?
Findings In this meta-analysis of individual participant data from
3876 adults, internet-based cognitive behavioral therapy was
more effective compared with controls. Adherence predicted
better treatment outcomes within the experimental condition.
Meaning Self-guided internet-based cognitive behavioral therapy
may be a viable alternative to current first-step treatment
approaches for symptoms of depression, particularly in those
individuals who are not willing to have any therapeutic contact.
Research Original Investigation Internet-Based Cognitive Behavioral Therapy for Treatment of Depression
E2 JAMA Psychiatry Published online February 22, 2017 (Reprinted) jamapsychiatry.com
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/0/ by a Vrije Universiteit User on 02/22/2017
Copyright 2017 American Medical Association. All rights reserved.
Traditional Meta-analysis
We conducted a traditional meta-analysis to examine differ-
ences among the 13 studies that provided the IPD and the 3
studies that did not. We used data reported in the articles to
calculate the effect sizes (Hedges g).
21
The reader is referred
to the eMethods in the Supplement for details regarding the
methods of the traditional meta-analysis.
IPD Me ta-analysis
Studies included in this IPD meta-analysis used measures such
as the Center of Epidemiologic Studies–Depression Scale,
22
the
Beck Depression Inventory I
23
or II,
24
(hereafter referred to as
Beck Depression Inventory) or the 9-item Patient Health
Questionnaire
25
to monitor change in depressive symptoms se-
verity. These depression measures were standardized by trans-
formation into z scores across the pool of the studies before
conducting the main analysis.
Missing outcome data at the posttreatment assessment
were estimated using multiple imputation under the missing-
at-random assumption (mi impute mvn in STATA software, ver-
sion 13.1; StataCorp). This method generated 100 imputed data
sets using data on baseline depressive symptoms scores, age,
sex, and group. These new imputed data sets included the ob-
served and the imputed standardized depressive symptoms
scores for the missing values. They were analyzed separately
using the selected model, and the results were averaged ac-
cording to Rubin ’s rules.
26
We also conducted sensitivity analy-
ses using only participants w ith complete data after treat-
ment to examine whether there was a difference between those
who dropped out of the RCTs and those who provided post-
treatment data.
In a 1-stage IPD meta-analysis, we merged all IPD from all
studies with participants nested within studies. One-stage IPD
meta-analysis y ields more precise and less biased estimates
of effect, maximizes the power, and accounts for parameter
correlation.
27,28
We calculated the standardized β coefficient
for the examined comparisons. This estimate indicates how
many SDs the dependent variable (depressive symptoms se-
verity or the log odds ratio [OR] of treatment response) changes
per SD increase in the predictor variable. Thus, the higher the
β is the greater the effect of the predictor variable on the de-
pendent variable, although there is no association among the
variables if the β is 0. All analyses were conducted with STATA
statistical software, version 131. The primary analysis was
2-fold. First, we analyzed the effects of the interventions on
depressive symptom severity at the end of treatment using a
multilevel mixed-effects linear regression (using a random in-
tercepts model with a random effect for each trial and fixed
effects for the intervention and the symptoms severity, using
STATA’s mixed command). The posttreatment depression
scores were used as the dependent variable and trial arm con-
dition (treatment vs control) as the independent variable, while
controlling for baseline depressive symptom severity.
Second, we analyzed the effects of the interventions on
treatment response (defined as a 50% reduction in baseline de-
pressive symptoms scores) at the posttreatment assessment
using a multilevel mix ed-effects logistic regression (using a ran-
dom intercepts model with a random effect for each trial and
fixed effects for the intervention and the depressive symp-
toms severity, using STATA’s melogit command). The re-
sponse (yes or no) was the dependent variable, and condition
was the independent variable.
Third, we ran a 2-stage IPD meta-analysis analyzing the IPD
separately in each study and then combining the estimates to
calculate the pooledeffectsizes(Hedges g)fordepressivesymp-
toms severity. Two-stage IPD meta-analysis facilitates analy-
sis standardization across the includ ed studies and estima-
tion of outcomes that are not available in the published reports,
such as treatment response.
29
Similarly, we calculated the OR
of treatment response and numbers needed to treat (NNTs),
which allowed us to compare the results of the present meta-
analysis with those reported in earlier meta-analyses. In ad-
dition, 2-stage IPD meta-analysis also allowed us to examine
the moderation effect of study -lev el variables. Thus, subgroup-
moderator analyses were conducted using a mixed-effects
model in which the random-effects model was used to pool
studies w ithin subgroups, whereas between-subgroup
differences were tested as fixed effects. We also ran meta-
regression analyses to examine the association between treat-
ment duration and treatment outcomes (severity of depres-
sive symptoms and treatment response).
Exploration of Variation in Effects:
Participant-Level Moderators
We tested whether available demographic and clinical char-
acteristics moderated the effect of self-guided iCBT on de-
pression outcomes (depressive symptoms severity and treat-
ment response). Not all included studies reported data on the
examined moderators (for precise numbers regarding the miss-
ing data, see Table 1 and Table 2). To examine moderators, we
added the interaction between each potential moderator and
