This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), 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.
Many studies have provided evidence for the effectiveness of Internet-based stand-alone interventions for mental disorders. A newer form of intervention combines the strengths of face-to-face (f2f) and Internet approaches (
The aim of this review was to provide an overview of (1) the different formats of blended treatments for adults, (2) the stage of treatment in which these are applied, (3) their objective in combining face-to-face and Internet-based approaches, and (4) their effectiveness.
Studies on blended concepts were identified through systematic searches in the MEDLINE, PsycINFO, Cochrane, and PubMed databases. Keywords included terms indicating face-to-face interventions (“inpatient,” “outpatient,” “face-to-face,” or “residential treatment”), which were combined with terms indicating Internet treatment (“internet,” “online,” or “web”) and terms indicating mental disorders (“mental health,” “depression,” “anxiety,” or “substance abuse”). We focused on three of the most common mental disorders (depression, anxiety, and substance abuse).
We identified 64 publications describing 44 studies, 27 of which were randomized controlled trials (RCTs). Results suggest that, compared with stand-alone face-to-face therapy, blended therapy may save clinician time, lead to lower dropout rates and greater abstinence rates of patients with substance abuse, or help maintain initially achieved changes within psychotherapy in the long-term effects of inpatient therapy. However, there is a lack of comparative outcome studies investigating the superiority of the outcomes of blended treatments in comparison with classic face-to-face or Internet-based treatments, as well as of studies identifying the optimal ratio of face-to-face and Internet sessions.
Several studies have shown that, for common mental health disorders, blended interventions are feasible and can be more effective compared with no treatment controls. However, more RCTs on effectiveness and cost-effectiveness of blended treatments, especially compared with nonblended treatments are necessary.
Empirical evidence suggests that Internet-based psychological interventions can be used to effectively treat adults, adolescents, and children for various mental disorders such as depression, anxiety, or problematic substance use [
On the other hand, Internet-based interventions may also have disadvantages when compared with face-to-face therapies. For instance, Internet interventions may require certain abilities such as computer and Internet skills, reading and writing skills, and, in comparison with traditional therapy settings, more self-reflection and eloquence when talking about one’s thoughts and feelings. Furthermore, it has been argued that this type of intervention may make it difficult for therapists to adequately react to crisis situations such as suicidality because nonverbal cues are missing as additional information when assessing whether dissociation of suicidal thoughts is possible [
As both face-to-face and Internet-based psychotherapy have advantages and disadvantages, combining the two approaches in a blended treatment might combine the best of two worlds.
As a clear definition of blended interventions is still missing [
Nonblended interventions: Face-to-face (f2f) treatments or stand-alone Internet-treatments only
Blended interventions: Treatment programs that use elements of both face-to-face and Internet-based interventions, including sequential use of both forms of treatment
Integrated blended interventions: Blended treatments where the Internet-based intervention part is arranged as an adjunct to face-to-face programs or vice versa so that face-to-face and Internet-based elements are provided within the same period. In integrated blended interventions, the focus can be either on the face-to-face treatment or on the Internet-based intervention.
Sequential blended interventions: Blended treatments where the Internet-based intervention part is arranged before or after the face-to-face treatment such as within stepped care approaches or aftercare interventions that directly follow the face-to-face intervention.
This systematic review supplies an overview of research into blended interventions for mental health. Specifically, we focus on the following questions: (1) Which blended intervention concepts have been proposed in the researched literature in the treatment of common mental disorders (anxiety disorders, depression, and substance abuse)? (2) In which stage of treatment (such as first step, acute phase treatment, and maintenance phase) do the Internet interventions take place? (3) Which types of problem and target group do the blended interventions focus on? (4) What is the objective in combining face-to-face and Internet-based approaches? and (5) What evidence is there for the effectiveness of blended interventions?