treatment outcome on depression severity into the multi-
level mixed-effects linear regression model. W e similarly added
the interaction between each potential moderator and treat-
ment response into the multilevel mixed-effects logistic re-
gression model. Each potential moderator was included in a
separate model as a main effect.
Treatment Adherence as a Predictor
Within the Treatment Group
We examined whether adherence to treatment predicted within
treatmentgroup effect size for the experimental condition only,
using a linear mixed model, which regressed posttreatment de-
pressive symptoms severity on treatment adherence and base-
line depressive symptoms severity (fixed effects) and using ran-
dom intercepts for the studies. Treatment adherence was
defined as the total number of sessions that each participant
completed divided by the total number of treatment ses-
sions.
Results
Study Selection and IPD Obtained
The systematic search resulted in 16 eligible articles of 1885
full-text articles screened. We were able to obtain IPD from
Internet-Based Cognitive Behavioral Therapy for Treatment of Depression Original Investigation Research
jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online February 22, 2017 E3
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/0/ by a Vrije Universiteit User on 02/22/2017
Copyright 2017 American Medical Association. All rights reserved.
13 of the 16 eligible trials (81%), y ielding a total of 3876
participants.
12,14,15,30-38
Three eligible data sets
39-41
were un-
availabl e a nd thu s c o uld not b e i n clud ed in the IPD me t a-
analyses. Figure 1 shows the study selection process.
Study and Participant Characteristics
Seven of the included studies
30,31,34,35,37,38
recruited partic i-
pants through the community. The included RCTs examined
iCBT, with interventions comprising 5 to 11online sessions.Four
of the included trials provided support related to the techni-
cal aspects of the online platforms,
15,31,33,36
whereas 9 trials
were purely self-guided.
12-14,30,32,34,35,37,38
The control condi-
tions used were attention placebo, no treatment, treatment as
usual, or waiting list. The included studies were conducted in
6countries:Australia,Germany,Spain,Switzerland,theNeth-
erlands, and the United Kingdom (eTable 1 in the Supplement
presents a summary of study characteristics).
Of the 3876 study participants, the mean (SD) age was 42.0
(11.7) years, 2531 (66.0%) of 3832 were female, 1368 (53.1%) of
2574 completed secondary education, and 2262 (71.9%) of 3146
were em p l oyed. The me a n b a s e l i n e d epre s s ive sy m p t o m s
scores were 25.7 on the Center of Epidemiologic Studies–
Depression S c ale, 28.3 on the B eck Depression Inventory, and
14.1 on the 9-item Patient Health Questionnaire in their re-
spective studies. Finally, 71 (1.8%) of 3876 randomized par-
ticipants did not start the treatment or did not provide base-
line and posttreatmen t data, and 1048 (27.0%) of 3876 dropped
out of the RCT and did not provide posttreatment depressive
symptoms scores. (eT able 2 in the Supplement provides a sum-
mary of participants’ characteristics.)
Table 1. Mixed-Effects Model Outcomes on Depressive Symptom Severity for 1-Stage Individual Patient Data
Variable
Full Sample Complete Cases Analysis
a
No. of Observations
(No. of Studies) Mean (SE) β
b
2-Tailed
P Value
No. of Observations
(No. of Studies) Mean (SE) β
b
2-Tailed
P Value
Main effects of depression severity
Baseline severity 3795 (13) 0.57 (0.02) <.001 2818 (13) 0.57 (0.02) <.001
Treatment group −0.21 (0.03) <.001 −0.19 (0.03) <.001
Age
Baseline severity 3786 (13) 0.58 (0.02) <.001 2809 (13) 0.57 (0.02)
Treatment group −0.32 (0.10) <.001 −0.33 (0.11) .003
Age × treatment group 0.003(0.002) .28 0.003 (0.002) .19
Sex
Baseline severity 3788 (13) 0.58 (0.02) <.001 2811 (13) 0.57 (0.02) <.001
Treatment group −0.22 (0.03) <.001 −.0.22 (0.04) <.001
Sex × treatment group 0.05 (0.06) .45 0.07 (0.06) .26
Educational level
Baseline severity 2538 (10) 0.58 (.024) <.001 1973 (10) 0.57 (0.02) <.001
Treatment group −0.031 (0.011) <.001 −0.31 (0.12) .00
Educational level × treatment group
Secondary vs primary education 0.15 (0.13) .21 0.19 (0.13) .14
Tertiary vs primary education 0.03 (0.13) .79 0.02 (0.13) .84
Relationship status
Baseline severity 3568 (12) 0.57 (0.02) <.001 2630 (12) 0.56 (0.02) <.001
Treatment group −0.20 (0.05) <.001 −0.18 (0.05) <.001
Relationship status × treatment group 0.006 (0.06) .91 −0.004 (0.06) .95
Employment status
Baseline severity 3067 (10) 0.55 (0.02) <.001 2194 (10) 0.53 (0.02) <.001
Treatment group −0.27 (0.06) <.001 −0.26 (0.07) <.001
Employment status × treatment group 0.12 (0.08) .11 0.14 (0.08) .07
Comorbid anxiety
Baseline severity 1728 (9) 0.62 (0.03) <.001 1447 (9) 0.62 (0.03) <.001
Treatment group −0.20 (0.05) <.001 −0.19 (0.05) <.001
Comorbid anxiety × treatment group −0.10 (0.07) .17 −0.11 (0.07) .13
Baseline severity of depression
Baseline severity 3795 (13) 0.59 (0.02) <.001 2818 (13) 0.59 (0.02) <.001
Treatment group −0.20 (0.03) <.001 −0.19 (0.03) <.001
Baseline severity × treatment group −0.03 (0.03) .22 −0.04 (0.03) .17
a
This is a sensitivity analysis that was conducted including only participants
who completed posttreatment depression questionnaires.
b
Standardized β weights of the composite z scores of the Beck Depression
Inventory I or II,
23,24
Center for Epidemiological Studies–Depression Scale,
22
and 9-item Patient Health Questionnaire.
25
Research Original Investigation Internet-Based Cognitive Behavioral Therapy for Treatment of Depression
E4 JAMA Psychiatry Published online February 22, 2017 (Reprinted) jamapsychiatry.com
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/psych/0/ by a Vrije Universiteit User on 02/22/2017
Citations
More filters
Journal Article•DOI•
An Empathy-Driven, Conversational Artificial Intelligence Agent (Wysa) for Digital Mental Well-Being: Real-World Data Evaluation Mixed-Methods Study

[...]

Becky Inkster1, Shubhankar Sarda, Vinod Subramanian•
University of Cambridge1
23 Nov 2018-Jmir mhealth and uhealth
TL;DR: A preliminary real-world data evaluation of the effectiveness and engagement levels of an AI-enabled, empathetic, text-based conversational mobile mental well-being app, Wysa, on users with self-reported symptoms of depression shows promise.
Abstract: Background: A World Health Organization 2017 report stated that major depression affects almost 5% of the human population. Major depression is associated with impaired psychosocial functioning and reduced quality of life. Challenges such as shortage of mental health personnel, long waiting times, perceived stigma, and lower government spends pose barriers to the alleviation of mental health problems. Face-to-face psychotherapy alone provides only point-in-time support and cannot scale quickly enough to address this growing global public health challenge. Artificial intelligence (AI)-enabled, empathetic, and evidence-driven conversational mobile app technologies could play an active role in filling this gap by increasing adoption and enabling reach. Although such a technology can help manage these barriers, they should never replace time with a health care professional for more severe mental health problems. However, app technologies could act as a supplementary or intermediate support system. Mobile mental well-being apps need to uphold privacy and foster both short- and long-term positive outcomes. Objective: This study aimed to present a preliminary real-world data evaluation of the effectiveness and engagement levels of an AI-enabled, empathetic, text-based conversational mobile mental well-being app, Wysa, on users with self-reported symptoms of depression. Methods: In the study, a group of anonymous global users were observed who voluntarily installed the Wysa app, engaged in text-based messaging, and self-reported symptoms of depression using the Patient Health Questionnaire-9. On the basis of the extent of app usage on and between 2 consecutive screening time points, 2 distinct groups of users (high users and low users) emerged. The study used mixed-methods approach to evaluate the impact and engagement levels among these users. The quantitative analysis measured the app impact by comparing the average improvement in symptoms of depression between high and low users. The qualitative analysis measured the app engagement and experience by analyzing in-app user feedback and evaluated the performance of a machine learning classifier to detect user objections during conversations. Results: The average mood improvement (ie, difference in pre- and post-self-reported depression scores) between the groups (ie, high vs low users; n=108 and n=21, respectively) revealed that the high users group had significantly higher average improvement (mean 5.84 [SD 6.66]) compared with the low users group (mean 3.52 [SD 6.15]); Mann-Whitney P=.03 and with a moderate effect size of 0.63. Moreover, 67.7% of user-provided feedback responses found the app experience helpful and encouraging. Conclusions: The real-world data evaluation findings on the effectiveness and engagement levels of Wysa app on users with self-reported symptoms of depression show promise. However, further work is required to validate these initial findings in much larger samples and across longer periods.