Studies of potential relevance were identified using a systematic search in the MEDLINE, PsycINFO, Cochrane, and PubMed databases. All studies up to December 2015 were included. Searches were performed using keywords indicating face-to-face interventions (“inpatient,” “outpatient,” “face to face,” or “residential treatment”), which were combined with terms indicating Internet treatment (“internet,” “online,” or “web”) and terms indicating mental disorders (“mental health,” “depression,” “anxiety,” or “substance abuse”). We focused on three of the most common mental disorders: depression, anxiety, and substance abuse. The bibliographies of the identified studies revealed additional sources.
Studies were included if they met the following inclusion criteria: (1) the study was on an intervention that was based on both an Internet and a face-to-face treatment element that was either integrated or delivered sequentially, (2) the study involved treatment for adults with depression, anxiety, or substance abuse, and (3) the study was published in English or German. Studies with mere self-help interventions were excluded.
After the initial database search and removal of duplicates, the title and abstract of the remaining studies were rated for the inclusion criteria independently by the first and second author. Interrater reliability was good (kappa, κ=.825,
In November 2016, we started a second search in the database of PubMed, searching for studies citing the studies we had found in the initial search. The extraction strategies as of the initial search were used for those studies, including the independent rating of studies by the first and second authors in the first and second steps.
New quality assessment criteria were created since there are no current guidelines for assessing the quality of blended intervention studies. The quality of each study was rated on five aspects: study design, randomization of study conditions, report of statistics, sample size (studies powered to detect effect sizes of a minimal important difference between blended and nonblended interventions of a priori defined Cohen
Study assessment criteria for scoring.
Aspect | Scoring |
Study design | Which design did the study have? 2=controlled trial, 1=pre-post, 0=case study or unclear. |
Randomization | Were participants randomized to conditions (depending on design)? 2=yes, 0=no. |
Report of statistics | Are relevant statistics reported? 2=mean and standard deviation for outcome measures, effect sizes, and |
Sample size | Was the sample size adequate to detect effect sizes of |
Nonblended control group | Did the design involve a nonblended active control group? 2=nonblended control group with same number of sessions, 1=active control group without assured same number of sessions, 0=no active control group |
See
Of the included studies, 27 were randomized controlled trials (RCTs), 4 were non-RCTs, 5 were pre-post studies without a control group, 4 were case studies, 3 were preliminary evaluation or acceptability studies, and 1 was a qualitative study. See
Of the 44 studies, 8 were rated high quality studies (18%). A control group was involved in 31 (70%, 18/44) of the studies. Twelve studies (27%, 12/44) involved a nonblended active control group with the same number of sessions as the intervention group. Only 6 studies (14%, 6/44) involved a sample of 264 or more. All relevant statistics, including effect sizes, were reported by 17 studies (39%, 17/44). See
Most studies were conducted or planned in the United States (n=12), followed by the Netherlands (n=9) and Germany (n=8). Six studies were conducted in Australia, four in Norway, three in the United Kingdom, and two in Sweden.
The Internet part of the studies’ blended interventions generally used Web-based programs with modules combining techniques such as cognitive, behavioral, and/or emotion-focused interventions, some of them with email support. However, there were two exceptions. One study [
Flowchart of included and excluded studies.
Of the included 44 studies, 20 studies focused on treating depression only, eight focused on anxiety disorders only, and eight focused on substance abuse only. One study described the treatment of comorbid depression and substance abuse; three studies treated both depression and anxiety. The remaining four studies had a transdiagnostic concept involving depression, anxiety, and other mental disorders. See
Following the study selection process, we clustered the studies into the following types of blended care (
Integrated blended interventions with face-to-face focus: These blended interventions are based on an face-to-face intervention that is complemented or partly replaced by an Internet intervention; face-to-face and Internet-based elements are provided within the same period.
Integrated blended interventions with Internet focus: These blended interventions are based on Internet interventions that are partly replaced or complemented by face-to-face sessions; face-to-face and Internet-based elements are provided within the same period.
Sequential blended interventions with Internet, then face-to-face: These blended interventions arrange the Internet intervention part before the face-to-face treatment, such as within stepped care.