327 citations

Journal Article•DOI•
Internet-delivered psychological treatments: from innovation to implementation

[...]

Gerhard Andersson, Nickolai Titov, Blake F. Dear, Alexander Rozental1, Per Carlbring •
University College London1
01 Feb 2019-World Psychiatry
TL;DR: It is argued that ICBT can be viewed as a vehicle for innovation and a significant minority of people do experience negative effects, although rates of deterioration appear similar to those reported for face‐to‐face treatments and lower than for control conditions.

321 citations

Journal Article•DOI•
Digital Mental Health Interventions for Depression, Anxiety, and Enhancement of Psychological Well-Being Among College Students: Systematic Review

[...]

Emily G. Lattie1, Elizabeth C Adkins1, Nathan Winquist1, Colleen Stiles-Shields2, Q. Eileen Wafford1, Andrea K. Graham1 •
Northwestern University1, Rush University Medical Center2
22 Jul 2019-Journal of Medical Internet Research
TL;DR: It is suggested that digital mental health interventions can be effective for improving depression, anxiety, and psychological well-being among college students, but more rigorous studies are needed to ascertain the effective elements of these interventions.
Abstract: Background: College students are increasingly reporting common mental health problems, such as depression and anxiety, and they frequently encounter barriers to seeking traditional mental health treatments. Digital mental health interventions, such as those delivered via the Web and apps, offer the potential to improve access to mental health treatment. Objective: This study aimed to review the literature on digital mental health interventions focused on depression, anxiety, and enhancement of psychological well-being among samples of college students to identify the effectiveness, usability, acceptability, uptake, and adoption of such programs. Methods: We conducted a systematic review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (registration number CRD42018092800), and the search strategy was conducted by a medical research librarian in the following databases: MEDLINE (Ovid), EMBASE (Elsevier), PsycINFO (EbscoHost), the Cochrane Library (Wiley), and Web of Science (Thomson Reuters) from the date of inception to April 2019. Data were synthesized using a systematic narrative synthesis framework, and formal quality assessments were conducted to address the risk of bias. Results: A total of 89 studies met the inclusion criteria. The majority of interventions (71/89, 80%) were delivered via a website, and the most common intervention was internet-based cognitive behavioral therapy (28, 31%). Many programs (33, 37%) featured human support in the form of coaching. The majority of programs were either effective (42, 47%) or partially effective (30, 34%) in producing beneficial changes in the main psychological outcome variables. Approximately half of the studies (45, 51%) did not present any usability or acceptability outcomes, and few studies (4, 4%) examined a broad implementation of digital mental health interventions on college campuses. Quality assessments revealed a moderate-to-severe risk of bias in many of the studies. Conclusions: Results suggest that digital mental health interventions can be effective for improving depression, anxiety, and psychological well-being among college students, but more rigorous studies are needed to ascertain the effective elements of these interventions. Continued research on improving the user experience of, and thus user engagement with, these programs appears vital for the sustainable implementation of digital mental health interventions on college campuses.

287 citations

Journal Article•DOI•
Internet-Based Cognitive Behavioral Therapy for Depression: A Systematic Review and Individual Patient Data Network Meta-analysis.

[...]

Eirini Karyotaki1, Eirini Karyotaki2, Eirini Karyotaki3, Orestis Efthimiou4, Orestis Efthimiou2, Clara Miguel3, Clara Miguel5, Frederic Maas genannt Bermpohl6, Toshi A. Furukawa6, Toshi A. Furukawa7, Pim Cuijpers3, Pim Cuijpers5, Heleen Riper3, Heleen Riper5, Vikram Patel1, Adriana Mira, Alan W Gemmil, Albert Yeung1, Alfred Lange8, Alishia D. Williams9, Andrew Mackinnon9, Andrew Mackinnon10, Anna C. M. Geraedts, Annemieke van Straten3, Annemieke van Straten5, Björn Meyer11, Cecilia Björkelund12, Christine Knaevelsrud13, Christopher G. Beevers14, Cristina Botella15, Cristina Botella16, Daniel R. Strunk17, David C. Mohr18, David Daniel Ebert19, David Kessler20, David Kessler21, Derek Richards22, Elizabeth Littlewood23, Erik Forsell24, Fan Feng1, Fang Wang25, Gerhard Andersson26, Gerhard Andersson24, Heather D. Hadjistavropoulos27, Heleen Christensen9, Iony D. Ezawa17, Isabella Choi28, Isabelle M. Rosso1, Isabelle M. Rosso29, Jan Philipp Klein30, Jason Shumake14, Javier García-Campayo31, Jeannette Milgrom, Jessica Smith32, Jesus Montero-Marin4, Jill M. Newby9, Juana Bretón-López15, Juana Bretón-López16, Justine Schneider33, Kristofer Vernmark26, Lara Bücker34, Lisa Sheeber35, Lisanne Warmerdam, Louise Farrer36, Manuel Heinrich13, Marcus J.H. Huibers3, Marcus J.H. Huibers5, Marie Kivi12, Martin Kraepelien24, Nicholas R. Forand37, Nicholas R. Forand38, Nicky Pugh27, Nils Lindefors24, Ove Lintvedt, Pavle Zagorscak13, Per Carlbring39, Rachel Phillips32, Robert Johansson39, Ronald C. Kessler1, Sally Brabyn, Sarah Perini, Scott L. Rauch29, Simon Gilbody23, Simon Gilbody40, Steffen Moritz34, Thomas Berger2, Victor J M Pop41, Viktor Kaldo24, Viktor Kaldo42, Viola Spek41, Yvonne Forsell24 •
Harvard University1, University of Bern2, VU University Amsterdam3, University of Oxford4, Public Health Research Institute5, University of Wuppertal6, Kyoto University7, University of Amsterdam8, University of New South Wales9, University of Melbourne10, City University London11, University of Gothenburg12, Free University of Berlin13, University of Texas at Austin14, Carlos III Health Institute15, James I University16, Ohio State University17, Northwestern University18, University of Erlangen-Nuremberg19, University Hospitals Bristol NHS Foundation Trust20, University of Bristol21, Trinity College, Dublin22, University of York23, Karolinska Institutet24, Peking Union Medical College25, Linköping University26, University of Regina27, University of Sydney28, McLean Hospital29, University of Lübeck30, University of Zaragoza31, Imperial College London32, University of Nottingham33, University of Hamburg34, Oregon Research Institute35, Australian National University36, The Ohio State University Wexner Medical Center37, Hofstra University38, Stockholm University39, Hull York Medical School40, Tilburg University41, Linnaeus University42
01 Apr 2021-JAMA Psychiatry
TL;DR: In this article, the authors conducted a systematic review and IPD network meta-analysis and estimated relative treatment effect sizes across different patient characteristics through IPD-network meta-regression, and found that both guided and unguided iCBT were associated with more effectiveness as measured by PHQ-9 scores than control treatments over the short term and the long term.
Abstract: Importance Personalized treatment choices would increase the effectiveness of internet-based cognitive behavioral therapy (iCBT) for depression to the extent that patients differ in interventions that better suit them. Objective To provide personalized estimates of short-term and long-term relative efficacy of guided and unguided iCBT for depression using patient-level information. Data Sources We searched PubMed, Embase, PsycInfo, and Cochrane Library to identify randomized clinical trials (RCTs) published up to January 1, 2019. Study Selection Eligible RCTs were those comparing guided or unguided iCBT against each other or against any control intervention in individuals with depression. Available individual patient data (IPD) was collected from all eligible studies. Depression symptom severity was assessed after treatment, 6 months, and 12 months after randomization. Data Extraction and Synthesis We conducted a systematic review and IPD network meta-analysis and estimated relative treatment effect sizes across different patient characteristics through IPD network meta-regression. Main Outcomes and Measures Patient Health Questionnaire–9 (PHQ-9) scores. Results Of 42 eligible RCTs, 39 studies comprising 9751 participants with depression contributed IPD to the IPD network meta-analysis, of which 8107 IPD were synthesized. Overall, both guided and unguided iCBT were associated with more effectiveness as measured by PHQ-9 scores than control treatments over the short term and the long term. Guided iCBT was associated with more effectiveness than unguided iCBT (mean difference [MD] in posttreatment PHQ-9 scores, −0.8; 95% CI, −1.4 to −0.2), but we found no evidence of a difference at 6 or 12 months following randomization. Baseline depression was found to be the most important modifier of the relative association for efficacy of guided vs unguided iCBT. Differences between unguided and guided iCBT in people with baseline symptoms of subthreshold depression (PHQ-9 scores 5-9) were small, while guided iCBT was associated with overall better outcomes in patients with baseline PHQ-9 greater than 9. Conclusions and Relevance In this network meta-analysis with IPD, guided iCBT was associated with more effectiveness than unguided iCBT for individuals with depression, benefits were more substantial in individuals with moderate to severe depression. Unguided iCBT was associated with similar effectiveness among individuals with symptoms of mild/subthreshold depression. Personalized treatment selection is entirely possible and necessary to ensure the best allocation of treatment resources for depression.