Sequential blended interventions with face-to-face, then Internet: These blended interventions arrange the Internet intervention part after the face-to-face treatment as in an aftercare program.
The majority of studies (n=29) used a concept of integrated blended intervention, with face-to-face and Internet-based elements being provided within the same period. Among those, 18 studies focused on the face-to-face intervention, considering the Internet intervention as a replacement of some of the face-to-face sessions or as an adjunct, whereas 11 studies focused on the Internet intervention as the basis of treatment where the face-to-face sessions served as an adjunct, for example, for increasing adherence to the Internet-based modules.
The remaining 15 studies presented sequential blended interventions, arranging the Internet intervention part before or following the face-to-face treatment. Nine of them placed the Internet intervention before the face-to-face treatment either as part of a stepped or matched care program (n=4) or for bridging waiting time for referrals on waiting lists for face-to-face psychotherapy (n=5). Six studies placed the Internet intervention after the face-to-face treatment as an aftercare concept.
The studies’ aims for choosing to use a blended intervention can be classified by the concepts that were used (see
Among the 18 integrated blended interventions with face-to-face focus, seven aimed at delegating some elements of face-to-face therapy to Internet-based cognitive behavioral therapy (iCBT) and thereby, saving clinician time and reducing overall costs [
Five of the 11 integrated blended interventions with an Internet focus aimed at improving the delivery of evidence-based treatment in primary care [
The nine sequential blended interventions that started with the Internet intervention either aimed at bridging waiting time (n=5) with iCBT until face-to-face therapy started [
Aims of blended therapy.
Five of the six sequential blended interventions that started with the face-to-face intervention were designed as aftercare programs aiming at maintaining therapeutic benefits of face-to-faceresidential or inpatient psychotherapy through subsequent iCBT [
See
Of the 44 identified studies, eight were study protocols and 36 had been completed. Among the 36 completed studies, 23 involved a control group. Out of the 23 completed studies with a control group, four studies compared the blended intervention with no treatment, for instance, a waiting-list group [
Among the eight study protocols, two studies compared the blended intervention with TAU without controlling for the number of sessions [
Although cost-effectiveness or cost-savings was in some way considered by almost all 44 studies, only three of the completed studies elaborated on it and evaluated potential cost-effectiveness or cost-savings [
Of the six studies that were delegating some elements of face-to-face therapy to iCBT, three studies were able to show that, by doing that, 50% to 86 % of clinician time could be saved without reducing the therapeutic outcome of depression and anxiety treatment [
Six of the nine studies aiming at supporting face-to-face therapy by delivering Internet elements were able to show that adding Internet elements can lead to lower dropout rates and/or greater abstinence rates of patients with substance abuse compared with stand-alone face-to-face interventions [
Aims of blended therapy of studies with published outcome.
One study aiming at establishing a proactive and long-term approach to the management of chronic mental diseases beyond the acute phase of face-to-face treatment [
Regarding the aim of integrating face-to-face sessions into Internet interventions to maximize effectiveness of iCBT, Sethi et al [
Using blended interventions to improve the delivery of evidence-based treatment in primary care has been successful in acceptability and in reducing symptomatology [
In a study aiming at increasing the flexibility of an Internet intervention by offering complementary face-to-face sessions as needed by the individual participant [
A study aiming at motivating participants to persist with iCBT through face-to-face support [
Of the five studies aiming at bridging waiting time with iCBT until face-to-face therapy starts, two studies [
Two of the completed studies working with a blended stepped care concept did not find a significant superiority of the blended intervention compared with face-to-face [
Two of the five studies with Internet intervention designed as maintenance treatment found a substantially and significantly lower relapse rate compared with access to TAU groups [
One case study aiming at long-term monitoring of patient progress beyond the acute phase of face-to-face treatment, as well as maintaining the therapeutic relationship in the absence of face-to-face contacts in a remote setting [
This study has shown that, in the past few years, a growing number of blended interventions that combine Internet and face-to-face interventions have been developed for common mental health problems. The interventions we found have different concepts and various aims. First results are encouraging and suggest that, compared with stand-alone face-to-face interventions, blended therapy may save clinician time without reducing therapy outcome, can lead to lower dropout rates and/or greater abstinence rates of patients with substance abuse, may help maintain effects of inpatient therapy, and may even increase the effects of psychotherapy, although results are mixed and more research is clearly needed.