271 citations

Journal Article•DOI•
Internet- and mobile-based depression interventions for people with diagnosed depression: A systematic review and meta-analysis

[...]

Koenigbauer Josephine1, Letsch Josefine1, Doebler Philipp2, Ebert David3, Baumeister Harald1 •
University of Ulm1, Technical University of Dortmund2, University of Erlangen-Nuremberg3
01 Dec 2017-Journal of Affective Disorders
TL;DR: IMIs significantly reduce depression symptoms in adults with diagnosed depression at the end of treatment and at follow-up assessments when compared to waitlist conditions, arguing for IMIs to be recommended in depression treatment guidelines.

262 citations

References
More filters
Book•
Statistical Power Analysis for the Behavioral Sciences

[...]

Jacob Cohen1•
University of North Carolina at Chapel Hill1
01 Dec 1969
TL;DR: The concepts of power analysis are discussed in this paper, where Chi-square Tests for Goodness of Fit and Contingency Tables, t-Test for Means, and Sign Test are used.
Abstract: Contents: Prefaces. The Concepts of Power Analysis. The t-Test for Means. The Significance of a Product Moment rs (subscript s). Differences Between Correlation Coefficients. The Test That a Proportion is .50 and the Sign Test. Differences Between Proportions. Chi-Square Tests for Goodness of Fit and Contingency Tables. The Analysis of Variance and Covariance. Multiple Regression and Correlation Analysis. Set Correlation and Multivariate Methods. Some Issues in Power Analysis. Computational Procedures.

115,069 citations

Journal Article•DOI•
The CES-D Scale: A Self-Report Depression Scale for Research in the General Population

[...]

Lenore Sawyer Radloff1•
National Institutes of Health1
01 Jun 1977-Applied Psychological Measurement
TL;DR: The CES-D scale as discussed by the authors is a short self-report scale designed to measure depressive symptomatology in the general population, which has been used in household interview surveys and in psychiatric settings.
Abstract: The CES-D scale is a short self-report scale designed to measure depressive symptomatology in the general population. The items of the scale are symptoms associated with depression which have been used in previously validated longer scales. The new scale was tested in household interview surveys and in psychiatric settings. It was found to have very high internal consistency and adequate test- retest repeatability. Validity was established by pat terns of correlations with other self-report measures, by correlations with clinical ratings of depression, and by relationships with other variables which support its construct validity. Reliability, validity, and factor structure were similar across a wide variety of demographic characteristics in the general population samples tested. The scale should be a useful tool for epidemiologic studies of de pression.

48,339 citations

Journal Article•DOI•
The PHQ-9: validity of a brief depression severity measure.

[...]

Kurt Kroenke1, Robert L. Spitzer2, Janet B. W. Williams2•
Indiana University – Purdue University Indianapolis1, Columbia University2
01 Sep 2001-Journal of General Internal Medicine
TL;DR: In addition to making criteria-based diagnoses of depressive disorders, the PHQ-9 is also a reliable and valid measure of depression severity, which makes it a useful clinical and research tool.
Abstract: OBJECTIVE: While considerable attention has focused on improving the detection of depression, assessment of severity is also important in guiding treatment decisions. Therefore, we examined the validity of a brief, new measure of depression severity.

26,004 citations

Book•
Multiple imputation for nonresponse in surveys

[...]

Donald B. Rubin
01 Jan 1987
TL;DR: In this article, a survey of drinking behavior among men of retirement age was conducted and the results showed that the majority of the participants reported that they did not receive any benefits from the Social Security Administration.
Abstract: Tables and Figures. Glossary. 1. Introduction. 1.1 Overview. 1.2 Examples of Surveys with Nonresponse. 1.3 Properly Handling Nonresponse. 1.4 Single Imputation. 1.5 Multiple Imputation. 1.6 Numerical Example Using Multiple Imputation. 1.7 Guidance for the Reader. 2. Statistical Background. 2.1 Introduction. 2.2 Variables in the Finite Population. 2.3 Probability Distributions and Related Calculations. 2.4 Probability Specifications for Indicator Variables. 2.5 Probability Specifications for (X,Y). 2.6 Bayesian Inference for a Population Quality. 2.7 Interval Estimation. 2.8 Bayesian Procedures for Constructing Interval Estimates, Including Significance Levels and Point Estimates. 2.9 Evaluating the Performance of Procedures. 2.10 Similarity of Bayesian and Randomization--Based Inferences in Many Practical Cases. 3. Underlying Bayesian Theory. 3.1 Introduction and Summary of Repeated--Imputation Inferences. 3.2 Key Results for Analysis When the Multiple Imputations are Repeated Draws from the Posterior Distribution of the Missing Values. 3.3 Inference for Scalar Estimands from a Modest Number of Repeated Completed--Data Means and Variances. 3.4 Significance Levels for Multicomponent Estimands from a Modest Number of Repeated Completed--Data Means and Variance--Covariance Matrices. 3.5 Significance Levels from Repeated Completed--Data Significance Levels. 3.6 Relating the Completed--Data and Completed--Data Posterior Distributions When the Sampling Mechanism is Ignorable. 4. Randomization--Based Evaluations. 4.1 Introduction. 4.2 General Conditions for the Randomization--Validity of Infinite--m Repeated--Imputation Inferences. 4.3Examples of Proper and Improper Imputation Methods in a Simple Case with Ignorable Nonresponse. 4.4 Further Discussion of Proper Imputation Methods. 4.5 The Asymptotic Distibution of (Qm,Um,Bm) for Proper Imputation Methods. 4.6 Evaluations of Finite--m Inferences with Scalar Estimands. 4.7 Evaluation of Significance Levels from the Moment--Based Statistics Dm and Dm with Multicomponent Estimands. 4.8 Evaluation of Significance Levels Based on Repeated Significance Levels. 5. Procedures with Ignorable Nonresponse. 5.1 Introduction. 5.2 Creating Imputed Values under an Explicit Model. 5.3 Some Explicit Imputation Models with Univariate YI and Covariates. 5.4 Monotone Patterns of Missingness in Multivariate YI. 5.5 Missing Social Security Benefits in the Current Population Survey. 5.6 Beyond Monotone Missingness. 6. Procedures with Nonignorable Nonresponse. 6.1 Introduction. 6.2 Nonignorable Nonresponse with Univariate YI and No XI. 6.3 Formal Tasks with Nonignorable Nonresponse. 6.4 Illustrating Mixture Modeling Using Educational Testing Service Data. 6.5 Illustrating Selection Modeling Using CPS Data. 6.6 Extensions to Surveys with Follow--Ups. 6.7 Follow--Up Response in a Survey of Drinking Behavior Among Men of Retirement Age. References. Author Index. Subject Index. Appendix I. Report Written for the Social Security Administration in 1977. Appendix II. Report Written for the Census Bureau in 1983.