Compared with the field of both Internet-based stand-alone treatment and face-to-face interventions, the field of blended interventions is, however, under development and still small. Most aims of the interventions stated in the studies have not been evaluated rigorously. For instance, only 19 out of 36 completed studies were RCTs, and only eight of them ensured comparability by involving a nonblended intervention control group with the same number of sessions [
Several questions remain. For instance, not much is known about the optimal ratio of Internet and face-to-facesessions that would allow costs to be minimized while maintaining or increasing effectiveness. Only one study in our review is moving toward answering this question: In the study by Jacmon et al [
Moreover, although a number of studies stated that increasing the effectiveness of face-to-face psychotherapy is an aim of the blending of treatments with Internet options, evidence for this hypothesis is limited to four studies with an integrated concept [
An interesting question we encountered is what blended treatments exactly entail. The identified studies in our review that explicitly use the term “blended” are the more recently published ones and describe integrated blended treatments, that is, combined treatments that provide face-to-face and Internet-based elements within the same period. Although there is currently no clear definition, this description might be an implicit understanding of blended interventions that is more limiting than our definition. We consider that blended interventions should also include the combination of the Internet-based interventions arranged before or after the face-to-face treatments following clear rules and procedures, such as within stepped care or aftercare that directly follows the acute phase treatment. In light of the numerous studies describing sequential blended interventions, we decided to use this wider definition.
This study has some limitations. As within every systematic review, the risk of selection bias when including relevant studies needs to be considered. However, our use of independent ratings by two of the authors worked against this bias. In addition, publication bias needs to be considered. We did not contact authors for additional data or additional studies, which would have automatically limited the number of studies that could be included. Also, we only reviewed bibliographies from included studies, so we possibly missed studies that were cited in papers that we reviewed but did not include. Furthermore, the types of studies we included were heterogeneous (for instance, we included study protocols). It is possible that more narrow inclusion criteria (eg, randomized controlled studies with nonblended active control groups only) would have produced more information about the effectiveness of blended interventions. However, the number of such studies in the field is yet very small, and future studies are needed to explore whether blended treatments can, for example, be superior compared with nonblended treatments with regard to effect sizes or lower costs.
For further research, it would be of interest to explore effectiveness and cost-effectiveness of blended concepts, especially concerning the optimal balance of face-to-face and Internet interventions. Information on this aspect would help determine which therapy with which theoretical foundation (such as cognitive behavioral therapy or psychoanalysis) is feasible for blended interventions, as well as where the optimal balance of therapy modules lies for individual patients in light of the type and severity of disorder, state of motivation, ability of introspection, and other variables such as age, gender, and computer skills.
To conclude, we have found that several studies have shown that blended interventions are feasible and effective compared with no treatment controls. There are many different kinds of blended concepts that, in every phase of treatment, may offer added value concerning either effectiveness or cost-effectiveness. However, to evaluate the actual benefit of blended concepts for mental health care, more RCTs on effectiveness and cost-effectiveness compared with traditional nonblended psychotherapy are required. Thus, more research is needed, especially concerning disorders for which blended interventions are particularly effective, the amount of face-to-face contact that is needed, and the parts of therapy that can be delegated to the Internet.
Previous psychological research combining Internet and face-to-face psychotherapy.
Outcome of previous psychological research combining Internet and face-to-face (f2f) psychotherapy.
Study quality assessment of blended treatments.
face-to-face
randomized controlled trials
None declared.