14,574 citations

Journal Article•DOI•
Cognitive therapy of depression: pretreatment patient predictors of outcome.

[...]

Kate E. Hamilton1, Keith S. Dobson1•
University of Calgary1
01 Jul 2002-Clinical Psychology Review
TL;DR: In this article, a review examines the role of patient predictors of outcome in cognitive therapy of depression and finds that high pretreatment severity scores are associated with poorer response to cognitive therapy, as are high chronicity, younger age at onset, an increased number of previous episodes, and marital status.

5,556 citations

Related Papers (5)
Computer-based psychological treatments for depression: a systematic review and meta-analysis.

[...]

01 Jun 2012-Clinical Psychology Review
Derek Richards, Thomas Richardson
The PHQ-9: validity of a brief depression severity measure.

[...]

01 Sep 2001-Journal of General Internal Medicine
Kurt Kroenke, Robert L. Spitzer, Janet B. W. Williams
The impact of guidance on Internet-based mental health interventions — A systematic review

[...]

01 Oct 2014-Internet Interventions
Harald Baumeister, L. Reichler, Marie Munzinger, Jihhsuan Lin 
Internet-based vs. face-to-face cognitive behavior therapy for psychiatric and somatic disorders: an updated systematic review and meta-analysis

[...]

01 Jan 2018-Cognitive Behaviour Therapy
Per Carlbring, Gerhard Andersson, Pim Cuijpers, Heleen Riper, Heleen Riper, Erik Hedman-Lagerlöf 
Guided Internet-based vs. face-to-face cognitive behavior therapy for psychiatric and somatic disorders : a systematic review and meta-analysis

[...]

01 Oct 2014-World Psychiatry
Gerhard Andersson, Pim Cuijpers, Per Carlbring, Heleen Riper, Heleen Riper, Erik Hedman 
Frequently Asked Questions (5)
Q1. What are the strengths of the present study?

Among the strengths of the present study was its high power to detect small statistically significant differences between intervention and controls and to yield more precise and robust evidence compared with traditional meta-analyses. 

Two-stage IPD meta-analysis facilitates analysis standardization across the included studies and estimation of outcomes that are not available in the published reports, such as treatment response. 

Studies included in this IPD meta-analysis used measures such as the Center of Epidemiologic Studies–Depression Scale,22 the Beck Depression Inventory I23 or II,24 (hereafter referred to as Beck Depression Inventory) or the 9-item Patient Health Questionnaire25 to monitor change in depressive symptoms severity. 

The current findings indicate that the authors need to treat 8 individuals with depressive symptoms with self-guided iCBT to expect a 50% symptom reduction. 

Although it is beyond the scope of this study, unguided iCBT has several limitations that should be addressed before it is disseminated as part of routine care (eg, high dropout rates, small effects compared with face-to-face and guided internet interventions, and possible participant selection bias).